Mapping GEF: Top-Down (TD) vs. Bottom-Up (BU) Approaches

Whereas political financing takes many forms, it is generally defined as “money for electioneering” (Pinto-Duschinsky Reference Pinto-Duschinsky2002, 70). Diamond and Morlino (Reference Diamond and Morlino2004, 29) argue that democratic quality requires the prevention of “subtle denigrations of electoral rights” and an awareness about how “equality in access to political finance” is often denied. Beetham further elaborates this position by posing the question: what is the value of the political right to stand for elected office if it is only accessible to the wealthy? He suggests that even if “civil and political equality does not require complete economic levelling” (Reference Beetham2000, 97), some sort of action is needed to address the political scope of wealth. For Beetham, the electoral playing field can be tilted by either adding “legal restrictions” on electoral financing to limit the wealthy or by better equipping those with meagre means through “positive attention.” Whereas Beetham pays less attention to what it takes for positive attention to bear fruit, he does indicate a pathway that is more complex, general, and situated in society compared with the state-based approach that is more regulatory.

While these discussions of the relationship between economic inequalities and democratic quality in the 2000s ignored gender issues, Mansbridge had already acknowledged a gendered effect in Reference Mansbridge1999. She argued that the uneven access to funding for marginalized groups, like women, over the long term had left these groups with fewer resources for organizations to make their voices heard. GEF initiatives are based on similar assumptions about electoral financing; there is a need for funding schemes that either motivate more women to stand for election or that convince parties there are no, or at least minor, financial drawbacks connected to backing women candidates, and potentially financial rewards. Given that to win elections parties must use ever-increasing expensive campaign techniques (Childs Reference Childs2013), party leaders are obliged to carefully consider the costs of selecting certain candidates over others, making GEF a potentially powerful tool to incentivize the selection of women candidates.

The logic of Beetham’s two categories helps to map GEF initiatives. “Legal restrictions” are government-led top-down approaches that penalize or pay parties through public funding to make parties comply with a given gender-balance goal. They are initiated and implemented by the government, often involving women’s policy agencies, and in many cases operate in conjunction with legislated quotas, like in France (Achin et al. Reference Achin, Lévêque, Durovic, Lépinard, Mazur, Muriaas, Wang and Murray2020; Mazur et al. Reference Mazur, Lépinard, Durovic, Achin and Lévêque2020). The “positive attention” strategy is bottom-up, as Beetham implies. It includes GEF instruments introduced by nongovernmental actors that cover electoral costs for women that make women candidates more financially appealing for parties to back. Examples include, EMILY’s List in the US (Kreitzer and Osborn Reference Kreitzer and Osborn2019) and the WIN WIN! initiative in Japan (Gaunder Reference Gaunder2011). While the arena is clearly different for each approach, state-based versus society-based, whether a penalty or payment is used does not necessarily match a particular approach. Top-down instruments can either provide a penalty or a payment for parties to comply, but bottom-up mechanism always involve “carrots” rather than “sticks”—giving a boost to women who run for office.Footnote ³ A top-down approach is not limited to legislation alone; funding initiatives that are facilitated and organized by the state bureaucracy may be included. The strategy adopted largely depends on the political financing regime at work in a particular context (Van Biezen and Kopecký Reference Van Biezen and Kopecky2007).

Variation across Top-Down and Bottom-Up Approaches: The Limits of Case Study Research

Country case studies reveal that success—a significant increase in gender balance in parliament—has been mixed regardless of approach and suggest some important lessons about the dynamics and determinants of the two different modes. Piscopo (Reference Piscopo2016), for example, shows in Mexico that top-down strategies, punishment as well as payment, are more likely to succeed. Ohman (Reference Ohman2018) concludes, in a study of top-down GEF, that simply introducing a connection between gender and public funding is unlikely to produce a change by itself and that these regulations are in danger of becoming purely symbolic. Alternatively, Piscopo (Reference Piscopo2015) argues that funding mechanisms are neither irrelevant nor symbolic, as they constitute highly significant interventions into what happens behind closed doors within political parties. Studying the top-down policies that have been implemented in Italy, Feo and Piccio (Reference Feo and Piccio2020a; Reference Feo and Piccio2020b) find that success is also limited; positive effects are only maximized when different policy instruments complement each other and share the same policy goals. In Cabo Verde, top-down measures were adopted in legislation but never implemented (Borges, Muriaas, and Wang Reference Achin, Lévêque, Durovic, Lépinard, Mazur, Muriaas, Wang and Murray2020). In Chile, the GEF scheme was insufficient to level the electoral playing field because of the disadvantage women have as electoral newcomers (Piscopo et al. Reference Piscopo, Hinojosa, Thomas and Siavelis2021).

