GENDER AND DEMOCRACY IN THE PACIFIC ISLANDS

In many parts of the world, democracies – both new and established – are ‘widely perceived to be suffering a serious crisis of representation, participation and legitimacy’ (Waylen Reference Waylen2015: 495). The political representation of women is significant both in that women’s under-representation is viewed as symbolic of this ‘crisis of democracy’, and in that increasing the presence of women is seen as a potential way to ameliorate it (Waylen Reference Waylen2015). In the academic literature on the politics of small states, it is often claimed that small states are more likely to be democratic (Anckar Reference Anckar2002; Hadenius Reference Hadenius1992; Srebrnik Reference Srebrnik2004; see also Veenendaal Reference Veenendaal2014). Indeed, the Pacific Islands region is made up entirely of small island developing states,Footnote ⁴ which have been – despite sometimes high levels of political instability, and much debate over whether democracy is relevant to and appropriate for Pacific cultures (see Larmour Reference Larmour1994) – largely continuously democratic, in the sense that elections are frequently held and that the executive is accountable to an elected legislature, since the post-colonial era began in the 1960s (see Corbett Reference Corbett2015b; Reilly Reference Reilly2002).Footnote ⁵ Yet despite this history, women remain severely under-represented, or absent altogether, from Pacific legislatures.

Various arguments are employed to make a case for greater women’s representation in politics, based on the importance of equal opportunity, the political advantages to increased women’s representation or the positive symbolic effects for wider society (see Sawer Reference Sawer2000). There is also the claim that the presence of women in legislatures makes a substantive difference to political debate and decision-making, and that women legislators are in a unique position to represent women substantively. According to Jane Mansbridge (Reference Mansbridge2005: 622), ‘descriptive representation by gender improves substantive outcomes for women in every polity for which we have a measure’. Critics of this argument note that women are not a homogeneous group with identical interests, and that even if ‘women’s interests’ could be articulated, the presence of women in legislatures does not guarantee the adequate representation of those interests (Sawer Reference Sawer2000).

In her work, Anne Phillips (Reference Phillips1995, Reference Phillips1998) has argued that while women do not all have common needs, the lived experiences of women create complex and not clearly defined, but nevertheless gender-specific, interests that only women can represent within legislatures. Iris Marion Young (Reference Young2002) claims that broad groups, including women, are impossible to define through needs and interests; they do, however, share a social status because of structural inequalities, and thus women can make a distinctive contribution to political debate. While one representative cannot fully represent any group as a whole, representation of societal groups such as women allow for the representation of this shared social perspective (Young Reference Young2002). Phillips and Young approach this issue from different perspectives, yet both advocate that women legislators are in a unique position to represent women substantively.

In many parts of the world today, gradually increasing levels of women’s representation have meant that the issue of one representative standing for the larger group is less relevant, as women from a wider variety of backgrounds are coming to be represented, albeit still at lower rates than men in most legislatures. Given the extremely low levels of women’s representation that endure in the Pacific Islands region, however, some Pacific female parliamentarians remain the sole voice for women. For example, in countries including Solomon Islands and Tonga, there has never been more than one woman member of parliament at any time. One-third of Pacific Islands countries have a single woman member of parliament, and a further one-quarter have no female parliamentarians.

The issue of the under-representation of women in politics has been identified as a key area of focus for political leaders and aid donors in the region (Corbett and Liki Reference Corbett and Liki2015; Zetlin Reference Zetlin2014). The absence of women from decision-making positions in the Pacific has been described as an ‘embarrassment’ by the prime minister of Samoa (quoted in Abbott Reference Abbott2012), and former Papua New Guinea national MP Dame Carol Kidu argued that it threatened the legitimacy of political institutions: ‘How you could legitimately call that a representative democracy is beyond my comprehension . . . when there is only less than 1% of half the population in Parliament’ (quoted in Garrett Reference Garrett2013). Gender equality, and in particular women’s political representation, were key issues discussed at the 2011 and 2012 Pacific Islands Forum leaders’ meetings, leading to the 2012 Gender Equality Declaration, which included support for the use of temporary special measures.

