Elections and politics

2 Elections and politics Michael Kabuni, Maholopa Laveil, Geejay Milli and Terence Wood Abstract In this chapter we provide an overview of electoral and parliamentary politics in Papua New Guinea (PNG). We cover electoral quality and trends. We examine voter choices. And we study the dynamics of parliamentary politics. As we do this, we pay particular attention to the challenges faced by female candidates. We also look for evidence of improvements stemming from the two most significant changes to postindependence electoral and political rules in PNG: the introduction of limited preferential voting and the Organic Law on the Integrity of Political Parties and Candidates. Little evidence exists that these rule changes have brought improvements. At present, many challenges plague electoral politics and political governance in PNG. Yet there are some signs of potentially positive social changes. There are also good grounds to believe that other rule changes, such as the proposed introduction of Temporary Special Measures to ensure women’s representation, can help. Introduction In the 45 years between independence in 1975 and 2020, Papua New Guinea (PNG) held nine general elections. It is one of only a small number of formerly colonised countries to have remained democratic throughout its history as an independent state. This is a proud record, yet 17 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY democracy has not delivered everything hoped of it – poor governance is a problem in PNG, and most of the country’s citizens have failed to see real development gains. In parts of the country elections have been marred by violence and cheating. Women have rarely been elected. PNG is a parliamentary democracy with a unicameral parliament. At general elections voters vote twice: once to select a candidate to represent their open seat (open seats typically share borders with PNG’s administrative districts) and once to select a candidate to represent their province. Members of parliament (MPs) from both open and provincial seats sit together in the same house of parliament. There are 89 open seats and 22 provincial seats in PNG’s parliament. Over the years, major changes have been made to PNG’s electoral and parliamentary rules in an attempt to improve democratic governance. Most notably, after the 2002 elections, PNG exchanged its first-past-thepost electoral system for limited preferential voting (LPV) and, in 2001, parliament passed the Organic Law on the Integrity of Political Parties and Candidates (OLIPPAC), which was subsequently amended in 2003. In this chapter we provide an overview of electoral and parliamentary politics in PNG. We start with the quality of elections. We then look at electoral trends and patterns. Then we cover female candidates. We subsequently examine why voters vote for the candidates that they do, and the consequences of their choices. Finally, we look at parliamentary politics, focusing on how PNG’s elected representatives interact once in parliament. As we do this, we pay particular attention to rule changes, such as the shift to LPV and OLIPPAC, and ask whether they have changed politics in helpful ways. We also look at proposed future changes such as Temporary Special Measures (TSMs) (rule changes to bring more women into parliament) and discuss their potential. Throughout the chapter we provide facts and figures about elections in PNG. Unless these figures are attributed to another source, the numbers come from the PNG election results database (Wood, 2019).1 1 18 Readers can access these data online at devpolicy.org/pngelections. 2. ELECTIONS AND POLITICS Electoral quality Before we cover electoral politics in detail, it is worth asking two basic questions about the nature of elections in PNG. First, are elections free? Or, to put it another way, are voters safely able to choose who they want to vote for? Second, are elections fair? Are winning candidates really the most popular candidates in their electorates, or do they win by cheating in some way? Figure 2.1 is based on data from the Electoral Integrity Project (Norris and Grömping, 2019). It shows the quality of 121 developing countries’ most recent elections based on expert assessments. (Data for PNG are from 2017.) Each country is represented with a bar. Higher bars mean better run elections. As can be seen, in comparison to the quality of elections globally, electoral quality in the 2017 election in PNG was very poor. Figure 2.1: Electoral quality internationally. Note: Each bar is a country. PNG is highlighted in red. Higher scores are better. Source: Data from the Electoral Integrity Project (Norris and Grömping, 2019). PNG’s poor score stems from a range of issues and we look at each in turn. As we do this, we draw on analysis from the 2017 election. However, while problems were particularly bad in 2017, similar issues have been present in other recent elections, particularly 2002 (Gibbs et al., 2004; Haley, 1997; Haley and Zubrinich, 2013, 2018; May et al., 2011). 19 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY Violence and coercion The 2017 PNG elections triggered violent conflict in places and led to over 200 deaths (Haley and Zubrinich, 2018; Lyons, 2018; Transparency International PNG, 2017). Violence occurred throughout the electoral period, including protracted conflict in some places after results were announced. A clear problem in parts of PNG during the 2017 elections was inadequate security at important times (Haley and Zubrinich, 2018). Even when there was no actual violence, sometimes threats were enough to influence who people voted for. One survey in 2017 found that just over a third of people reported that they were intimidated when voting (Haley and Zubrinich, 2018, p. 60). In parts of the country, polling officials also reported being intimidated and threatened by candidates (Markiewicz and Wood, 2018). Problems with the roll and polling fraud There were major problems with the roll in 2017. There were many more names on the roll than could reasonably be expected based on estimates of PNG’s population (Markiewicz and Wood, 2018). This roll inflation allowed people to vote more than once in some places and ineligible voters to vote in others. In 2017, the roll was not just inflated but also inaccurate: there were many instances of eligible voters not being able to vote because their names were missing from the roll (Markiewicz and Wood, 2018). Polling was also problematic. Proper polling procedures were often neglected. In some instances, polling officials turned a blind eye to, or were complicit in, candidates cheating. In some electorates, supporters of particular candidates were able to take control of polling stations to ensure all votes were cast for their candidates (Haley and Zubrinich, 2018; Transparency International PNG, 2017). Vote buying In 2017, as in earlier elections, in many parts of PNG voters reported being offered money to vote for candidates. Although not everyone sells their vote in PNG, vote buying is common enough to influence election results in some electorates (Haley and Zubrinich, 2013, 2018; Transparency International PNG, 2017). 20 2. ELECTIONS AND POLITICS Better news This is a worrying list of problems, but not the whole story. Most of the worst problems in 2017 came from parts of the Highlands or from large cities. In much of the country the election was peaceful and cheating less common. Many people were free to choose who to vote for. Also, in most of PNG the actual process of vote counting, while slow, has generally been fair (Markiewicz and Wood, 2018; Wood, 2014). Most electoral officials do their jobs well. The electoral process has not completely collapsed in PNG. This has an important ramification for this chapter: it means that we can still meaningfully talk about election results and the nature of electoral competition. Elections are not completely fraudulent. Electoral trends and patterns Voter participation As discussed above, in recent years PNG has suffered significant roll inflation (i.e. more people being on the electoral roll than are eligible to vote). Given how inflated the roll has been, rather than calculate voter participation in the conventional manner, comparing votes cast to the number of names on the roll, a more appropriate calculation is the ratio of votes cast to the estimated voting-age population. Based on this calculation, voter participation is higher in PNG than in PNG’s closest neighbours. Nationally, the average voter participation rate in the most recent election in PNG was 90 per cent, higher than Indonesia (73 per cent) and Solomon Islands (80 per cent). It is only marginally lower than Australia (92 per cent) where voting is compulsory (International IDEA, 2020; and authors’ calculations for Solomon Islands). Trends in voter participation over time in PNG are shown in Figure 2.2. Voter participation reached its highest (we estimate it was over 100 per cent) in 2002. Voter participation nationally masks variation across regions. Variation can be seen in Figure 2.3, using the 2017 election as an example. Voter participation in the Highlands in 2017 was about 120 per cent. Voter participation was lower in the other three regions, with 60 per cent in the Islands Region, 77 per cent in Momase Region, and 69 per cent in the Southern Region. The Highlands Region has seen votes exceed 21 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY the voting-age population in the previous five elections. High voter participation in the Highlands is probably a product of intense electoral competition, roll inflation and issues with electoral fraud. It remains a puzzle, though, why voter participation is comparatively low in the Islands Region. Figure 2.2: Voter participation. Source: For full details on sources and calculations, see Laveil (2020). Figure 2.3: 2017 voter participation by region. Source: For full details on sources and calculations, see Laveil (2020). 