Introduction

Over the past four decades, decentralization reforms have become an enduring political and public policy strategy across various countries, especially in the developing world. The increasing emphasis on decentralization as a concept and practice is mainly due to its underlying tenets associated with democratic governance such as promoting grassroots participation in politics, improving local governance and empowering local actors (Adjei et al. 2017; Beall 2005). Consequently, many developing countries have adopted various forms of decentralization as evident in an array of institutional, structural and administrative reforms geared towards improving performance, responsiveness, efficiency and effectiveness of governing institutions including public, private and civil society organizations (Adjei et al. 2017; Jabeen and Iqbal 2010).

As an important element of participatory democracy, decentralization is usually viewed as essential for achieving women’s political representation and participation in decision-making processes (Pini and McDonald 2011). This is because, it constitutes the closest form and most accessible level of government for marginalized groups such as women due to its link with the concept of subsidiarity (Adjei et al. 2017). Again, it has been noted as an important ingredient for championing women’s political empowerment and gender equality (Tsikata 2009), in line with the Sustainable Development Goals 5 and 16. As Beall (2005) points out, decentralization holds real opportunities for women and can serve as an important site for their empowerment.

Despite the centrality of decentralization in advancing women’s political representation and gender equality, few studies have focused on women’s political involvement in local-level politics and the local government arena (see for example, Adatuu and Apusigah 2018; Amponsah et al. 2019; Gyan and Mfoafo-M’Carthy 2022; Gyan et al. 2022; Pini and McDonald 2011). Existing research has largely studied women’s political representation in national parliaments and legislatures, national political executive offices as well as those occupying ministerial positions (Adams et al. 2016). A notable aspect of this phenomena has been the exponential research on gender quotas and reserved seats for women across various polities to address women’s political under-representation (Dahlerup 2013; Tsikata 2009). As such, it is common in the literature to see a great bias of women’s democratic representation having a ‘national focus’. This bias neglect local politics and its decentralized political structures which mostly undermine women’s political representation. As Pini and McDonald (2011) argue, women’s experiences and perspectives on local government representation and their substantive representation remain ‘overwhelmingly gender blind.’

Shifting focus from the dominant ‘national’ orientation, this study investigates women’s political representation at the sub-national level, which is often glossed over in the literature. The importance of researching women’s representation in local government is because women are not only disproportionately affected by services under local government control, such as social care and childcare, but also because local governments serve as important conduits for regional and national governments (Pini and McDonald 2011; Roberts 2017). Turning to women’s representation in local government units can, therefore, help address pervasive structural conditions and gender differentials that influence women in politics and to better understand local-level political dynamics that affect women. For instance, in the United Kingdom, recent evidence shows that women constitute only 33% of local councillors in England, 27% in Wales and 24% in Scotland as well as just 15% of local authority leaders in England (Roberts 2017). Besides, studying the local government arena is also crucial for highlighting the interface between rural politics and gender, where global evidence about women’s actual presence in local government is still very scarce.

Our study addresses two main questions. First, how have assemblywomen under Ghana’s local government system performed in the delivery of democratic representation defined as a function of responsive and downwardly accountable leadership? Second, why have women’s political representation and participation as local government authorities remained relatively lower compared to their male counterparts? These questions are framed in light of ongoing academic, policy and practical discussions that point to increasing women’s political representation nationally and internationally. While some scholars argue that women’s descriptive representation can translate into substantive representation of women (Celis et al. 2008; Lovenduski 2005), others also contend that it is also required of women political leaders, especially those in local government, to perform their roles effectively (Bari 2005; Graff 2003).

This article is organized as follows. After this introductory section, the next section provides a theoretical overview on women’s political representation. This is followed by a section that situates the political representation of women in the Ghanaian context. The fourth section details the methodology used for the study. Next, is the results section which provides the findings of the study. This is followed by the discussion section which sheds light on the findings and its broader implications for women’s representation. The seventh section offers a conclusion and policy recommendation in light of the findings.