With respect to bottom-up approaches, Kayuni and Muriaas (Reference Kayuni and Muriaas2014) find in Malawi that campaigns directly targeting women can contribute to an increase in gender-balance if the context is conducive for change. Gaunder (Reference Gaunder2011) studying Japan and Bauer and Darkwah analyzing Ghana (Reference Bauer, Darkwah, Muriaas, Wang and Murray2020) paint a bleaker picture. In Ghana, a reduced filing fee for women aspirants is not enough for women to make it to parliament. While women’s recruitment groups in the US, a potentially bottom-up avenue, have been studied by Kreitzer and Osborn (Reference Kreitzer and Osborn2019), these groups focus on the recruitment and training of women candidates but not funding and, thus, are not considered as GEF in action.

Qualitative Comparative Analysis as the Best Path to Construct a Theory of GEF

The country case research provides interesting insights; however, the absence of uniform measurements undermines the validity of individual case studies and the reliability of cross-case comparison. Nonetheless, this body of work shows that progress can be made through both types of GEF through a shifting interplay between certain conditions. Qualitative comparative analysis, therefore, based on premises of configurational causality—outcomes are a product of a combination of conditions—and of equifinality—multiple combinations of conditions or paths to the same outcome—is a useful analytical tool to understand the interplay of GEF with other conditions.Footnote ⁴ Scholars in gender and politics often opt for QCA over statistical analysis in larger-n studies due to the highly complex nature of gendered dynamics at play (e.g., Ciccia Reference Ciccia2016; McBride and Mazur Reference McBride and Mazur2010). Krook (Reference Krook2010) and Lilliefeldt (Reference Lilliefeldt2012) specifically assert in their studies of gender and representation that QCA is a more appropriate methodological approach than correlational analysis. For Krook (Reference Krook2010), reflecting a broader consensus in the QCA community, the use of interaction terms in regression is not an acceptable substitute for the configurational logic of QCA, a rationale often used by critics of QCA who argue correlational analysis does the same job as configurational methods.

Potential Paths to GEF Success

This section turns to the configurational hypotheses about GEF implementation success to be assessed in the mixed methods analysis of the 31 elections. While the literature identifies many different conditions that can potentially combine with GEF implementation to significantly enhance gender balance in legislatures, the QCA conducted here focuses on the potential combination of six of these conditions.Footnote ⁵ The reasons for omitting other important conditions are discussed at the end of this section. As previous research indicates and is shown in greater detail below, it is the presence or absence of each condition that counts in successful outcomes rather than gradated levels of each condition. Therefore, crisp set QCA is more appropriate to be used with distinct cutoffs established for “presence” and “nonpresence,” rather than fuzzy set or multivalue QCA.Footnote ⁶ The six conditions, expressed in crisp set logic, where capital letters indicate presence and small letters absence, include whether GEF is state-driven or not (SD/sd), whether there is a quota in action at the time of GEF implementation (Q/q), whether there is a PR electoral system in place or not (PR/pr), whether there is centralized candidate selection or not (CCS/ccs), whether political parties are publicly funded or not (PF/pf), and whether there is a 15% minimum level of women members of parliament or not (WMP/wmp). The configurational hypotheses for success are presented first for top-down state-driven instruments (Box 1) and then for bottom-up mechanisms developed in society or parties (Box 2). The specific operationalization of each condition, as well as the omitted conditions discussed below, can be found in the codebook in online Appendix 1, which also includes the source literature used for coding.

Box 1. Hypotheses for Top-Down Success Paths

Note: KEY = State Driven Regulatory Approach of GEF (SD); Quota in Action (Q); Proportional Representation Electoral System (PR); Public Funding of Parties (PF); Minimum Level of Women MPs (WMP).