22 2. ELECTIONS AND POLITICS Candidates As can be seen in Table 2.1, candidate numbers have grown with each election except 2007 and 2017. The total number of candidates that contested the 1977 election was 878, and by 2017 candidates numbered 3,335. The voting-age population only grew by 293 per cent in the same period. One possible reason for growing candidate numbers is that political parties are of limited electoral importance, and local issues dominate, thus allowing more independents to run, free of the need for party endorsement. However, this does not appear to be the main driver of increasing candidate numbers: in 2017 the share of all candidates who were independents was lower than in 1992. Another plausible explanation for increased candidate numbers is increased fracturing and less cohesion in communities, which prevents cooperation and coordinated support for candidates. This is just a possibility, however, and more research is required in this area. Table 2.1: Candidate numbers over time. Year Seat numbers Candidate numbers Increase (%) Candidates/ electorate Share of independents (%) 1977 109 878 1982 109 1,124 28 8 No data 10 No data 1987 109 1,515 35 14 No data 1992 109 1,644 9 15 74 1997 109 2,373 44 22 73 2002 109 2,875 21 26 43 2007 109 2,748 –4 25 No data 2012 111 3,443 25 31 64 2017 111 3,335 –3 30 57 Source: PNG Elections Database (Wood, 2019). Average candidate numbers vary across regions but, as can be seen in Figure 2.4, all have increased considerably since 1977. Although the Highlands Region has the most candidates in total, since 1992 the Southern Region has had the highest number of candidates contesting per seat. The Islands have the lowest number of candidates per seat. Chimbu provincial seat in 2012 had the highest number of candidates ever (73). Looking at open seats only, Kerema open in 2012 had the highest number of candidates ever (70), while Namatanai open in 1992 and Pomio open in 1977 both only had two candidates. 23 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY Figure 2.4: Average candidates per electorate by region. Source: Data from the PNG Elections Database (Wood, 2019). It is unclear why candidate numbers vary so much between electorates. One obvious explanation for variation would seem to be variation in electorate size. Using roll numbers for 2017, Laveil and Wood (2019) found that the largest electorate (Laigap-Porgera in the Highlands) is six times larger than the smallest electorate (Rabaul in the Islands). However, Wood (2017) found that larger electorates and provinces did not tend to have more candidates on average. Winner vote shares Figure 2.5 shows the lowest, median and highest winning candidate’s vote share in each post-independence election in PNG. At times, candidates have won elections with incredibly low vote shares. The lowest vote share won by a candidate was 6 per cent, which has happened twice: Ben Kerenga Okoro in Sinasina-Yungomugl open seat in 1992, and Bani Hoivo in the Northern provincial seat in 2002. Since the introduction of the LPV electoral system, the minimum winning vote share has lifted somewhat. It was 13 per cent in 2007, 15 per cent in 2012 and 18 per cent in 2017. This is an increase, but it is a small gain considering winners’ vote shares now include second and third preferences too. 24 2. ELECTIONS AND POLITICS Figure 2.5: Minimum, median and maximum winning candidate vote shares over time. Source: Data from the PNG Elections Database (Wood, 2019). There is considerable variation in winning candidate vote shares between electorates. This variation can be seen when comparing maximum, median and minimum winner vote shares in each election (see Figure 2.5). Sir John Guise won the Milne Bay provincial seat with 90 per cent of total votes in 1977. Winner vote shares are typically highest when there are fewer candidates, as evidenced in 2017 when the winning candidate with the highest vote share of 77 per cent was the incumbent prime minister, Peter O’Neill, who contested the Ialibu-Pangia open seat, the seat with the second lowest number of candidates. Incumbent turnover Incumbent turnover (the rate at which sitting MPs lose their seats) has been high in all of PNG’s nine elections. Figure 2.6 shows incumbent turnover in different elections. There is no clear national trend, but the highest incumbent turnover rate was experienced in 2002, when almost three-quarters of sitting MPs lost their seats. Turnover rates have averaged 54.8 per cent: on average, more than half of MPs lose their seats in every election. A few notable politicians such as Sir Michael Somare (East Sepik provincial), Sir Julius Chan (Namatanai open) and Sir Peter Lus (Maprik open) have been successful in consecutive elections. Sir Michael Somare, in particular, was triumphant in all elections before retiring from politics prior to the 2017 elections. 25 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY Figure 2.6: Incumbent turnover over time. Source: Data from the PNG Elections Database (Wood, 2019). Unlike some other aspects of elections, incumbent turnover rates do not vary much between regions: sitting members everywhere have a high chance of losing their seats. In the early 1980s, David Hegarty (1982) noticed a correlation at each election between candidate numbers and the likelihood an MP would lose their seat. Recent statistical analysis by Wood (2017) shows that this relationship can still be found. When more candidates stand, the sitting member is more likely to lose their seat. What remains unclear though is why this relationship exists. Do more candidates stand when they think an MP is weak? Or is it the case that more candidates weaken an MP’s chances of re-election by eating into their support base? Future case study research could shed light on this. Women candidates Since independence, PNG’s parliament has seen a total of only seven women parliamentarians. At present, there are no women in PNG’s parliament. PNG sits alongside Micronesia and Vanuatu as one of the few countries on earth that does not have any female representation in parliament (IPU, 2020). The number of women MPs elected in each postindependence general election in PNG is shown in Figure 2.7. The absence of women in PNG’s parliament does not reflect disinterest on behalf of women candidates. Candidate numbers have risen: 10 women stood in 26 2. ELECTIONS AND POLITICS the 1977 elections; in 2007, 105 women stood; in 2017, 179 women contested the elections (Laveil and Wood, 2019, p. 22; Sepoe, 2002, p. 39). Yet, the number of women elected at each election is not growing. Figure 2.7: Number of women elected in general elections. Source: Data from the PNG Elections Database (Wood, 2019). The impact of LPV PNG has used two electoral systems in its time as an independent nation.2 The first electoral system, used from 1977 to 2002, was a single member district plurality (SMDP) system, commonly known as first-past-the-post. Voters were allowed a single preference under this system. The second electoral system, first used in a 2003 by-election (Reilly, 2006), is the LPV system, which allows voters three preferences. Winning candidates under the LPV are those who earn more than 50 per cent of those ballots that have not yet been exhausted.3 In this section we examine whether changing electoral systems has changed electoral politics in PNG. 2 A third system, similar to LPV, was used in national elections prior to independence. SMDP was adopted at independence on the recommendation of the constitutional commission, which concluded it would be simpler logistically (Reilly, 2002). 3 The LPV electoral system has two counting phases: one in which first preferences are counted, and another where second and third preferences are counted. A winning candidate is one who receives either more than 50 per cent of total valid first preference votes or, failing this, receives more than 50 per cent of those ballots not yet exhausted. A ballot is said to be exhausted once a voter’s first, second and third preference candidates have already been eliminated from the count. In practice a significant number of ballots are eliminated in most electorates, meaning the total ballots counted in the round in which a winner is selected is usually considerably lower than the total number of ballots cast. 27 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY Violence Election-related violence does not appear to have changed much when elections under the SMDP and LPV systems are compared. SMDP elections had been believed to encourage clan-based candidates, with violence a reflection of ethnic differences (Reilly, 2002, 2006). Violence was an issue in SMDP elections and was particularly bad in 2002 (May et al., 2011). The 2007 elections, the first general elections run under LPV, were relatively peaceful. This seemed like an early achievement for LPV. Violence has increased since then, however, with the 2017 elections recording 204 election-related deaths (Haley and Zubrinich, 2018). It is possible that the 2007 improvement was driven largely by an increased security presence in the Highlands in that election. Another possibility is that while LPV initially triggered a reduction in violence as approaches to campaigning changed, the changes were not sustained, and the norm of electoral violence, particularly in parts of the Highlands, ultimately proved resistant to changes in electoral rules. Invalid ballots Figure 2.8 shows invalid ballots for elections with available data. Unfortunately, spoiled ballot data are not available for 2007, the year LPV was first used in a general election. However, comparing numbers in the previous elections with those after 2007 is instructive. Comparing the last election that the SMDP system was used (2002), and the first election under the LPV system for which data are available (2012), invalid ballots increased from 0.7 per cent in 2002 to 2.6 per cent in 2012. The more complex system of LPV where voters have to choose three preferences seems to have increased invalid ballots. Possibly, however, the share of ballots that are spoiled is now decreasing as more voters understand the LPV system: only 2 per cent of ballots were invalid in 2017. 