From the perspective of donors, women’s political representation is seen as important through two lenses: firstly, from a standpoint that stresses the importance of equal opportunities for women and men to participate in politics; and secondly, because female representatives are expected to act for women (see Corbett and Liki Reference Corbett and Liki2015). The first perspective is preoccupied with a rights-based approach; a UN representative speaking on temporary special measures noted that it was important ‘to ensure that women increasingly take their rightful places in Pacific parliaments’ (UNDP 2015). The second is largely focused on how women’s representation could advance development outcomes. Pacific Women, an Australian government aid programme, identifies as one of its four key outcomes increased women’s participation in decision-making. In its design document, it notes that ‘increasing women’s voice in formal and informal decision-making processes from community through to national levels is important to ensure their inclusion in and contribution to development’ (DFAT 2014: 4–5). Its website argues that there is a link between increased women’s political representation and economic and social development (Pacific Women n.d.). In interviews, representatives from several international aid agencies stressed a connection between women’s representation and development:

The goal of having more women parliamentarians in the Pacific Islands is seen as improving women’s representation both descriptively, in terms of the number of women involved in politics, and substantively, through better development outcomes.Footnote ⁶ This article argues that a focus on the latter tends to reinforce the burden of extremely high expectations that Pacific women in politics face, especially when they are the sole female representative – or one of very few – in parliamentary politics. A strategy for increasing women’s representation with a goal of greater descriptive representation would be likely to help ease this burden, and be more likely to succeed in its aim.

WOMEN AS CANDIDATES

In the aftermath of elections, the question of why female voters do not vote for female candidates is often raised, not only by candidates but also academic observers (Zimmer-Tamakoshi Reference Zimmer-Tamakoshi2013) and political actors. In his inaugural address as prime minister of Tonga, ‘Akilisi Pohiva (Reference Pohiva2015) raised the issue of the absence of women from parliament following the 2014 election and advocated voter education for women:

The lack of support given by female voters to female candidates is often listed as a key obstacle to political power (see Huffer Reference Huffer2006; PPSEAWA et al. Reference PPSEAWA2004).

For female candidates, blame for an election loss is often placed on female voters in the constituency. One candidate in the 2015 Bougainville election claimed before counting: ‘if I win, it’s a victory for the men. If I lose, it’s thanks to the women’ (personal communication, May 2015). A Tongan candidate in the 2014 election claimed a lack of support from female voters as the reason she was unsuccessful: ‘I didn’t feel supported by women. No woman wants the success of another woman . . . that’s why I failed. I blame the women’ (personal communication, March 2015). At the same time, if women are successful it is often exclusively male voters that are the subject of praise for voting for them. One woman described going on a radio programme following the 2012 Papua New Guinea election to thank male listeners for voting for female candidates, claiming you have to ‘pat them on the back’ so they will continue to support women in politics (personal communication, March 2015).

Women who had contested elections in Papua New Guinea noted that one barrier was that female voters were not interested in supporting or voting for women (personal communication, March 2015). Women running against other women in elections was also seen as a problem, as it potentially split the loyalties of female voters (personal communications, March 2015, March 2016). A female candidate in Tonga alleged that a high-profile advocate of greater women’s representation was publicly encouraging voters to vote for women, yet was supporting a male relative in her own constituency. One Tongan public servant expressed her support for the use of temporary special measures in parliament to increase the number of women over voter education initiatives, arguing that ‘you can’t educate women to vote for women’ (personal communication, March 2015).