Box 2. Hypotheses for Bottom-Up Success Paths

Note: KEY = State Driven Regulatory Approach of GEF (SD); Quota in Action (Q); Proportional Representation Electoral System (PR); Centralized Candidate Selection (CCS); Public Funding of Parties (PF); Minimum Level of Women MPs (WMP).

While the presence of a state-driven GEF may determine the overall approach, it is not always part of a successful path. As Lépinard (Reference Lépinard2016) for France and Espírito-Santo (Reference Espírito-Santo, Lépinard and Rubio-Marín2019) for Portugal find, parties that spearhead the adoption of GEF measures may not comply, either due to lack of commitment to authoritative policies or because feminist activists who see the policies as half measures do not hold political parties’ feet to the fire. Given that much research asserts that quotas are important in promoting gender balance in representation and that an initial comparative study of GEF finds that having a quota in place is a sufficient but not necessary condition for increased gender balance (Muriaas, Wang, and Murray Reference Muriaas, Wang, Murray, Muriaas, Wang and Murray2020), it is crucial to include quotas as a condition in combination with state-driven GEF.Footnote ⁷ In France and Ireland, for example, financing through state-driven instruments is used as a remedy to penalize parties that do not comply with a gender quota by reducing the amount of funding for parties that do not pass the threshold (Achin et al. Reference Achin, Lévêque, Durovic, Lépinard, Mazur, Muriaas, Wang and Murray2020; Buckley and Gregory Reference Buckley, Gregory, Muriaas, Wang and Murray2020). Although they may be administratively connected, as in France and Ireland, GEF measures operate independently from any gender quota and thus are conceptualized here as two distinct instruments. As shown in Box 1, all four hypotheses for top-down paths to success include a quota in action combined with state-driven instruments, even though quotas are not always present in all top-down GEF. Research shows that these two conditions need to combine with at least one other condition to facilitate successful outcomes. Different options are plausible.

The first option includes the presence of proportional representation electoral systems (H1). Many comparative studies have shown the importance of PR for increasing women’s representation, but the studies also find that women’s increased representation does not occur in all PR systems (Matland and Studlar Reference Matland and Studlar1996; Norris and Lovenduski Reference Norris and Lovenduski1989; Sawer, Tremblay, and Trimble Reference Sawer, Tremblay and Trimble2006; Tremblay Reference Tremblay and Tremblay2012). The presence of multimember districts in PR systems has the potential to reduce resistance to GEF because influential men party members may still be placed in winnable seats even when the party complies with the threshold of women candidates needed to receive a payment or avoid a penalty. While there are a range of different types of PR systems, what is at play is that compliance to place women candidates on lists is easier to manage in multimember districts than in single-member districts in first-past-the-post electoral systems; thus, the crisp set logic of the presence or absence of PR makes empirical sense.

A second option brings in a third condition developed specifically for this study, a 15% minimum of women MPs (H2). Krook’s (Reference Krook2010) study of the different pathways to enhanced women’s representation specifically identifies “women’s status in politics” as a crucial ingredient for success. Others have focused on the importance of feminist coalitions, called strategic partnerships and triangles of empowerment, between women’s policy agencies, women’s movements, and women MPs as being a key part of feminist agendas as well as being indicative of a broader acceptance of values about women’s rights and gender equality.Footnote ⁸ Early work on critical mass by Kanter (Reference Kanter1977) on “skewed groups” and Dahlerup (Reference Dahlerup1988) on “tilted groups” provide the conceptual foundation for this condition.Footnote ⁹ Both assert that with a minimum of 15% women in parliament women can begin to form alliances within and outside of parliament and change institutional culture. In addition, the presence of at least 15% women MPs in parliament indicates elite support for gender equality—Krook’s “women’s status in politics”—as well as providing a potential for strategic alliances and a guaranteed reserve of experienced women when there is a sudden demand for women candidates when GEF is put into action. Such a reserve of eligible women also helps the eligibility of political parties for financial rewards or to avoid a penalty for noncompliance.