28 2. ELECTIONS AND POLITICS Figure 2.8: Share of votes that were invalid. Source: Data from the PNG Elections Database (Wood, 2019). Years with no bars are missing data. MPs’ electoral mandates One argument made in favour of LPV was that it would increase MPs’ electoral mandates. It was thought that because MPs would win votes from a larger share of their electorates under LPV they would govern in a manner that better served their whole electorate, rather than just their supporters. As the median winner’s vote share line in Figure 2.5 above shows, LPV has clearly increased the typical MP’s mandate in the sense that they have won with the support of more voters. However, as the figure also shows, contrary to a common mistaken belief about LPV, the typical winner in elections since LPV was introduced has not won with more than 50 per cent of the total number of votes cast in their electorate (the actual figure for the median winning candidate is closer to 30 per cent). Nevertheless, the increase in winner vote share under LPV suggests the system is more representative in the sense that MPs are now being elected to parliament with the support of more voters. Even so, persistent high incumbent turnover suggests a continued dissatisfaction with MPs’ performance. LPV does not seem to have left voters much more satisfied with their MPs on average. 29 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY Governance Another claim sometimes made for LPV was that it would improve the quality of governance in PNG. Figure 2.9 shows the average of five governance indicators provided by the World Bank: political stability and the absence of violence, government effectiveness, regulatory quality, rule of law, and control of corruption (World Bank, 2019). In the figure, these are measured in global percentile rankings for years when data are available. Figure 2.9: Quality of governance under SMDP (1996–2006) and under LPV (2007–18). Note: WGI = Worldwide Governance Indicators. Source: Data from World Bank (2019). Although the average appears volatile, a simple comparison suggests PNG’s governance score is marginally better now than it was in 2007 (the first general election with LPV). This would also seem to be a reversal of a previous trend of deterioration. However, a closer look at trends reveals more. In particular, governance started improving in 2005, before LPV was first used in a general election, and the trend of improvement also appears to have reversed again since 2014.4 In 2018, governance was still worse than it had been under SMDP in 1996. If LPV has improved governance in PNG, there is no clear evidence in the available data. 4 There are two plausible alternative explanations for improving governance in this period. First, it coincides with the period of improved economic performance stemming from PNG’s resource boom. It is plausible the resource boom is the source of better governance scores, either because more government revenue afforded improved scope for better services, or because perceptions of governance (the indicators are largely perception based) were influenced by PNG’s rising affluence. Second, improvement may have stemmed from reforms introduced by Sir Mekere Morauta’s 1999–2002 government. 30 2. ELECTIONS AND POLITICS Has LPV helped women? Another hope was that LPV would improve the electoral fortunes of female candidates. At this point, limited data have been an impediment to research on the question of whether LPV has worked to help women candidates or not. When we examine the number of women elected to parliament, as we did in Figure 2.7, it seems clear that LPV has not yet brought any change to the number of female MPs who have been elected. However, it is also true that the average performance of women who stood in the 2017 elections tells a more positive story than the one that emerges if we look at winners alone. Average performance suggests women candidates became slightly more competitive between 2012 and 2017 (Laveil and Wood, 2019). This trend is more encouraging, although, unfortunately, insufficient data exist on women candidates in pre-LPV elections to state for sure that the trend is a result of the change in electoral systems. The belief that LPV might help women candidates hinges on the assumption that women might win by the second and third preferences, or the ‘sori vote’ (the sympathy vote, a concept many female candidates would view as patronising). Empirical analysis of election results suggests female candidates have benefited somewhat more from second and third preference votes than men have (Laveil and Wood, 2019, p. 21). This is encouraging and once again suggests LPV may have helped women candidates. However, gains have been small and there is no evidence LPV has seriously addressed the challenges female candidates face (Baker, 2018). Ultimately, it will need more than three elections to get a clear picture of LPV’s effect on female candidates. Thus far, however, if it has helped women candidates, any benefits of LPV have been modest at best. Women’s political participation and Temporary Special Measures PNG’s constitution strongly advocates for equality for all citizens. The second National Goal and Directive Principle, ‘Equality and Participation’, stipulates that citizens should have equal opportunity to participate in and benefit from the country’s development. The principle makes explicit mention of women’s political participation (Government 31 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY of Papua New Guinea, 1975). However, as we have shown, election results strongly suggest women candidates have not been able to participate equally in elections. Three key impediments serve as common challenges for nearly all women aspiring to political office in PNG: cultural factors, insufficient finance, and violence and intimidation. Being a predominantly patriarchal society with strong cultural norms contributes significantly to low female representation in parliament. Patriarchal societies, particularly in the Highlands, still have conservative attitudes regarding a woman’s ‘place’, which is not in positions of leadership. The challenge is not limited to the Highlands either. During observations of the 2017 national elections in Madang Province, strong patriarchal norms were observed. Many of the women voters interviewed indicated their family and tribal affiliations dictated who they voted for. Voters’ comments often suggested strong opposition to women candidates with the argument that women did not belong in a ‘men’s house’ (parliament). As we highlighted earlier in this chapter, elections in PNG have been marred by electoral corruption and money politics. The costs of money politics, combined with legitimate expenses, make campaigning an expensive exercise. Costs associated with running a successful campaign include, but are not limited to, posters, advertisements, logistics and transportation, security, supplying ‘campaign houses’ with food and beverages, meeting the demands of voters, financing vote buying, hiring sing-sing and supporter groups, and allowances for scrutineers. Funding all of this can easily exhaust candidates who do not possess substantial financial resources. The gendered nature of PNG’s economy makes it harder for women than men to obtain much-needed electoral finances. This puts female candidates at a great disadvantage. Election-related violence and intimidation of female candidates are recurring problems. Policing is usually reserved for polling stations and during counting. With insufficient resources, female candidates are not guaranteed security and are prone to intimidation and violence. Sarah Garap, who contested in 2002 in Simbu Province, faced violence and intimidation from supporters of male candidates (Radio New Zealand, 2002). Mary Kaman, who has run for the Madang provincial seat four times, has spoken of how her supporters and those of other women candidates in Madang were openly intimidated. 32 2. ELECTIONS AND POLITICS Temporary Special Measures – a pathway to change There is need for serious action at the national level to adequately address the issue of women’s representation in parliament. Given all the constraints to women winning elections, introducing some form of TSM has been proposed as an action the government should take if it is serious about including women in national decision-making. TSMs are not new: about 40 countries have already introduced gender quotas for parliamentary elections by constitutional amendment or electoral law. These provisions include reserved seats, which set aside a certain number of places for women (Krook, 2006). Also, in more than 50 countries, quotas requiring that a certain minimum of parties’ candidates for election to national parliament must be women are now stipulated in major political parties’ own statutes (Dahlerup and Freidenvall, 2005). In 2011, the then minister for community development and three-term parliamentarian Dame Carol Kidu spearheaded the push to introduce 22 reserved (regional) seats for women in the national parliament. Elections since the early 1980s had brought disappointing results for women, and Dame Carol was the only woman in parliament at the time. The Equality and Participation Bill was tabled in parliament on 9 September 2011 and passed by 72 votes to 2 on 23 November 2011. This resulted in an amendment of the constitution to include the provision for 22 reserved seats for women, one for each province (Baker, 2014a). However, for the act to be implemented, a constitutional amendment on the Organic Law on National and Local-level Government Elections (amendment no. 2) also had to be passed. This amendment failed to get the two-thirds majority of 73 needed to pass it. Only 58 members voted for the Bill, while 21 members walked out of the chamber and one member voted against it (Elapa, 2012). With this effective shelving of the Equality and Participation Bill, the push for 22 reserved seats was not pursued, even when three women were elected to parliament in the 2012 elections. 