This discourse contributes to a sense of responsibility for the success or failure of female candidates that is conveyed almost exclusively onto female voters. In this discourse, female voters are expected to give precedence to gender above all other considerations, even though their expectations and aspirations for their constituency representatives are often the same as or very similar to those of men – namely, service delivery as a key priority, especially in Melanesia (McLeod Reference McLeod2002). Male voters, on the other hand, are allowed a more complex calculation, taking into account kinship ties, political beliefs and customs of reciprocity. Elise Huffer (Reference Huffer2006: 45) in her account of women’s under-representation in the Pacific drew attention to this double standard: ‘Why is solidarity expected from other women when it is not expected of men with respect to other men?’

An approach that focuses on winning the ‘women’ or the ‘women and youth’ vote is commonly utilized, with the fact that women generally make up close to or over half of registered voters in the region used as justification for this strategy (personal communication, May 2015). Evidence from recent elections in the region, however, shows that those who espouse this strategy tend to perform poorly (see Baker Reference Baker2015). It inadvertently creates the impression that women are viable candidates for female voters only, as well as underestimating the impact of the dynamics within families and communities that influence voting. The approach can create a perception of female candidates as ‘women’s candidates’, a potentially significant obstacle. In the 2015 Bougainville election, one candidate reported that male voters would not attend her campaign events as it was thought that only female voters should go to listen to female candidates (personal communication, April 2015). Other female candidates noted that prominent men in the community might publicly encourage women to vote for female candidates, yet would often stop short of promising their own votes. Community gatekeepers – leaders and other politically influential people who have key roles in shaping political thinking – tend to be older men, and their lead is often followed at the ballot box as a result of explicit or implicit pressure. This gendered bias is reinforced by campaign activities which are often traditionally centred on masculine fora, such as kava circle campaigning in Tonga (Campbell Reference Campbell2011; see also Corbett Reference Corbett2015a; Huffer Reference Huffer2006).

One election-related issue that is often discussed is the ability of voters, and in particular women, to cast their votes freely (see Haley and Zubrinich Reference Haley and Zubrinich2013; Haley et al. Reference Haley, Zubrinich, Barbara, Baker and McMahon2015; Huffer Reference Huffer2006; Sepoe Reference Sepoe2013). In the debate around the shift from a first-past-the-post to a limited preferential voting system in Papua New Guinea, proponents argued that limited preferential voting would be advantageous to female candidates; it was thought that even if women were pressured by their family or community to cast their first preference vote a certain way, they might have more freedom to allocate their second and third preferences (May et al. Reference May, Wheen and Haley2013). Of course, this rests on the assumption that women, in the absence of intimidation, would choose to vote for female candidates.

In the 2007 general election – the first held using limited preferential voting – it was noted that women’s choices were often influenced by male relatives and that bloc voting was commonplace, enforced by the threat of violence or estrangement (Sepoe Reference Sepoe2013). Money politics – the distribution of gifts and cash in exchange for votes – was also more frequently practised by male candidates (Sepoe Reference Sepoe2013). In the 2012 general election, it was noted that while turnout by female voters was high, assistance in casting votes was often forced on women and many were observed to not have filled out their own ballot paper (Haley and Zubrinich Reference Haley and Zubrinich2013). Women were more likely than men to report either having experienced intimidation or having been prevented from voting (Haley and Zubrinich Reference Haley and Zubrinich2013). The freedom of women to vote without intimidation is not of concern only in Papua New Guinea, with intimidation of female voters prevalent in other countries as well (Huffer Reference Huffer2006). In neighbouring Solomon Islands, almost one in three women surveyed after the 2014 election reported having experienced voter intimidation, as opposed to one in four men (Haley et al. Reference Haley, Zubrinich, Barbara, Baker and McMahon2015). Other research on Solomon Islands has also highlighted the impact of pressure from male heads of households on the voting behaviour of women (see Corbett and Wood Reference Corbett and Wood2013).

This intimidation takes many forms. In some areas of Papua New Guinea, female voters have been assaulted at polling stations when attempting to mark their own ballots (Haley and Zubrinich Reference Haley and Zubrinich2013). The threat of violence is a frequently used tactic to exert pressure on women to vote for a certain candidate (personal communication, March 2015). Elsewhere, the intimidation experienced by women takes place for the most part in the home or community. In these cases too, however, there is often an explicit threat of retaliation that prevents women from exercising their right to vote freely (Billy Reference Billy2002).