The third and fourth options for which conditions combine with state-driven GEF and quotas include the presence and absence of public funding for political parties (H3 and H4), often the vehicle through which GEF is administered. While public funding has been identified as a key mechanism for achieving equality of competition (Van Biezen and Kopecký Reference Van Biezen and Kopecky2007), work on the gender effects of public party funding is quite limited (Ohman Reference Ohman2018). Nonetheless, the country-based studies do provide some insights. Research on top-down initiatives suggests that when public funding is significant for parties, the door is opened for fines to be given for noncompliance with a gender-balance target (Buckley and Gregory Reference Buckley, Gregory, Muriaas, Wang and Murray2020; Achin et al. Reference Mazur, Lépinard, Durovic, Achin and Lévêque2020). Top-down strategies may also occur in countries with limited public support for campaigning or party activities. Thus, the absence of significant public funding is also likely to make the prospects of a reward more appealing for parties just as much as its presence (Feo and Piccio Reference Feo and Piccio2020a; Ohman Reference Ohman2018).

Bottom-up strategies, found in Malawi, the US, and other countries, take place when a state-driven GEF is absent, instead relying on initiatives from private actors like political parties, companies, NGOs, and individual actors. It is expected that bottom-up schemes may be less successful than top-down instruments given there is no official involvement or endorsement by state or political party officials, as is the case for political action committees for women in the US (Gichohi Reference Gichohi, Muriaas, Wang and Murray2020). However, research shows that top-down approaches fail as well, for example, in Portugal (Espírito-Santo Reference Espírito-Santo, Lépinard and Rubio-Marín2019) and in France (Lépinard Reference Lépinard2016), and that bottom-up approaches succeed, as in Malawi (Muriaas and Kayuni Reference Kayuni and Muriaas2014). Furthermore, directly targeting women aspirants or candidates as a group by reducing the campaign costs is often a response to failed attempts at establishing a legislative quota. As Box 2 indicates, there are four potential paths to bottom-up success.

The presence of centralized candidate selection (CCS) across all parties may make-up for the shortfalls of a bottom-up instrument (H5). Previous studies find that CCS favors women because party leaders can ensure compliance with a gender target with sometimes the help of a quota (Caul Reference Caul1999; Hinojosa Reference Hinojosa2012). Hinojosa (Reference Piscopo, Hinojosa, Thomas and Siavelis2021, 159) suggests that avoiding primaries or enforcing clear limits on primary spending could increase women’s access to political offices. Building on this, if women candidates are assured extra funding for campaigning in elections, this might be the needed incentive for party leaders in centralized nomination systems to select more women.

Centralized candidate selection may not make it alone as is indicated in Hypothesis 6. If there are few women in elected office, party leaders may struggle to identify women that they can nominate. Therefore, CCS in combination with the presence of a 15% minimum of women MPs, indicating a certain level of support for women’s status and rights and gender equality, may make-up for the weaknesses of a GEF instrument itself. One flaw of a centralized system is that there are limits to how well party leaders know what is happening on the ground because leaders depend on networks present in the party organization or networks of their representatives when incentivized to do so (Van Houten Reference Van Houten2009). Alternatively, a quota in combination with centralized candidate selection may contribute to successful outcomes (H7). In the absence of any sanction mechanism strengthening the quota or engagement from the government, parties with centralized candidate selection may be more able to manage a gender-balanced list; thus, a bottom-up GEF could provide the incentive needed to follow the imperative of a quota regulation.

Another enabler for successful bottom-up approaches may be the presence of public funding (H8). Studies that identify money as a barrier to political recruitment are often done in contexts where there is no public funding to political parties. In the absence of public funding, parties need to rely on the candidates being able to fund their own election campaigns, as is the case in Zambia (Wang and Muriaas Reference Wang and Muriaas2019). If public funding is available for political parties, this reduces the risk involved in selecting less affluent or less well-known candidates and trying out new women candidates, particularly if women candidates are self-funded and are obliged to ask for funding from party coffers.

What Was Left Out and Why?

The six conditions to be considered in the QCA are by no means the only factors identified by researchers that may combine with GEF implementation to enhance gender balance in representation. Here, the omitted factors are discussed including the rationale for their exclusion from the QCA, keeping in mind that they are taken into account in the causal mechanism case analyses below and should be considered in any future study of GEF.