33 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY One persistent argument against reserved seats is that it would give women an unfair advantage. According to this argument, if women truly want equality, they should contest elections without special privileges. As shown above, however, electoral contests are already unequal: female candidates face more challenges than men. Despite the failure of the 2011 initiative, PNG does have experience of TSMs at the subnational level. With the finalisation of the Bougainville constitution in 2004, three seats in the Bougainville House of Representatives were reserved for women, along with three seats for ex-combatants (Bougainville has a 39-seat legislature) (Baker, 2014b). The existence of reserved seats in the Autonomous Region of Bougainville has helped increase women’s political participation. The 2020 Bougainville elections yielded positive and promising results for women’s representation, with a total of four women elected to the Bougainville House of Representatives. In addition to the three women elected to reserved seats, 29-year-old activist, social worker and mother Theonilla Roka Matbob was elected to represent the Loro Constituency, which includes the Panguna mine, where fighting in Bougainville’s civil conflict first started in the 1980s (Whiting, 2020). Matbob’s win is significant, as she is only the second woman to win an open seat in Bougainville. When the first woman to win an open seat in Bougainville, Josephine Getsi, won in 2015, her success was celebrated and credited as an indication of the growing understanding of the role women can play in politics (Radio New Zealand, 2016). The 2020 election results suggest this understanding is continuing to grow. More than 40 women contested the 2020 elections in Bougainville (SBS News, 2020). This represents a significant increase from 2015 and is evidence that more women than ever are interested in vying for leadership positions in the region. Francesca Semoso, former deputy speaker for the third Bougainville House of Representatives, was elected twice to the north Bougainville reserved seat (in 2005 and 2015); she took second place in the 2010 elections and third in the 2020 elections for an open seat. She is an ardent advocate for TSMs. Semoso describes reserved seats as a ‘launching pad’ for women politicians who face challenges that men do not often face when they enter politics (Pacific Women, 2016). PNG can learn a lot from the measures Bougainville has taken to increase equal political representation. 34 2. ELECTIONS AND POLITICS What is the way forward for the inclusion of women in politics? Although female electoral success in PNG appears elusive, there are many small victories that must be acknowledged. First, the awareness around TSMs that was brought to the national consciousness in late 2011 as well as the successful use of TSMs in the Autonomous Region of Bougainville. Second, the 2012 victory of three women: Loujaya Toni (Lae open), Delilah Gore (Sohe open) and Julie Soso (Eastern Highlands provincial). Soso’s win was testimony to years of planning that eventually ensured victory in a seat that demands widespread popularity in a particularly patriarchal part of the country. The 45 years of independence and the nine elections that have occurred have brought useful lessons on equal political representation. There are two points that aspiring women politicians, advocates of equal political participation, civil society and international organisations should take into account in their attempts to get more women into parliament. The first point is for individual candidates themselves; the second point relates to the broader collective task of changing electoral rules. Aspiring female candidates need to focus on the long-term planning associated with the challenge of winning elections. Given the disadvantages posed by culture and lack of finance, female candidates must be willing to plan ahead, putting in extra effort. In particular, female candidates need to maximise their financial resources to sustain their campaigns. Candidates also need to get to know their community’s needs by being on the ground and being involved. Winning takes groundwork. When interviewed in 2017, Dame Carol Kidu recalled how she would visit settlement communities in Port Moresby South and villages frequently as part of her groundwork to get a sense of the issues people faced. Research by the National Research Institute also confirms voters stated they valued hanmak as a show of concern with the community’s affairs (Fario et al., 2020). In addition, candidates need to focus on networking and utilising connections with community groups. There are many examples of how women used their networks and connections to help their campaigns. 35 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY Long-term strategising is an arduous journey, and there is no guarantee that the efforts put in over the years will translate into earning a seat in parliament. Yet female candidates can improve their chances by campaigning strategically and over the long-term. The collective task associated with improving women’s representation involves lobbying and increasing parliamentary support for TSMs. To do this, reserved seats for women must be aggressively pursued. Despite its rejection by many in political office, reserved seats can be lobbied for by sitting parliamentarians, civil society and women’s groups. Leaders’ views can be changed. Public awareness of the nature of TSMs also needs to be raised. TSMs are not unequal as is sometimes claimed. They are designed to correct existing inequalities in electoral competition and to boost the equality of political participation until other changes occur. When these changes do occur, TSMs can be discontinued, hence the ‘temporary’ nature of these quotas. Another approach would be to push for party gender quotas. The Integrity of Political Parties and Candidates Commission registrar, Dr Alphonse Gelu, has stated that proposed revisions to OLIPPAC include a provision that 20 per cent of a political party’s nominated candidates must be female candidates. If the revised law is passed, it will be a big improvement (Radio New Zealand, 2020). Whatever the mechanism, the aim must be that through collaboration, consultation and lobbying, PNG moves towards measures that will allow women more representation in parliament. Voter choice and clientelism As we have discussed, voters are not always free to choose who they vote for in PNG. Sometimes they are forced or intimidated into voting for a candidate. This is a real problem in parts of PNG. However, when they were surveyed after the 2012 election, the majority of surveyed voters said they were free to choose who to vote for. And in 2017 the majority of voters said they were free to choose except in the Highlands Region and National Capital District (Haley and Zubrinich, 2018, p. 61). Voter coercion is a real problem in PNG, but enough voters appear free to choose for the question ‘why do voters vote for the candidates they vote for?’ to be an important one for study. 36 2. ELECTIONS AND POLITICS What are voters looking for when they vote? Before going any further, one point needs to be clarified: not all voters think or behave in the same way. For every rule there is an exception. In this section we are going to talk about the choices of the majority of voters. There is an obvious reason for doing this: in elections, in most circumstances, majorities decide the results. In 1989, an academic who ran a large voter survey in PNG wrote the following about voter choice based on the survey findings: ‘Recurring themes are the overwhelming importance of local factors in candidate evaluation, the corresponding insignificance of party, and the virtual absence of [national] issues in the decision calculus’ (Saffu, 1989, p. 15). Nearly 40 years later, the authors of another large survey-based election study summarised their findings about voter choice by stating that: ‘voter choice remains driven by local politics, personal interest and personal gain, and not by party policies’ (Haley and Zubrinich, 2018, p. 41). These survey-related findings are corroborated by the findings of case study research (e.g. Anere, 1997; Osi, 2013; Standish, 2007).5 When voters are free to vote in PNG, most make choices based on local factors. They do not normally choose their preferred candidates based on national politics or national issues. Typically, voters vote for candidates who they think will help them, their family or their community if they are elected (Wood, 2018). Under these circumstances, parties – as the quotations from Haley and Saffu suggest – are much less important than they are in many democracies. In PNG very few voters vote for candidates because they like their party’s national policy platform or share its ideology (Okole, 2001). As we