Voter intimidation, while widespread, is not experienced by all female voters in the Pacific Islands. Yet even where women can cast their votes freely, it would be wrong to assume that they would always choose a female candidate. The idea that being a woman trumps competing identities in choosing how to cast a vote is not borne out by past experience. In a 2014 survey in Fiji, only 17 per cent of female respondents reported that they would prefer to vote for a woman; 73 per cent reported they had no gender preference; and 11 per cent reported they would prefer to vote for a man (Dumaru and Pene Reference Dumaru and Pene2014). Some Pacific women would not agree with the statement that increased women’s political representation is a good thing (Sepoe Reference Sepoe1996). Yet even for those who agree in principle that there should be more women in politics, it does not follow that they would necessarily give their own vote to a female candidate.

Political parties can be influential. Several Pacific Islands have strong political party systems, including countries with relatively high women’s representation, such as French Polynesia, New Caledonia and Fiji. Parties are often seen as having an important gatekeeping role in recruitment processes that impact the number of female candidates (Franceschet 2005; Matland 2005; Norris Reference Norris1997; Norris and Lovenduski Reference Norris and Lovenduski1995). Party affiliation can also significantly influence voter choice. In the 2014 Tonga election, one unsuccessful female candidate noted the influence of the prevailing political ideology on the voting patterns of her local community, most of whom ultimately chose to vote for her male opponents (personal communication, March 2015). The influence of political parties, however, varies significantly throughout the region. In many parts of the Pacific, party systems are weak and fragmented and parliaments have large numbers of independent MPs, while some Pacific parliaments have no parties represented at all (Fraenkel and Grofman Reference Fraenkel and Grofman2005). In these cases, party affiliation has little or no impact on the election process and is often only important in the ‘election after the election’ – the selection of prime minister and cabinet.

In much of the Pacific Islands, politics is intensely local. One of the higher-performing female candidates running in a non-reserved seat in the 2010 Bougainville election, while acknowledging the gendered nature of some of the challenges she faced in campaigning, attributed her loss mainly to kinship networks within the constituency. She stated that while she had the support of her own community, they were a minority in the constituency, and members of a larger group voted for another candidate because of community ties: ‘He’s their wantok,Footnote ⁸ so that comes first’ (personal communication, September 2013). These kinship ties are often of significant importance, and female voters can prioritize kinship networks over gender at election time (McLeod Reference McLeod2002).

This paramountcy of kinship ties is an oft-criticized element of Pacific politics. Yet it is, as Terence Wood argues in the case of Solomon Islands, frequently a rational response to the political context:

The clientelist strategy of ‘people tending to vote for candidates who are related to them on the assumption that the winning candidates will reward wantok before passing on largesse to more distant associates’ is well documented (Zimmer-Tamakoshi Reference Zimmer-Tamakoshi2013: 413). In this political environment, concerns about women’s representation at the national level are superseded by concerns about service delivery at the local level – and the ability of the local MP to provide this. Attempts by female candidates to capitalize on gender identity to win elections through pushing notions of women’s rights and gender equality in election campaigns have typically had little resonance, especially in rural areas of Melanesia (Corbett and Liki Reference Corbett and Liki2015: 332; Sepoe Reference Sepoe2013), and led to accusations that the candidate does not understand or respect local culture (personal communication, November 2014).