Theoretical and contextual review

Gender and political representation: a theoretical framework

According to Pitkin (1967), there are four types of representation—formalistic, descriptive, symbolic and substantive representation. Formalistic representation is used to define the formal and institutional arrangements that precede and initiate representation by conferring authority to a person to act for others. In her view, descriptive representation refers to the extent to which a representative resembles those being represented. As Pitkin (1967) further asserts, descriptive representation places too much emphasis on political institutions and individuals within such organizations rather than their activities. Thus, it becomes problematic since individuals cannot be held accountable for ‘who they are’ but only for ‘what they have done’. Symbolic representation refers to the meaning that a representative has for those being represented and is greatly based on the beliefs and attitudes of the represented. Lastly, Pitkin defined substantive representation as a form of representation where representatives act on behalf of those they represent. Pitkin viewed substantive representation as the one true type of representation because the actions by representatives are taken on behalf of, in the interest of, as agents of, and as substitute for the represented (Pitkin 1967).

Descriptive representation and substantive representation remain the foremost types of Pitkin’s conceptualization most widely discussed in the gender and politics literature. This often stems from the argument that descriptive representation does little for women’s representation vis-à-vis substantive representation which is seen as the ultimate form for counting women’s gains in politics. Indeed, Pitkin (1967) noted that descriptive representation often breeds self-centeredness without responsiveness and accountability and often fails to evaluate representatives by whether they are good delegates or good trustees. The ability of representatives to act for the good of those they represent and their accountability for their actions makes substantive representation the best form of representation from Pitkin’s viewpoint.

Some school of thought emphasize that increasing women’s representation may not necessarily create an environment favorable for women-friendly policies (Dovi 2007). For instance, as Walsh (2012) has demonstrated, the effects of party quota by South Africa’s leading political party, African National Congress (ANC), while increasing the number of women in parliament, had little impact on championing women-friendly policies. She further asserts that, the ANC only handpicked women party loyalists, thereby allowing party leaders to centralize their power even though they had the idea of being committed to women’s equality. To this end, some scholars argue that the performance women actually deliver based on their capabilities as political authorities should become the basis for their increased participation under substantive representation (Graff 2003; McGregor and Webster 2017).

In contrast, proponents of descriptive representation also argue that the numerical representation of women in politics and the public sphere is crucial for making inroads for women’s participation given their historically relegated roles. As Lovenduski (2005) contends, descriptive representation is imperative for furthering women-specific policies. In particular, some feminist scholars and political scientists have argued that descriptive representation of women matters in terms of whether the people who occupy institutional positions are women given the gendered nature of politics and the public sphere (Duerst-Lahti and Kelly 1995). Arguments for descriptive representation espouse the importance of adequate representation of women in contemporary politics and provide six main reasons why it matters for women’s representation (Dovi 2007). These six distinct arguments include the role model argument, the justice argument, the overlooked interests argument, the transformative argument, the trust argument and the legitimacy argument (Phillips 1998; Dovi 2007).

The role model argument makes a strong case for women’s descriptive representation on the grounds that the presence of female representatives in a political system improves female citizens’ self-esteem and sense of political efficacy (Dovi 2007). For instance, seeing a woman in a high-rank political office such as president can encourage female citizens, especially young females to aspire such leadership positions and also boost their self-confidence and sense of possibility (Campbell and Wolbrecht 2006). The justice argument emphasizes that fairness demands that men and women be present in a roughly equal number among political representatives across various political institutions (Dovi 2007). This argument further critiques the monopolization of representation by men (Phillips 1998) and the ‘masculinization of politics’ across various societies. The overlooked interest argument contends that the presence of female representatives improves democratic deliberations and political agendas because such female representatives are believed to have more insight and concern about women’s issues than do male representatives (Childs 2006). The transformative reasons for women’s descriptive representation rests on the idea that having more women political leaders improves democratic institutions through enhancing the approximation of their democratic ideals (Dovi 2007). Hence, the presence of female representatives is noted for shaping the norms and practices of democratic institutions (Phillips 1998), which ultimately transforms the political system. The trust argument contends that due to the past betrayals of female citizens by male representatives, increasing the number of female representatives is necessary for renewing the confidence of female citizens in political institutions (Gay 2002). Lastly, some scholars stress that having more female representatives increases the legitimacy of democratic institutions (Phillips 1995, 1998).