discuss in the section on parliamentary politics, parties are still important in government formation. And parties can still play a role in elections in PNG. In the past, parties sometimes provided candidates with a network of contacts who could help win votes (e.g. Filer, 1996). In more recent years, parties have sometimes served as vehicles through 5 Many of the best qualitative studies on elections in PNG, along with Yaw Saffu’s quantitative work, can be found for free online in a series of downloadable books on PNG elections. To access these books, go to pacificinstitute.anu.edu.au/resources, then scroll down to ‘PNG Elections’. This is a fantastic resource if you are interested in elections in PNG. 37 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY which powerful politicians provide resources to aspiring candidates. But it has been rare for voters to vote for candidates simply because they like their party. Interestingly, in 2017 it seems as if voters were willing to vote against a party – the People’s National Congress (PNC) – to punish it for what they saw as poor performance in government. Particularly, voters appeared to show their displeasure with PNC candidates by not giving them second and third preference votes, and this caused a number of PNC candidates to lose electoral contests (Laveil and Wood, 2019). This ‘protest’ vote is interesting and may be evidence that voters in PNG are starting to think about politics in different ways. However, the protest only involved second and third preferences, not first preferences, and it may be a one-off, stemming from unusual circumstances. For now, most of the time, most voters do not normally vote along party lines or consider national issues. How do people choose who to vote for? If most voters do not vote for candidates on the basis of national issues or on related grounds such as their party, how do they choose who to vote for? Researchers have identified a number of different influences. Some of these are stronger in some parts of PNG than in others and some influences are becoming more or less important with time. Often people vote for candidates who they are in some way related to (colloquially, their wantoks). In parts of the country this can be a central organising principle. In some parts of the Highlands, for example, voters are expected to vote for the candidate from their haus lain (clan), and entire communities will vote as a bloc for an individual candidate. Often, however, when people vote for relatives or people who they share similar ties with, they are making a well-considered calculation. All candidates promise voters they will help them if elected, but not all candidates follow up on those promises. Rules of reciprocity within extended families are not always honoured, but they can be a strong cultural force. This means candidates are often more likely to help supporters who are their wantoks if they are elected. As a result, voters often vote for wantoks because they think they can trust them to help if they win (Okole, 2002, 2005; Wood, 2016). In other instances, voters 38 2. ELECTIONS AND POLITICS simply feel obliged to vote for relatives: a voter from Central Province told one of the authors that they liked LPV because they could vote for all three of their cousins and keep each happy. Voters sometimes vote for relatives who have no chance of winning simply for the sake of maintaining good relations with the relatives in question. Another factor influencing voters’ choices can be support from community groups and recommendations from influential community figures. Strong candidates often campaign in part through support of this kind. Voters are more likely to trust a candidate if they know and trust their main, influential supporters (Filer, 1996; Osi, 2013). Beyond relational ties, another way a voter can get a sense of whether a candidate will keep their promises and help them if elected is if a candidate has helped in the past. This is easy if the candidate they are considering voting for is the sitting MP (has the MP helped in the last five years?), but it can also be a strategy for appraising candidates who are standing for the first time. Candidates can win voter loyalty by having helped previously with community projects or with families’ needs. Closely associated to voters’ considerations of whether a candidate will help in the future based on their past track record of helping is the desire among some voters for some form of hanmak or tangible material assistance. We have already discussed this in the context of female candidates, but the calculation is present regardless of a candidate’s gender. Sometimes voters use giving of this nature as a guide as they assess candidates’ longer histories and how helpful they are (Anere, 1997; Osi, 2013; Wood, 2016). At other times, giving, particularly during campaign periods, primarily serves as a form of vote buying or money politics (Haley and Zubrinich, 2018). Clientelism and politics The fact that voters appraise candidates based on their likelihood of helping directly, rather than on national issues, is not unique to PNG. It is a key feature of politics in many countries. Political scientists use the term ‘clientelism’ to refer to politics that is based on local, personal benefits rather than national issues or policies (Hicken, 2011). 39 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY Figure 2.10: Clientelism in PNG and internationally. Source: Varieties of Democracy (Coppedge et al., 2020). Figure 2.10 uses data from a large international study of countries’ politics (Coppedge et al., 2020). It charts countries based on how strongly clientelist their politics are. Each bar is a country. Higher bars are countries with more clientelism. All the world’s countries with data are included (179 countries in total). The score is an average across the years 2015–19. PNG is shown in red. PNG has one of the highest bars on the chart, reflecting the fact that it is one of the world’s most clientelist countries. However, the figure also shows that PNG is not unique – clientelism is prevalent in many other countries too. Sometimes in discussions of voting in PNG, people blame voters for being selfish and choosing their own personal benefit over the national good. However, this view is unfair. Voting for local or personal support is a reasonable way for voters to vote in PNG. In rural areas in particular, voters have immediate material needs. They also receive very little or no support from the state. And they have never experienced politics where national policy changes have had a big impact on their lives. To further complicate matters, national change requires national political movements (to really change a country, a cohesive majority of MPs is needed in parliament). Genuinely national political movements do not exist in PNG. Under these circumstances, a rational voter wanting to see any positive change emerge from an election has good grounds to vote for someone who will help locally. This is the best they can hope for (Haque, 2012; Wood, 2016). 40 2. ELECTIONS AND POLITICS There is a problem, however, with voters choosing candidates based on whether they will provide direct, local support. Voting this way selects and incentivises most MPs to focus on delivering support to individuals and local communities, and to neglect national governance. This is not the only political problem PNG faces, but the dynamic does contribute to poor political governance. Voters choosing who to vote for on the basis of whether they will help locally has other impacts too. For example, clientelism has almost certainly contributed to the rise of MP-influenced electorate funds such as the District Services Improvement Program – funds that tend to be an inefficient way of delivering services (Howes et al., 2014). A challenge for PNG in the future will be finding a way of building genuinely national, issues-based politics. There is potential for this. New social movements and activists are starting to grow in PNG. The protest vote against the government in 2017 also suggests that voters may be increasingly willing to vote on national issues. At the same time, however, other forces, in particular the influence of money on politics and people’s electoral choices, threatens to push the other way, away from issueoriented national politics. Parliamentary politics Coalition governments in PNG Since the first post-independence elections in 1977, no political party has had the majority to form the government on its own in PNG, either after a national election or a vote of no confidence. All governments have been coalition based. Reflecting the clientelist nature of PNG’s politics, political parties in PNG have no clear ideological divides and, as we discussed in the section on voter choices, voters rarely vote along party lines. Parties largely operate as parliamentary factions, especially after elections, acting in the hope of forming government. Party allegiance among politicians is fluid. Parties are often based around personal ties and the financial assistance party strongmen provide to candidates (Laveil and Wood, 2019). 