In her study of the campaign of a female candidate in the 2007 Papua New Guinea national election, the anthropologist Laura Zimmer-Tamakoshi (Reference Zimmer-Tamakoshi2013: 408) drew on the theory of Ira Bashkow (Reference Bashkow2006) of ‘whiteness’ as a moral concept that implies a disconnect from tradition in her discussion of how the candidate was often accused of being ‘white’. Similar accusations of ‘whiteness’ and related cultural ignorance were made of female candidates in the 2014 Solomon Islands election (personal communication, November 2014) and the 2015 Bougainville election (personal communication, May 2015). As Zimmer-Tamakoshi (Reference Zimmer-Tamakoshi2013) points out, male candidates are often subject to the same scrutiny and have their connections to the local community debated and tested during election campaigns. Yet there is a gendered aspect to this criticism of female candidates that is linked to the traditional role of women within society and the masculinized ideal of the political candidate. For some voters, the simple fact of standing for election as a woman implies an unacceptable distance from traditional custom.

Candidate training for women organized by international organizations has potentially contributed to this perception of female candidates as ‘white’ and culturally removed. Training for female candidates generally involves workshops led by an international facilitator, possibly in coordination with some local organizations, to teach campaigning and leadership skills. Past training initiatives have been criticized for focusing on urban women (personal communication, March 2015). It has also been argued that this training emphasizes women’s rights and gender equality rather than issues that have more resonance for voters, such as service delivery (Corbett Reference Corbett2015a: 72). There is limited evidence that candidate training as it has been carried out in the past has produced stronger female candidates; in the 2010 Solomon Islands election, women who had attended training on average received just 3 per cent of the vote share (Wood Reference Wood2015).

Training also often focuses on the difference of women in comparison to male candidates. One difference often emphasized in their subsequent campaigns is that women are ‘clean’ – that is to say, not corrupt – candidates. In the 2015 Bougainville election, this apparent point of difference between female and male candidates was often emphasized by female candidates: ‘men don’t do clean campaigning’ (personal communication, May 2015). Clean campaigning was defined by female candidates as not getting involved in money politics or gifting, and not attacking other candidates. Vote-buying and gifting are practices that can easily be construed as corrupt; yet they are also, in many parts of Melanesia, an accepted – and expected – component of contemporary politics (see Allen and Hasnain Reference Allen and Hasnain2010; Haley and Dierikx Reference Haley and Dierikx2013; Standish Reference Standish2007). In the 2012 Papua New Guinea election, money politics was a noted feature of the campaign period in all regions, with candidates who participated in these activities in general performing better than other candidates (Haley and Zubrinich Reference Haley and Zubrinich2013).

The perception of women as ‘clean’ candidates who do not engage in money politics can be advantageous. In Bougainville, Josephine Getsi in Peit constituency campaigned – and won – on a strong anti-corruption platform that helped to distinguish her from her male opponents, especially those who were former politicians (personal communication, May 2015). For other female candidates in that election, however, anti-corruption messaging resonated less in their communities. In a political context in which service delivery is considered a primary function of a politician’s role, the idea that women are ‘clean’ candidates – and may not deliver for constituents if elected in the way members are expected to – can work against them.

For many candidates in Papua New Guinea, election losses were blamed on money politics (personal communications, March 2015, March 2016): ‘The men on the other hand use money, they use money, they gave money out freely. That was quite obvious . . . that’s where women could not come face-to-face and be able to meet the challenge on par with the men. We didn’t have the money’ (personal communication, September 2013); in these cases voters – and especially female voters – are often seen to share the blame with male candidates, because they were willing to receive money or gifts in exchange for their votes (personal communication, March 2016).

Another stated point of difference for female candidates is that women are good managers, because of their roles as wives and mothers. This idea was used by female candidates in both the 2014 Solomon Islands election (personal communication, November 2014), and the 2015 Bougainville election (personal communication, May 2015). A common sentiment was: ‘Women are good managers in the house . . . [they] manage their families so that prepares them for parliament’ (personal communication, May 2015). This discourse was so widespread that one of the female candidates who described her skills as a manager in terms of her qualifications as a wife and mother was in fact unmarried and had no children. These types of essentialist notions of women’s qualifications for parliament contribute to the idea that women are a homogeneous group with common interests, and that gender should be the overarching concern for female voters.