Two major conditions were initially coded and considered for QCA: GEF target (candidate or party) and GEF as a party penalty rather than a payment. When the GEF target was included in a six condition QCA model, the limit for a population of 34 or less, extreme model ambiguity and failure to generate parsimonious explanations without contradicting simplifying assumptions resulted. In addition, the GEF target is more or less covered by the state-driven condition given that most of the candidate targeted instruments were non-state-driven and the party targets were state-driven. Whether GEF penalizes or provides payment as a condition is omitted given that less than one third of the cases have the presence of a party penalty, thus, not enough variation.

While feminist coalitions of state and society-based actors have been found to be crucial in promoting effective gender equality policy (e.g., Mazur Reference Mazur2002), given they mobilize around all cases of GEF implementation in the study, their presence or absence makes no difference in the outcome. Furthermore, the condition, 15% minimum of women MPs, allows for examining what might be the more essential part of feminist coalitions. The influence of international norms from extranational organizations like the UN, ILO, or EU, also identified as an important driver in gender equality policy (e.g., Zwingel Reference Zwingel2016), is present across all cases as well, except for the US, and thus is omitted from the QCA. The ideology of the majority party in power has also been considered important for gender equality policy with left-wing majorities often found to be the most supportive (e.g., Caul Reference Caul1999); however, for many new democracies in this study, the ideological right-left cleavage is not politically meaningful. Once again, the condition of a minimum of 15% women MPs may serve as a proxy for this omitted condition as well given that it indicates support for gender equality values in the political elite, which actually may be what is at play more than the presence of a left-wing majority. Finally, feminist institutionalists have shown path dependent dynamics or gender-biased norms established over a long period to be important obstacles to feminist change (e.g., Krook and Mackay Reference Krook and Mackay2011). In this study, gender-biased path dependency is a possible explanation for cases of failure; however, its influence can only be detected over time, not through the QCA, and only in countries that have more than one case of GEF implementation, Therefore, the question of path dependency can only be taken-up in the causal mechanism case studies.

Putting it All Together So Far: Expected and Unexpected Results

The csQCA generated expected and unexpected findings, presented in Table 3, laying solid groundwork for the two causal mechanism cases studies presented in the next section, particularly given that Creswell and Creswell (Reference Creswell and Creswell2017) argue it is necessary to move from larger-n analysis to case studies when results are unexpected in mixed methods studies.

Table 3. csQCA Findings: Expected and Unexpected

A CLASSIC TOP-DOWN APPROACH IN FRANCE (2002–2017): Parity Penalties through State Feminist Alliances and a Minimum Level of Women MPs Incentivize Women Candidates and Party Leaders

In France, the state-driven GEF scheme has been implemented since 2002 in conjunction with constitutionally mandated quotas, called parity.Footnote ¹⁵ France had been renowned for low and stagnant levels of women MPs prior to 2002, going from 4% in 1978 and reaching only 10% in 1997. A large and vocal coalition of feminist groups joined with women’s policy agencies inside the cabinet and parliament, using policy pressure from the UN, EU, and the Council of Europe to convince the Socialist government to adopt a constitutional amendment in 1999. A separate law came soon after on the specific mechanism of the penalties in National Assembly elections; political parties were targeted through a reduction in their government funding in proportion to the percentage difference between men and women candidates presented in the first round of elections across all constituencies.

In the first election in 2002, the penalties had little effect given their low level. The larger parties chose to take the relatively low financial hit in government funding rather than present women candidates in half of the constituencies and the result was a failure. In a context without a minimum level of women MPs and their networks of eligible women, the party penalties were not a strong enough incentive for party leaders to change their behavior. A 2007 law, adopted under the right-wing government, through the same coalition of feminist forces behind the parity reforms, although this time without any extra national influence, increased the level of the party penalty to 75%, and it was applied in the 2012 June parliamentary elections, with women winning 27% of the seats. In 2014, the powerful Ministry of Women’s Rights and Equality convinced the Socialist government to increase the fines one more time to 150% of party grants. In the 2017 legislative elections, women received 38.7% of seats in the National Assembly, an all-time high in France and a 28% increase since the parity sanctions had been implemented, but still 12% from the goal of numerical parity.