41 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY Governing coalitions in PNG’s parliament are not normally based on similarities in parties’ policies. For instance, despite having very different policies on education, Pangu Pati and the PNC formed a coalition government following a vote of no confidence in 2019 (Kabuni, 2019c). Instead, MPs and parties join coalitions with the hope of attaining positions of power, such as ministerial roles. The Organic Law on the Number of Ministers sets the maximum number of ministers as a quarter of the total number of MPs in the legislature, and the minimum at 18 (Government of Papua New Guinea, n.d.). This translates into a maximum of 28 ministerial portfolios as there are 111 MPs. The ViceMinisters Act 1994 provides 12 vice-ministers to assist the ministers (Government of Papua New Guinea, 1994). In addition, section 118 of the constitution provides for the creation of permanent parliamentary committees, without prescribing the number of committees (Government of Papua New Guinea, 1975). The chairmen of the committees are MPs from the government side who are not ministers, while deputy chairmen come from the Opposition. After the 2017 elections, 17 permanent committees were formed for the duration of the term. MPs can also be given positions on the boards of various state-owned companies and authorities. With the exception of the roles of committee deputy chairmen, all of the above positions are reserved for MPs in the government. These positions serve as an incentive to parties and their MPs to join the government. Figure 2.11: MPs and ministers by party in 2019. Source: Data from Kabuni (2019b). 42 2. ELECTIONS AND POLITICS Parties with more MPs in a governing coalition usually have more portfolios. Figure 2.11 shows how ministerial portfolios were distributed when James Marape replaced Peter O’Neill as the prime minister in May 2019, and illustrates this point. Political instability in PNG A major consequence of coalition governments that are office oriented and not based on shared beliefs about policy is political instability. Successive votes of no confidence in PNG’s parliament have led to changes in the position of the prime minister during every electoral term from 1977 until the Somare government completed the 2002 to 2007 term. As Figure 2.12 shows, between 1975 and 2020, even though PNG has had only nine national elections, there were 17 changes to the prime minister’s position. Governing coalitions are rearranged with every vote of no confidence. Sometimes, changes see the cabinet replaced in its entirety. Ministerial reshuffles further add to instability, as prime ministers attempt to secure support for their position by replacing individual ministers mid-term (Ivarature, 2016). Figure 2.12: Length of prime ministers’ tenure. Source: Adapted from Kabuni (2018b). Additions were made for 2019. 43 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY In the 2000s, the prime minister’s position became more stable. Michael Somare became the first prime minister to complete a parliamentary term between 2002 and 2007, and he formed government again in 2007 after re-election. His government lasted until 2011.6 Peter O’Neill, who replaced Somare as prime minister in 2011, formed government after the 2012 elections and again in 2017. Somare and O’Neill are the only prime ministers to complete their terms and successively form government in subsequent elections. James Marape replaced Peter O’Neill as the prime minister in 2019 when O’Neill resigned in the face of a vote of no confidence (Kama, 2017; May, 2020). Engineering political stability OLIPPAC was passed in 2001 and amended in 2003 (Government of Papua New Guinea, 2003). It contained a number of provisions designed to promote political stability and parliamentary integrity. Section 63 stipulates that the party that has the most MPs after a national election will be invited by the governor-general to form government. The aim of section 63 is to ensure that a prime minister is appointed after a general election in an orderly manner with a direct relationship to the way voters expressed their wishes (Gelu, 2005). Only if the largest party’s nominated candidate failed to win a majority of MPs’ votes would the process be thrown open to other contenders, enabling any group that could pull together enough support to form government. This requirement was intended to prevent lobbying for government formation, as that had given rise to corruption and bribery in the past, as even parties with small numbers competed to form government (Okole, 2012). However, there is no provision in the OLIPPAC requiring the party with the largest numbers to form the government if the prime minister is removed mid-term. Peter O’Neill, who replaced Somare as prime minister in 2011, had only five MPs in his PNC party. The PNG party had 25 MPs in the same coalition when O’Neill was elected prime minister (Kabuni, 2019a). OLIPPAC was also enacted with the aim of bringing party discipline and political stability. Central to this were provisions that required: 6 The replacement of Somare by O’Neill occurred, it should be noted, under highly irregular circumstances, and in a manner that appears to have violated PNG’s constitution. For a full discussion, see May (2017, 2020). 44 2. ELECTIONS AND POLITICS • MPs to give substantive reasons for leaving their parties, and the Ombudsman Commission to subsequently investigate these reasons. This was aimed at preventing MPs from joining other parties as this usually led to a vote of no confidence (sections 57–59). • All MPs in a political party to remain a single cohesive body in making decisions such as to determine who will be prime minister and in supporting or opposing the prime minister during a vote of no confidence (section 65). If MPs voted for the incumbent prime minister at the time of their nomination, they were required not to vote against the prime minister in any subsequent votes of no confidence. MPs could, however, abstain from voting against the incumbent prime minister they voted for originally in a vote of no confidence. Votes of no confidence and the OLIPPAC As can be seen in Figure 2.12 above, it does appear as if no confidence motions decreased after OLIPPAC was introduced. However, the effectiveness of OLIPPAC has generated considerable debate (Okole, 2012). Some scholars have claimed that OLIPPAC has had limited success (Sepoe, 2005). Others have attributed Somare’s nine-year rule from 2002 to 2011 to OLIPPAC (Fairweather, 2019). In our view, recent history provides little or no evidence that OLIPPAC has played an important role in increasing the longevity of governments. The executive government in a parliamentary democracy derives its legitimacy from the legislature and, in theory, remains accountable to the legislature. One of the ways in which the legislature exercises control over the executive is through votes of no confidence. In PNG, this process is provided for by section 145 of the constitution. In 2010, the Supreme Court ruled the OLIPPAC provisions regarding changing parties and votes of no confidence unconstitutional for, among other things, placing unreasonable restrictions on the rights and freedoms of MPs (Kabuni, 2018b). However, even before these provisions were ruled unconstitutional, MPs were moving from one party to another. Although the Ombudsman Commission was empowered to investigate these changes under OLIPPAC, it failed to do so (Gelu, 2005). 45 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY Peter O’Neill’s completion of the 2012 to 2017 term is even more interesting as it was achieved after the parts of OLIPPAC intended to enhance political stability were ruled unconstitutional in 2010. The next section looks at possible explanations for this. Peter O’Neill’s term: 2012–17 The first factor that helped prevent Peter O’Neill from being removed as prime minister after the 2012 elections was the extension of the grace period. A grace period is a period in which the prime minister cannot be removed through a vote of no confidence after their election, either following a national election or after a vote of no confidence. At independence, the grace period was six months. In 1991, this was extended to 18 months. In 2012, the O’Neill government amended the constitution to increase the grace period after the election of the prime minister to 30 months (Radio New Zealand, 2015). There is also a rule that stipulates that if a vote of no confidence is instituted within 12 months before the fifth anniversary of the date fixed for the return of writs at the previous general election, parliament must be dissolved and national elections conducted. Because of the high incumbent turnover rate (as discussed in the section on electoral trends), MPs are usually not enthusiastic about votes of no confidence during these final 12 months. There are 60 months in a parliamentary term of five years. Because votes of no confidence are rarely instituted within the 12 months before the next election, and another 30 months was covered by the grace period (a total of 42 months), during O’Neill’s term the Opposition only had a short 18-month window to challenge the prime minister. Also, during O’Neill’s tenure, the minimum parliamentary sitting days in a year were reduced from 63 to 40 (Radio New Zealand, 2015). This left only 60 days in the 18-month window. The government then used tactics such as adjourning the parliament when there were threats of a vote of no confidence to stay in power. One such adjournment in 2016 was made as the final 12 months before the fifth anniversary of the date fixed for the return of writs at the 2012 general election drew near. The next parliamentary sitting was scheduled to fall well within the grace period, eliminating any chance of conducting a vote of no confidence without dissolving the parliament. However, the Supreme Court then ruled that 46 2. ELECTIONS AND POLITICS both this adjournment and the increase of the grace period to 30 months were invalid (Radio New Zealand, 2015). The parliament met but the prime minister successfully overcame the vote of no confidence. The second important factor that helped secure O’Neill in office was the increased use of the District and Provincial Services Improvement Program (DSIP/PSIP) funds. MPs have considerable discretion over the spending of these funds. They are a powerful tool