WOMEN AS REPRESENTATIVES

For those few women who are elected to Pacific parliaments, preconceived notions of what a female politician should be can constitute a significant hurdle. In Melanesia in particular, turnover is high in elections (Fraenkel and Grofman Reference Fraenkel and Grofman2005). In Papua New Guinea, Solomon Islands and Vanuatu, historically around 50 per cent of incumbent politicians lost their seats every election (Morgan Reference Morgan2005; Van Trease Reference Van Trease2005; Wood Reference Wood2014). Even in countries with lower turnover, female politicians can face difficulties in winning re-election, especially those who have won with the ‘sympathy vote’, being elected after a male relative or spouse who was in politics has died (Corbett Reference Corbett2015a: 71). In these contexts, where re-election is already a difficult prospect, the high expectations on female representatives can be a significant impediment.

Female politicians are often expected to be less corrupt than their male counterparts. This is linked to the commonly cited argument for increased women’s representation that it leads to good governance and lower levels of corruption (see Dollar et al. Reference Dollar, Fisman and Gatti1999; Swamy et al. Reference Swamy, Knack, Lee and Azfar2001). Such arguments have been critiqued by many scholars, who argue that many of these claims are ‘based upon essentialist notions of women’s higher moral nature and their propensity to bring their finer moral sensibilities to bear on public life, and particularly on the conduct of politics’ (Goetz Reference Goetz2007: 90; see also Sung Reference Sung2003).

While the perceived link between good governance and increased women’s representation has been challenged, the perception remains. Women are seen as better than men at resisting bribes (Dumaru and Pene Reference Dumaru and Pene2014). Former Papua New Guinea MP Josephine Abaijah (Reference Abaijah1991: 306) wrote that her constituents ‘put me on a pedestal and said that, being a woman, I was beyond the corruptions and foibles that afflicted many men’. While this perception can be an advantage for women in earning the trust of voters, for those who fall off the pedestal the consequences can be worse than for men involved in scandal. Male politicians are often assumed to be corrupt (Corbett Reference Corbett2015a), while the expectation on women is that they are a ‘better’ class of politician.

The behaviour of female MPs in Papua New Guinea is heavily scrutinized online and in the media. Shortly after she was elected, the marital issues of Lae Open MP Loujaya Kouza were widely reported on; after she requested an Interim Protection Order against her husband (Buckley Reference Buckley2012), a call was made for Kouza to resign from her ministerial position, with a senior public servant claiming: ‘we expect the minister responsible [for child and family welfare] to be the leader in promoting good family lives’ (quoted in Papua New Guinea Post-Courier 2013a). A letter to the editor published in a major national newspaper stated that Kouza’s marital troubles proved ‘that women are irrational beings who use their emotions rather than logic to assess situations and make decisions’ (Papua New Guinea Post-Courier 2013b). She has been publicly critiqued for breaching cultural norms of how women should act. After claiming that the governor of Morobe had called her a ‘wicked witch’, Kouza was then criticized for making the allegation by a local woman leader who said it was disrespectful and further stated: ‘It is against our cultural principle for Ahi women to criticise Ahi men’ (quoted in Kivia Reference Kivia2014).

Criticism of the behaviour of female MPs is often articulated in terms of their status as role models for women. After Sohe Open MP Delilah Gore was ejected from a flight for refusing to turn off her phone, a community leader in her electorate said: ‘The Minister’s actions were unbecoming of a leader and a female politician whom the women of this country can look up to’ (quoted in Salmang Reference Salmang2014). A witness to the incident stated: ‘She is one of the only three female parliamentarians and is looked up to by many young ladies’ (quoted in Willie Reference Willie2014). Even some women leaders expressed concern over the public images of the female MPs, arguing that they impacted on the reputations of all female candidates, and that voters wanted to vote for women with husbands (personal communication, March 2015).