Situating women’s political representation in the Ghanaian context

The political representation of women in formal politics in Ghana remains very low (Ministry of Local Government and Rural Development 2016; hereafter MLGRD; Odame 2010), despite growing efforts to increase women’s political participation. At present, there is a general consensus regarding the under-representation of women in politics, which is often attributed to the gendered political order in Ghana’s formal politics (Bawa and Sanyare 2013). Within this order, men are conceived as the front runners, occupying political positions. This is informed largely by established norms and practices that regard men rather than women as leaders. Further, with ownership and control of household assets, vested in men, they emerge as the ideal candidates for political party flag bearer positions (Bawa and Sanyare 2013). Women, however, remain in the background and merely participate in programs that court public support for male political party candidates. Following this gendered political order, women’s role until recently was narrowly confined to membership of campaign teams, while the opportunity to assume political leadership roles was considered the reserve of men (Bawa and Sanyare 2013).

To this end, increasing women’s representation in parliament, national executive offices and other high rank political offices has been an important agenda among various stakeholders (Adams et al. 2016; MLGRD 2016). Women’s descriptive representation has been highlighted as an important avenue for empowering Ghanaian women, increasing women’s participation in politics, de-gendering the political arena and promoting women’s political rights on grounds of fairness (Bawa and Sanyare 2013; Ministry of Gender, Children and Social Protection 2015; hereafter MoGCSP).

Consequently, there have been some positive feats for women’s representation in Ghana considering the strides made by some notable women political leaders in the recent past. For instance, Ghanaian women have held political positions such as Speaker of Parliament, Chairperson of the Council of State, Chief Justice, National Electoral Commissioner (MoGCSP 2015), and flagbearers of smaller political parties. Besides, for the first time in the nation’s history, the National Democratic Congress (NDC), one of Ghana’s two mainstream political parties, presented a female vice-presidential candidate during the 2020 general elections. These attempts have all been geared towards improving women’s national political participation and leadership capacities, which could translate to the substantive representation of women and also creating a well-functioning democratic system (Adams et al. 2016).

At the local government level, the few existing studies have highlighted various initiatives put in place to promote women’s political participation in local governance and development (MLGRD 2016; Odame 2010). For instance, Odame (2010) notes that sensitization efforts should be directed towards women to erase deep-rooted misconceptions that politics is for men, in order for women to compete favorably with their male counterparts in local government. As data from the MLGRD (2016) has shown, only 11 out of the 170 metropolitan, municipal and district chief executives (MMDCEs) in 2008 were women. The MLGRD data further demonstrates a marginal rise in this number from 11 to 12 and later 17 female local assembly executives out of 216 MMDCEs in Ghana. Again, existing research on gender and politics, have also provided some insights into the major impediments to women’s political participation and representation in Ghana (Bawa and Sanyare 2013; MoGCSP 2015; Odame 2010).

However, a critique of the current literature in Ghana is that aside demonstrating the barriers to women’s political representation (see for instance, Tsikata 2009), it does not give critical insights into what female representatives do for those they represent. As such, our understanding of the actual performance of women political leaders towards advancing women’s interests, on the one hand and serving their electoral areas as a whole on the other hand is often not discussed. As argued by Ballington and Karam (2005), calls for gender equality and women’s empowerment should go beyond window dressing of political institutions with greater number of women to empower female representatives to deliver beneficial roles to those they represent.