that the MP can use to generate political support within their electorate. During O’Neill’s time as prime minister, these funds were greatly increased. In addition, MPs in the Opposition complained that the O’Neill government deliberately withheld their DSIP and PSIP funds, while releasing the funds to MPs who supported the government. For instance, Opposition MPs Basil and Juffa, who were critical of the O’Neill government in 2016, claimed that they did not receive their full share of the funds (Kabuni, 2018a). O’Neill was not the first prime minister to use DSIP and PSIP funds in this way (Ketan, 2007); however, O’Neill appears to have been particularly astute in his use of this tool and was able to further strengthen his governing coalition. MPs had a clear financial incentive not to cross the floor. The third factor aiding O’Neill’s tenure was the politicisation of some parts of the public service (Kama, 2017; May, 2017, 2020). The executive arm of the government secured control of appointments of senior civil service positions (Kama, 2017). This can be seen in the appointment of the police commissioners. In 2014, the O’Neill government dismissed Tom Kulunga when he signed the necessary documents to effect the arrest of Prime Minister Peter O’Neill in relation to alleged illegal payments made to Paul Paraka lawyers. Geoffrey Vaki replaced Tom Kulunga. Geoffrey Vaki prevented the arrest of the prime minister, but when he became embroiled in a contempt of court case in the same year for conspiring to prevent that arrest, the O’Neill government appointed Gary Baki again as the police commissioner. Gary Baki then again prevented the arrest of the prime minister. Arguably, were it not for political appointments in key roles, O’Neill’s term would have been truncated sooner on legal grounds. Finally, it is possible that the resource boom from about 2003 to 2014 helped both O’Neill and Somare stay in office by providing funding for increasing constituency funds, and more generally by providing increased resources that could be used to help cohere governing coalitions. 47 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY The consequences of political instability Fraenkel et al. (2008) attribute political instability in Melanesia to MPs’ belief that access to elected office is a major avenue to power and wealth, and the outcome of a struggle by those MPs not in control of the resources to oust those in power. This constant struggle for power diverts parliament from its main task, which is to debate and devise legislation for the country. Instead, the MPs are occupied by the desire to form government or maintain their portfolios in the existing governing coalition. MPs in the Opposition await the expiry of the grace period and plot to replace the government. OLIPPAC was intended as a means of legislating political stability, but key provisions have been struck down by the courts and it is unclear whether OLIPPAC achieved its goals. Rather, recent political stability appears to have stemmed from recent prime ministers being particularly astute in their use of DSIP and PSIP funds, alongside lengthened grace periods and the proroguing of parliament. As O’Neill’s defeat in 2019 shows, these tools do not render sitting prime ministers invulnerable. Whether the political stability seen between 2002 and 2019 will be the new normal or whether PNG will return to more frequent changes of prime minister remains to be seen. Yet one fact is clear, the tools used by prime ministers in their quest to stay on top have further detracted from the important role that parliament should be playing. Conclusion In this chapter we have described democratic politics as it currently exists in PNG. We have highlighted a number of problems. We have also argued that ambitious reforms to PNG’s electoral system and its parliamentary rules have largely been ineffective: neither LPV nor OLIPPAC has delivered as hoped. And yet, we have also shown that a different type of reform – TSMs – could help in increasing the number of women elected to PNG’s parliament. In the case of TSMs, there is a clear link between the law change and the desired outcome. There is also a precedent: such approaches have succeeded in other countries and have had a very promising start in the Autonomous Region of Bougainville. We also have demonstrated that many of PNG’s electoral and political problems do not stem from the shortcomings of voters. Rather they stem from structural issues, such as the clientelism that characterises the relationship between voters and the state in PNG, and the unstable 48 2. ELECTIONS AND POLITICS dynamics of parliamentary coalition politics. Such structural problems may be hard to shift. They would require voters, whose needs are immediate and significant, to forego the short-term assistance currently on offer, and vote in search of something less tangible and less likely to be achieved: genuine national change. Change would also require sustained parliamentary cooperation from reformers. All of this would have to occur in the face of growing problems, such as electoral fraud and the influence of money on politics. Nevertheless, there is some cause for hope: other countries have changed the dynamics of their electoral politics through concerted efforts from both political reformers and civil society. And in PNG at present, both reformers and change-oriented civil society groups do exist. It would be naive to claim that democracy in PNG has delivered all that was hoped of it. Yet it would be unduly pessimistic to abandon hope in the democratic process. PNG’s future is yet to be written. There is scope for positive change in the country’s politics. Democracy is messy, but its future in PNG is still one of potential. Acknowledgements We are very grateful to Stephen Howes for detailed and helpful comments on earlier drafts of this chapter. We would also like to thank Dr Kerryn Baker for her insightful comments and assistance. We are similarly grateful to Ron May for his review and insights. We are also grateful to participants in two virtual University of Papua New Guinea/The Australian National University workshops for their feedback. As always, any errors or omissions are ours alone. References Anere, R. (1997). Milne Bay Provincial: Independents versus parties. In R. May and R. Anere (Eds), Maintaining democracy: The 1997 elections in Papua New Guinea (pp. 88–97). University of Papua New Guinea. Baker, K. (2014a). Explaining the outcome of gender quota campaigns in Samoa and Papua New Guinea. Political Science, 66(1), 63–83. doi.org/10.1177/ 0032318714531428. Baker, K. (2014b). Women’s representation and the use of reserved seats in Bougainville. SSGM In Brief Series, 2014/2. dpa.bellschool.anu.edu.au/sites/ default/files/publications/attachments/2015-12/SSGM_IB_2014_2_0.pdf. 49 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY Baker, K. (2018). Women’s representation and electoral system reform in Papua New Guinea: The limitations of limited preferential voting. Asia & the Pacific Policy Studies, 5(2), 208–19. doi.org/10.1002/app5.235. Coppedge, M., Gerring, J., Lindberg, S., Skaaning, S-E., Teorell, J., Altman, D., Bernhard, M., Fish, M. S., Glynn, A., Hicken, A., Knutsen, C. H., Krusell, J., Lührmann, A., Marquardt, K. L., McMann, K., Mechkova, V., Olin, M., Paxton, P., Pemstein, D., Pernes, J. … Wilson, S. (2020). V-Dem Dataset v10. Varieties of Democracy (V-Dem) Project. doi.org/10.23696/vdemds20. Dahlerup, D. and Freidenvall, L. (2005). Quotas as a ‘fast track’ to equal representation for women. International Feminist Journal of Politics, 7(1), 26–48. doi.org/10.1080/1461674042000324673. Elapa, J. (2012, 23 February). Reserved seats for women bill shelved. The National. www.thenational.com.pg/reserved-seats-for-women-bill-shelved/. Fairio, M., Kaut Nasengom, S. and Keimelo, C. (2020). Challenges and critical factors affecting women in the 2017 national elections: Case of Lae and Huon Gulf. National Research Institute Discussion Papers, 178. www.pngnri.org/images/ Publications/DPNo178_Challenges_and_critical_factors_affecting_women_ in_the_2017_national_elections-_Case_of_Lae_and_Huon_Gulf_.pdf. Fairweather, K. (2019). Farewell white man. Jabiru Publishing. Filer, C. (1996). ‘Steak and grease’: A short history of political competition in Nuku. In Y. Saffu (Ed.), The 1992 Papua New Guinea elections: Change and continuity in electoral politics (pp. 168–218). The Australian National University Department of Political and Social Change. Fraenkel, J., Regan, A. and Hegarty, D. (2008). The dangers of political party strengthening in Solomon Islands. SSGM Working Paper Series, 2008/2. dpa.bell school.anu.edu.au/sites/default/files/publications/attachments/2015-12/08 _02wp_Fraenkel_0.pdf. Gelu, A. (2005). Failure of the Organic Law on the Integrity of Political Parties and Candidates (OLIPPAC). Pacific Economic Bulletin, 20(1), 83–97. devpolicy. org/PEB/2019/06/09/the-failure-of-the-organic-law-on-the-integrity-ofpolitical-parties-and-candidates-olippac/. Gibbs, P., Haley, N. and McLeod, A. (2004). Politicking and voting in the Highlands: The 2002 Papua New Guinea national elections. State, Society and Governance in Melanesia Discussion Paper, 1. hdl.handle.net/1885/42056. Government of Papua New Guinea. (1975). Constitution of the Independent State of Papua New Guinea. www.paclii.org/pg/legis/consol_act/cotisopng534/. 50 2. ELECTIONS AND POLITICS Government of Papua New Guinea. (1994). Vice-Ministers Act. www.paclii.org/ pg/legis/consol_act/va1994173/. Government of Papua New Guinea. (2003). Organic Law on the Integrity of Political Parties and Candidates. www.paclii.org/pg/legis/consol_act/olotioppac 2003542/. Government of Papua New Guinea. (n.d.). Organic Law on the Number of Ministers. www.paclii.org/pg/legis/consol_act/olotnom353/. Haley, N. (1997). Election fraud on a grand scale: The case of the KorobaKopiago Open Electorate. In R. May and R. Anere (Eds), Maintaining democracy: The 1997 elections in Papua New Guinea (pp. 123–39). University of Papua New Guinea. Haley, N. and Zubrinich, K. (2013). 