Once elected, female representatives can find themselves in precarious positions, often with slim majorities and seen as ‘easy targets’ for electoral petitions or challenges to re-election. A long-serving female MP in Samoa noted the difficulties of retaining a parliamentary seat as a woman: ‘Women are seen as easy targets, you know, especially for court challenges and so forth, so the getting in is hard enough, staying in is even harder, whether it’s through court challenges or coming back for re-election’ (personal communication, August 2012).

Of the 13 women who served as MPs in Samoa between 1970 and 2011, only four served multiple consecutive terms, and over half served only one term or less.Footnote ⁹ A further three women have been elected to parliament, but lost their seats due to electoral petitions. In Tonga, no woman representative – either elected or appointed – has served more than one parliamentary term.

In countries with a history of extremely low levels of women’s political representation – as in most of the Pacific Islands – female MPs are seen as carrying the responsibility to represent women as a group, as well as their own individual constituencies. This can constitute a ‘dual sense of ownership’ (Corbett Reference Corbett2015a: 72), and furthermore a dual set of expectations on their performance as MPs. Some Pacific female MPs reported often being sought out by women from other constituencies who preferred to deal with another woman, rather than their own member (Corbett and Liki Reference Corbett and Liki2015: 337). Hilda Kari (quoted in Pollard and Waring Reference Pollard and Waring2009: 73) said that, as the first female MP in Solomon Islands, ‘I was everything to everybody’. In addition to these perceived extra responsibilities, female parliamentarians find themselves treated as outsiders in a masculinized environment, which can lead to a sense of isolation in parliament (Corbett Reference Corbett2015a; Corbett and Liki Reference Corbett and Liki2015).

For women in reserved seats, such as the three regional seats that are set aside for women in the Bougainville House of Representatives,Footnote ¹⁰ the situation is different albeit with its own difficulties. In Bougainville, the regional women’s members are tasked with representing women, meaning the dual constituency problem outlined above is not an issue. Yet despite the regional seats covering much larger areas than the open constituencies – the largest regional seat, North, covers the same area as 14 open constituencies – regional members receive the same amount in constituency development funds as open members.Footnote ¹¹ Bougainvillean politicians often claim the amount of constituency development funds that they receive is inadequate to deliver projects on the scale demanded by their constituents (Baker and Oppermann Reference Baker and Oppermann2015), and this issue is exacerbated for the regional members given their significantly larger electorates. Regional members in the 2010–15 term and their staff reported their difficulties in demonstrating service delivery to their constituents: ‘My region is a very big area. I haven’t reached the other end yet . . . [at the next election] people will probably vote me out, and another woman can come in’ (personal communication, September 2013); and ‘For our current member, [voters] expect much more involvement from her. They complain that she does not do much for the people of Konnou because she’s from our area. So they are complaining that she did not do much . . . they see her as a[n open] constituency member’ (personal communication, May 2015). Turnover for the regional seats was 100 per cent in both the 2010 and 2015 Bougainville elections, compared with 69.7 per cent and 57.6 per cent respectively in the open seats.

In addition to the disadvantages that stem from serving larger electorates, the women’s regional members were also frequently criticized for not collaborating on projects. A common refrain was: ‘These three women leaders in parliament are not really working together’ (personal communication, September 2013). As representatives for women, they were expected to work collaboratively, an expectation that was not directed towards either the open constituency members (who in the 2010–15 term were all men) or the regional ex-combatants’ members.

When a woman is elected in the Pacific Islands, she is often the only woman, or one of just a handful of women, in parliament. This can mean that expectations of their efficacy are high, given that being elected as a woman in the Pacific Islands is a significant achievement. Conversely, it can mean that they are isolated in a parliament where women form a token presence, making it hard to effect legislative change. A former MP in the Tongan Parliament said: ‘it’s a different story walking in, especially as the only woman. I was nervous and had no confidence in the beginning’. She further noted that being the sole woman hampered her efforts to pursue substantive reform on gender issues: ‘it’s difficult for male MPs to support women’s issues. We need a number of women, not just one or two, for when the voting happens’ (personal communication, March 2015).