The increase clearly incentivized party gatekeepers to assure that women were selected for winnable seats. The steady rise of women MPs above the minimum level of 15% meant that more women were encouraged to run for office as well. The 2017 elections provided an unprecedented context for success. A new reform that limited individuals from holding more than two elected offices opened incumbent seats in the National Assembly for women newcomers, and the landslide victory of Emmanuel Macron’s new party paved the way for more women MPs given that the party had followed 50-50 parity in candidate selection. While it is difficult to parse out the effects of the reforms and the Macron landslide from the increase in penalties, the wide elite acceptance of parity indicated by the spread of “parity grammar” (Bereni and Revillard Reference Bereni and Revillard2007) and the increasing presence of women MPs certainly played an important role in the significant rise. Despite these successes, research has shown persistent gender biases; women are ignored when seats are in clearly winnable districts and women MPs continue to be kept away from positions of power and influence.

COMPLEXITY WITHOUT QUOTAS IN MALAWI (2009 to 2019): Feminist Mobilization and Women MPs Incentivize Women Candidates and Party Leaders

Failure of GEF implementation in the 2014 Malawi election was turned into success in 2019. What are the mechanisms that explain why a new type of GEF introduced prior to the 2019 was more successful? In Malawi, there has been constant engagement with trying to develop a GEF design without a legislated quota or a PR electoral system. The first GEF initiative was launched prior to the 2009 elections, as donors and domestic women’s organizations were concerned about the 87% share of male MPs. A 50:50 Campaign, with a parity goal and funded by donors, gave cash funds and other resources to women to facilitate their electoral campaigns. However, the role of the government was unclear, and elections were not completely free. The 50:50 Campaign was continued in the two following elections. The Ministry of Gender, Children, Disability, & Social Welfare coordinated the campaign in 2014. The president, Joyce Banda, appointed after Bingu wa Mutharika died in office, initiated several gender reforms and had a particular interest in the continuation of the GEF scheme (Wang et al. Reference Wang, Kayuni, Chiweza, Soyiyo, Muriaas, Wang and Murray2020). As is typical for the region, it is difficult to place Malawian parties on a left–right spectrum. In 2014, the implementation of the more top-down approach targeting women candidates resulted in failure, and the strong ties between a government and the GEF initiative were seen to be problematic. Joyce Banda, implicated in a corruption scandal, lost her bid for reelection and the poor election results for women were seen as a spillover effect of dissatisfaction with her performance (Kayuni Reference Kayuni, Banik and Chinsinga2016).

The top-down approach was also discredited; thus, for the 2019 election an NGO, the 50:50 Management Agency, took over the organization of the campaign in which political party leaders were more willing to participate. This time, its main funder, the Norwegian Embassy, collaborated with the management agency to reimburse the registration fees of women candidates. Parties also took an interest in the scheme because they did not have to pay the fee for women candidates (Malawi News, February 16–20, 2019). The situation was a win-win: the constituency branches did not pay registration fees for women, and party leaders could attempt to comply with calls to advance gender-balance. The combination of an independent agency in charge of GEF and that not only party leaders but also regional and local branch leaders were incentivized to select women, encouraged more women to step up as well. The newcomers had seen that other women had been successful in the past, thus showing that a 15% minimum level of women MPs prior to the implementation of a GEF was an important contributing factor. In all, Malawi can be seen as both a top-down and bottom-up success story, as the influence of GEF grew when there was a stronger demarcation between the government and the initiative.

Analytical Implications

So, what were the causal mechanisms behind failures being turned into success in the two countries? In both countries, this development occurred without PR, but with the incentives of GEF mechanisms, reduced costs for parties and women candidates in Malawi without a quota and increased costs to political parties for noncompliance with the parity quota in France. Another contributing dynamic in both countries was the presence of a 15% minimum of women MPs, which indicated the previous success rates of women candidates and party willingness to mobilize women’s networks. Feminist alliances were also instrumental in putting the GEF schemes in place and making sure they were implemented as well as evolving gender norms that represented critical junctures in gender-biased path dependencies. Despite the similar outcomes, the impetus and arena for the successful turn of events in both countries were quite different. In France, the legal restriction through penalties was a result of state feminist-based alliances and was implemented by the national administrative process through the public funding of the political parties. In Malawi, international donors funded a bottom-up process with feminist NGOs trying to keep the government at a distance through positive attention through payments that boosted the financial advantage of women candidates. Whether the GEF was targeted at parties or candidates made little difference in moving failure to success in either country.