Methodology

Research design

This study adopted a cross-sectional case study research design. This approach was employed to solicit the responses of community members as well as to ascertain the experiences of multiple local political actors (such as assembly members, Metropolitan, Municipal and District Assemblies (hereafter MMDAs) officials, Electoral Commission officials and household heads) concerning women’s representation in three district assemblies in the Ashanti Region of Ghana. There were eight selected communities from the three districts namely Abrepo Junction, North Suntreso, Aketego, Kwamo, Bomso, Fumesua and Nyameani and Konkoma. The case study was appropriate for the study focusing on the Ashanti Region, which is the most populated region in Ghana with the highest number of MMDAs in the country.

Study setting

This study was conducted in selected communities from three MMDAs in the Ashanti Region of Ghana, namely the Kumasi Metropolitan Assembly (KMA), Ejisu-Juaben Municipality (EJM) and Bosomtwe District (BD) from January to March 2016. Figure 1 is the map of Ghana showing the MMDAs chosen for the study. The communities in the MMDAs were purposively selected for the study because of their interesting dynamics concerning the election of male and female representatives within the last two decades. The three MMDAs were also strategically chosen for their geographical relevance to women’s representation in local government: the KMA is highly urbanized, followed by the EJM which has peri-urban characteristics and then the BD which is more ruralized.

Fig. 1
figure 1

Map showing the study districts in regional and national context

Full size image

From the 2010 Population and Housing Census (PHC), the population of the Kumasi Metropolitan Area was estimated at 1,730,249 representing 36.2% of the total population of Ashanti Region. As of 2010, it composed of 826,479 males (47.8%) and 903,779 females (52.2%). With 52.2% of the total population being women, it is ironical that women have low representation in almost every level of decision-making especially at the local level. For instance, from the 2010 and the 2015 District Assembly elections, women respectively constituted only 16.1% and 9.4% of the total elected members for the Oforikrom sub-Metropolis selected from the KMA to be studied.

Concerning the Ejisu-Juaben Municipality, it shares boundaries with the Asante Akim North to the East and the Kumasi Metropolis to the West; the Sekyere East and Afigya Kwabre Districts to the North–East and North–West respectively and the Bosomtwe and Asante Akim South Districts to the South. The EJM has a surface area of 582.5 km2 with Ejisu as at the district capital (GSS 2010). The estimated population of EJM from the 2010 PHC was 143,762 comprising 68,648 males (47.8%) and 75,114 females (52.2%) females. In 2010, only 6.4% of the elected assembly members were women as against 93.6% who were men. In the district level elections in 2015, the percentage of women as elected assembly members increased marginally to 14.9% as against 85.1% of men despite women constituting more than half of the total of the district.

The Bosomtwe District Assembly shares boundaries with the Ejisu-Juaben Municipality to the East and Atwima Kwanwoma District to the West, the Kumasi Metropolis to the North and with Bekwai Municipality and Bosome-Freho District to the South. The district has a land size of 422.5 km2 with Kuntanase as the District Capital (GSS 2010). The estimated population of the Bosomtwe District, according to the 2010 PHC, was 93,910 representing 2.0% of the region’s total population. Males constituted 47.7% and females represented 52.3%. Despite majority of the population being females, women’s representation at the district assembly level is very low. In the 2010 district level elections, women made up only 5.7% of the total elected Assembly members for the district, which increased only marginally to 6.3% in the 2015 district level elections. Figure 2 shows the total population of males and females in the study areas and corresponding disparities of elected males and females to the respective local government assemblies for 2010 and 2015 election years.

Fig. 2
figure 2

Gender representation of assembly members in the study areas

Full size image

Data collection methods and sampling

Data collection for the study involved surveys and in-depth interviews with community members and local government leaders in selected communities in KMA, EJM and BD. In all, 180 participants comprising selected community members, officials from each of the selected MMDAs and the electoral commission, assembly members (both past and present), heads of households and household members in the selected study areas were involved in the study. Questionnaires were the principal means of data collection employed to ascertain how communities perceive and operationalize responsive and accountable local representation, as well as the performance of assemblywomen. This primarily involved heads of households and other household members.