2012 Papua New Guinea general elections domestic observation report. Cardno Emerging Markets. Haley, N. and Zubrinich, K. (2018). 2017 Papua New Guinea general elections: Election observation report. Department of Pacific Affairs. Haque, T. A. (2012). The influence of culture on economic development in Solomon Islands: A political-economy perspective. SSGM Discussion Paper, 2021/1. dpa. bellschool.anu.edu.au/sites/default/files/publications/attachments/2015-12/ 2012_03_tobias_a_haque_0.pdf. Hegarty, D. (1982). Electoral vulnerability in Papua New Guinea. Political studies seminar. University of Papua New Guinea. Hicken, A. (2011). Clientelism. Annual Review of Political Science, 14(1), 289–310. doi.org/10.1146/annurev.polisci.031908.220508. Howes, S., Mako, A., Swan, A., Walton, G., Webster, T. and Wiltshire, C. (2014). A lost decade? Service delivery and reforms in Papua New Guinea 2002–2012. The National Research Institute and the Development Policy Centre. International IDEA. (2020). Voter turnout database. www.idea.int/data-tools/ data/voter-turnout. IPU. (2020). Monthly ranking of women in national parliaments. data.ipu.org/ women-ranking?month=10&year=2020. Ivarature, H. (2016). Unravelling parliamentary instability: Ministerial durations in Papua New Guinea, 1972–2012. International IDEA Discussion Paper, 18. www.idea.int/publications/catalogue/unravelling-parliamentary-instabilityministerial-durations-papua-new-guinea. 51 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY Kabuni, M. (2018a, 16 July). Does political stability make governments unaccountable? Devpolicy Blog. devpolicy.org/does-political-stabilityconsolidate-irresponsible-government-png-2012-2018-20180716/. Kabuni, M. (2018b). Explaining the paradox of internal party instability and political stability in PNG: 2012–2017. Journal of South Pacific Culture and Philosophy, 13, 50–63. Kabuni, M. (2019a, 9 May). Changing prime ministers in PNG: Does party size matter? Devpolicy Blog. devpolicy.org/changing-prime-ministers-in-pngdoes-party-size-matter-20190509/. Kabuni, M. (2019b, 19 June). PNG’s fluid politics: Winners and losers from O’Neil to Marape. Devpolicy Blog. devpolicy.org/pngs-fluid-politics-winnersand-losers-from-oneill-to-marape-20190619/. Kabuni, M. (2019c, 23 September). Assorted MPs, assorted parties: James Marape’s coalition. Devpolicy Blog. devpolicy.org/assorted-mps-assorted-parties-jamesmarapes-coalition-20190923/. Kama, B. (2017). PNG in 2017: An analysis of Papua New Guinea’s political condition and trends through to 2025. Lowy Institute. interactives.lowyinstitute. org/archive/png-in-2017/png-in-2017-png-political-condition-to-2025.html. Ketan, J. (2007). The use and abuse of electoral development funds and their impact on electoral politics and governance in Papua New Guinea. CDI Policy Papers on Political Governance, 2007/2. archives.cap.anu.edu.au/cdi_anu_edu_au/. png/2007-08/2007_07_PPS_4_Ketan.htm. Krook, M. L. (2006). Reforming representation: The diffusion of candidate gender quotas worldwide. Politics & Gender, 2(3), 303–27. doi.org/10.1017/ S1743923X06060107. Laveil, M. (2020, 16 September). PNG: Where voter turnout is too high. Devpolicy Blog. devpolicy.org/png-where-voter-turnout-is-too-high-20200916/. Laveil, M. and Wood, T. (2019). The 2017 election in Papua New Guinea. Development Policy Centre Discussion Paper, 83. doi.org/10.2139/ssrn. 3422550. Lyons, K. (2018, 20 June). ‘Pushing for civil war’: Fears riots could turn into widespread conflict in PNG. The Guardian. www.theguardian.com/world/ 2018/jun/20/pushing-for-civil-war-fears-riots-could-turn-into-widespreadconflict-in-png. Markiewicz, A. and Wood, T. (2018). Evaluation of Australia’s electoral assistance to Papua New Guinea 2015–2017. DFAT. 52 2. ELECTIONS AND POLITICS May, R. J. (2017). Papua New Guinea under the O’Neill government: Has there been a shift in political style? SSGM Discussion Paper, 2017/6. dpa.bellschool.anu. edu.au/sites/default/files/publications/attachments/2017-08/dp_2017_6_ may.pdf. May, R. J. (2020). Politics in Papua New Guinea 2017–20: From O’Neill to Marape. DPA Discussion Paper, 2020/3. dpa.bellschool.anu.edu.au/sites/default/files/ publications/attachments/2020-11/dpa_dp_may_20203.pdf. May, R. J., Anere, R., Haley, N. and Wheen, K. (2011). Election 2007: The shift to limited preferential voting in Papua New Guinea. National Research Institute. Norris, P. and Grömping, M. (2019). Perceptions of electoral integrity (PEI-7.0). Electoral Integrity Project. dataverse.harvard.edu/dataset.xhtml?persistentId= doi:10.7910/DVN/PDYRWL. Okole, H. (2001). The fluid party system of Papua New Guinea: Continuity and change in a third wave democracy [Unpublished doctoral dissertation]. Northern Illinois University. Okole, H. (2002). Political participation in a fragmented democracy: Ethnic and religious appeal in Papua New Guinea. Development Bulletin, 59, 31–34. Okole, H. (2005). The ‘fluid’ party system of Papua New Guinea. Commonwealth & Comparative Politics, 43(3), 362–81. doi.org/10.1080/14662040500304924. Okole, H. (2012). A critical review of Papua New Guinea’s Organic Law on the Integrity of Political Parties and Candidates: 2001–2010. SSGM Discussion Paper, 2012/5. ssgm.bellschool.anu.edu.au/sites/default/files/publications/ attachments/2015-12/2012_5_0.pdf. Osi, H. (2013). The 2012 elections in Oro Province: The emerging candidate and voter behaviour and distinct differences between the open constituencies, vis-à-vis the provincial constituency. Paper Presented at the SSGM Pacific Week, Canberra. Pacific Women. (2016). Deputy Speaker of the House: ‘Women leaders vital for Bougainville’. pacificwomen.org/news/deputy-speaker-of-the-house-womenleaders-vital-for-bougainville/. Radio New Zealand. (2002, 7 August). Women in PNG’s Chimbu Province say successful candidates won through violence and intimidation. www. rnz.co.nz/international/pacific-news/140356/women-in-png%27s-chimbuprovince-say-successful-candidates-won-through-violence-and-intimidation. Radio New Zealand. (2015, 7 September). PNG Supreme Court overturns grace period extension. www.rnz.co.nz/international/pacific-news/283484/pngsupreme-court-overturns-grace-period-extension. 53 PAPUA NEW GUINEA: GOVERNMENT, ECONOMY AND SOCIETY Radio New Zealand. (2016, 31 May). Greater understanding of women in politics in Bougainville. www.rnz.co.nz/international/pacific-news/305281/ greater-understanding-of-women-in-politics-in-bougainville. Radio New Zealand. (2020, 12 March). PNG moves to get more women into parliament. www.rnz.co.nz/international/pacific-news/411521/png-movesto-get-more-women-into-parliament. Reilly, B. (2002). Political engineering and party politics in Papua New Guinea. Party Politics, 8(6), 701–18. doi.org/10.1177/1354068802008006004. Reilly, B. (2006). Democracy and diversity: Political engineering in the Asia-Pacific. Oxford University Press. Saffu, Y. (1989). Survey evidence on electoral behaviour in Papua New Guinea. In M. Oliver (Ed.), Eleksin: The 1987 national election in Papua New Guinea (pp. 15–36). University of Papua New Guinea. hdl.handle.net/1885/133355. SBS News. (2020, 13 August). Elections on the island of Bougainville this month feature a record number of women candidates. www.sbs.com.au/ news/a-record-number-of-women-are-contesting-elections-in-bougainvillethis-month. Sepoe, O. (2002). To make a difference: Realities of women’s participation in Papua New Guinea politics. Development Bulletin, 59, 39–42. Sepoe, O. (2005). Organic Law on the Integrity of Political Parties and Candidates: A tool for political stability. SSGM Working Paper Series, 2005/4. dpa.bell school.anu.edu.au/sites/default/files/publications/attachments/2015-12/05_ 04wp_Sepoe_0.pdf. Standish, B. (2007). The dynamics of Papua New Guinea’s democracy: An essay. Pacific Economic Bulletin, 22(1), 135–57. devpolicy.org/PEB/2019/06/29/ the-dynamics-of-papua-new-guineas-democracy-an-essay/. Transparency International PNG. (2017). TIPNG observation report: 10th national parliamentary elections. Transparency International Papua New Guinea Inc. Whiting, N. (2020, 23 September). Bougainville elects new president Ishmael Toroama as negotiations on independence from PNG ramp up. ABC News. www.abc.net.au/news/2020-09-23/bougainville-elects-new-president-ishmaeltoroama/12692158. Wood, T. (2014). The three political economies of electoral quality in Papua New Guinea and Solomon Islands. Challenges of electoral integrity in Asia-Pacific workshop, 28 September, Sydney. doi.org/10.2139/ssrn.2497206. 54 2. ELECTIONS AND POLITICS Wood, T. (2016). Is culture the cause? Choices, expectations, and electoral politics in Solomon Islands and Papua New Guinea. Pacific Affairs, 89(1), 31–52. doi.org/10.5509/201689131. Wood, T. (2017). Papua New Guinea election results: Trends and patterns 1972– 2012. Development Policy Centre Discussion Paper, 55. doi.org/10.2139/ ssrn.2926707. Wood, T. (2018). The clientelism trap in Solomon Islands and Papua New Guinea, and its impact on aid policy. Asia & the Pacific Policy Studies, 5(3), 481–94. doi.org/10.1002/app5.239. Wood, T. (2019). The Papua New Guinea election results database. devpolicy.org/ pngelections/. World Bank. (2019). Worldwide governance indicators. www.govindicators.org. 55 This text is taken from Papua New Guinea: Government, Economy and Society, edited by Stephen Howes and Lekshmi N. Pillai, published 2022, The Australian National University, Canberra, Australia. doi.org/10.22459/PNG.2022.02