Of course, the idea that increased women’s representation will make it easier to effect substantive change on gender issues also assumes that women in politics will cooperate with each other. This expectation that female representatives should be working together is indicative of the assumption that they share a common agenda. Women in politics are often expected to act as the default spokespeople for international norms of women’s rights and gender equality, as set out in a number of regional and international declarations and development strategies.Footnote ¹² Some female MPs resent this role, and the implication that gender advocacy should be their primary function as politicians (Corbett and Liki Reference Corbett and Liki2015). While small states are more likely to be democratic, they are also likely to have a more conservative political culture (see Baldacchino Reference Baldacchino2012; Corbett Reference Corbett2015c; Lawson Reference Lawson1996), meaning that social change can be a slow and difficult progress requiring significant political capital. Even for women who are committed to these gender equality goals, it is difficult to balance this work with the work of representing their own constituencies. While the former may even be their primary motivation to go into politics, the latter is essential to win re-election. Thus, the pursuit of national-level policy aims can be difficult to reconcile with the constituency-based work necessary to remain in politics, an issue by no means restricted to female politicians in the Pacific Islands (see Morgan Reference Morgan2005). Carol Kidu was a member of the Papua New Guinea Parliament for three consecutive terms from 1997 to 2012 and was minister for community development from 2002 to 2011. In her time in politics, Kidu spearheaded the passage of the Lukautim Pikanini (Child Protection) Act 2009 as minister for community development, as well as the Equality and Participation Act, which amended the constitution to allow for the possibility of reserved seats for women in parliament. She credited her re-election in 2002 to the strategic use of constituency development funds in her first term to, as she put it, ‘leave my “handmark” on the electorate’ (Kidu and Setae Reference Kidu and Setae2002: 51).

As Kidu demonstrates, a ministerial portfolio is one way to pursue substantive policy goals effectively. Women are under-represented at the executive level in most countries, and where women are represented in cabinet, they are often found in portfolios that are considered ‘soft’ or less prestigious, such as education, health or social welfare (Davis Reference Davis1997; Reynolds Reference Reynolds1999). While there are signs of gradual change in some areas, where women are present in greater numbers and in more prestigious positions (see Escobar-Lemmon and Taylor-Robinson Reference Escobar-Lemmon and Taylor-Robinson2005; Krook and O’Brien Reference Krook and O’Brien2012), the number of female ministers in the Pacific Islands region remains very low. While more than one in three women MPs in the region also hold a ministerial position, only 7.2 per cent of cabinet ministers in the region are women. Four Pacific countries have all-male cabinets, and no country has more than two female cabinet ministers (see Table 2).Footnote ¹³ The first female head of government in the Pacific region is Hilda Heine,Footnote ¹⁴ who assumed office as president of the Marshall Islands in January 2016. Also in 2016, the first female deputy prime minister of a Polynesian country, Fiame Naomi Mata’afa of Samoa, was appointed.Footnote ¹⁵

Table 2 Women’s Representation in Cabinets in the Pacific Islands Region (as at 16 December 2016)

Notes: ^aThe cabinet of the Federated States of Micronesia is appointed and includes non-elected members.

^bThe cabinet of Palau is appointed and includes non-elected members.

While there is a female presence in most Pacific cabinets, women can be sidelined within political executives when they are given unequal responsibilities or less prestigious portfolios (Murray Reference Murray2013). Many – but not all – past and present women cabinet ministers in the Pacific Islands have held so-called ‘soft’ portfolios, most commonly community development (which often includes responsibility for women’s affairs), health and education. Even where women hold different portfolios, many are conscious of their status as the sole – or one of two – women in cabinet, and feel a responsibility to represent women at the executive level regardless of whether they hold the women’s affairs portfolio. One articulated this sense of responsibility: ‘I am the minister representing women, women’s interests. That’s why even though the minister [for women’s affairs] is a male minister now, I continue to help him with women’s issues’ (personal communication, September 2013).