Considering the gendered focus of our study with emphasis on women’s political representation and performance, 116 out of the 150 questionnaires were administered at electoral areas where women had previously been elected as their district assembly members but were replaced with men in the most recent local government elections. The remaining 34 respondents were administered in communities where men had previously been elected as district assembly members but had also been replaced with women as assembly members in the earlier mentioned elections. Out of the eight electoral areas chosen for the study, Abrepo Junction, North Suntreso, Aketego, Kwamo and Konkoma had previously elected a female as district assembly member but replaced them with males; whereas Bomso, Fumesua and Nyameani had males as their previously elected district assembly members but had replaced them females.

In addition to the surveys, focus group discussions (FGDs), in-depth interviews and observations were used to solicit the views of community members, officials of district administration, assembly members, Electoral Commission officials and traditional authorities from the selected MMDAs used in this study. The in-depth interviews and FGDs were conducted to offer insights from different local actors on the performance of assemblywomen. This was done to better understand the responses provided by the survey data from household heads and household members. Interviews were conducted in Asante Twi, which is the widely spoken language in the study communities and lasted on average for 40 min. The responses from interviewees were written in English to confirm content validity and for easy understanding. Again, six FGDs (two each in the KMA, EJMA, BDA) involving at least four and maximum eight community members, were also conducted to supplement the in-depth interviews.

For the surveys, household members and heads of households were recruited using a simple random sampling. These respondents were randomly selected from the study communities to ensure an unbiased representation of community views on women’s representation and performance in local government. Again, purposive sampling technique was used to select community residents and key informants including officials of the district administration, electoral commission, traditional authorities and district assembly members for the study. The study communities were North Suntreso, Abrepo (in the Bantama Sub-Metropolis), Aketego and Bomso (in the Oforikrom Sub-Metropolis) electoral areas from the Kumasi Metropolitan Area, Kwamo and Fumesua electoral areas from the Ejisu-Juaben Municipality as well as Konkoma and Nyameani communities from Bosomtwe District.

Data analysis

Descriptive statistics with frequencies and percentages and factor analysis were used to organize, analyze and present quantitative data obtained with the aid of the Statistical Package for Social Scientists (SPSS) software. The factor analysis tool was used to extract the primary and most significant factors that undermine effective political representation of women as local government authorities in the selected communities. The quantitative analysis was supported with relevant transcriptions of focus group discussions, in-depth interviews and observations using a thematic content analysis.

Results from the study

Background of study participants

A total of 180 research participants were involved in the study. Out of this, 150 people were residents of the eight local communities in the Kumasi Metropolis, Ejisu-Juaben Municipality and the Bosomtwe District. Concerning the 150 community residents, 60 were from North Suntreso, Abrepo Junction, Aketego and Bomso in the Kumasi Metropolis, 54 were from Kwamo and Fumesua in the Ejisu-Juaben Municipality and the remaining 36 were from Nyameani and Konkoma in the Bosomtwe District. The remaining 30 respondents comprised 6 officials of the Electoral Commission and the District Assemblies recruited for interviews and 24 community members selected for the FGDs. From the eight chosen communities, Abrepo Junction, North Suntreso, Aketego, Kwamo and Konkoma had previously elected a female as district assembly member but replaced them with males whereas Bomso, Fumesua and Nyameani had males as their previously elected district assembly members but had replaced them with females. This was considered suitable for gathering relevant and adequate data on how communities perceive the performance of women as local government authorities compared to their male counterparts in the study areas. Table 1 shows the sample distribution and communities selected for the study.

Table 1 Sample distribution and gender representation
Full size table

Delivery of accountable and responsive local representation

One of the aims of our study was to examine the performance of assemblywomen in Ghana’s local government system. This was done to ascertain how female local government representatives delivered accountable and responsive local government representation in their local communities. Concerning accountable representation, our analysis was based on assembly members’ awareness of decentralized responsibilities, ability to mobilize resources for projects considered as preferred by the communities, effectiveness of project delivery, reporting (information delivery) and reaction to sanctions. With a total multiple response of 464 counts for accountability assessment in all eight study communities, our study results show that, 68% (34% each) of the respondents ranked the performance of their district assemblywomen as either very good or good. In contrast, 24% and 8% of the respondents respectively perceived the performance of their assemblywomen as being poor and very poor. In this regard, assemblywomen’s ability to demonstrate awareness of their decentralized responsibility, mobilize resources to effectively deliver community projects and also report on their performance to their constituents was considered to be good. Almost all the participants in the study emphasized empowerment of women to participate in local government as a necessary condition considering their ability to deliver accountable leadership at the MMDAs. The detailed results of respondents’ perceptions about the performance of assemblywomen in the delivery of accountable representation are shown in Fig. 3.

Fig. 3
figure 3

Respondents’ perceptions of assemblywomen’s performance in the delivery of accountable local representation

Full size image

Furthermore, our study sought to examine the responsiveness of assemblywomen concerning their delivery of demands signaled as preferred or prioritized by their local communities. The key indicators used to assess responsiveness of assemblywomen in the study areas included their ability to carry out needs assessments in their communities, their engagement with community members for their opinions and views and the actual delivery of prioritized development projects. Our findings show that, with total multiple responses of 348 counts, 67% of the respondents ranked the performance of their district assemblywomen concerning responsive representation as either very good (37%) or good (30%). On the other hand, 22% and 11% of respondents perceived the performance of their assemblywomen as being poor and very poor, respectively. Generally, the responses from our research participants show that, district assemblywomen performed well by carrying out needs assessment through community engagements and delivering what their communities prioritized or signal as preferred (see Fig. 4). For instance, At Bomso, one of the study communities where their male assembly member was replaced with a female, a respondent mentioned during a focus group discussion that, ‘when the assemblywoman came, our road was very bad but she has managed to fix it even though she hasn’t been long in office.’ As such, community members were of the view that female local assemblywomen were effective at delivering responsive local government through their representation.

Fig. 4
figure 4

Respondents’ perceptions of assemblywomen’s performance in the delivery of responsive local representation

Full size image

Performance of assemblywomen compared to their male counterparts

The performance of female representatives towards the promotion of women’s interests and that of their polities as a whole, remains a keenly contested phenomenon. With calls for more women in politics and the public realm, there are also critical voices to move beyond ‘mere representation’ of women in politics to consider how well they perform towards improving the lives of those they represent. After all, it could be argued that some male representatives may perform better at representing women’s interest and beyond. To this end, we investigated the opinions of community members on the performance of their district assemblywomen compared to their male counterparts concerning delivery of responsive and downwardly accountable local representation. Using the same accountability and responsiveness indicators shown in Figs. 3 and 4, the aggregate responses show that the performance of assemblywomen ranked slightly below that of their male counterparts. Using the various accountability indicators as shown in Fig. 5, 81% of respondents considered the performance of their assemblymen as generally good (53% = very good, 28% = good) compared to 68% of the respondents who indicate same (34% = very good, 34% = good) for assemblywomen. From our findings, male representatives at the local government level were generally rated higher than their female counterparts in their respective communities.

Fig. 5
figure 5

Respondents perceptions of assembly members performance in the delivery of downwardly accountable representation

Full size image

Similarly, majority of the respondents in the study communities ranked the performance of district assemblywomen below their male counterparts regarding responsive representation as presented in Fig. 6. We find that 88% of respondents ranked the performance of their assemblymen as generally good (54% = very good, 34% = good) compared to 67% of the study participants who reported same for their assemblywomen (37% = very good, 30% = good). Overall, while assemblywomen and assemblymen were perceived by their community members as doing well in the delivery of responsive and accountable local government representation, their responses suggest that assemblymen perform better than their female counterparts.

Fig. 6
figure 6

Respondents perceptions of assembly members performance in the delivery of responsive representation

Full size image

Explanatory factors for low women’s representation and performance as local authorities

In Table 2, we present 12 factors emphasized by respondents as reasons for women’s low representation in local government as assembly members in the study communities. These factors were broadly socio-cultural, political and behavioral factors that served as barriers for women’s political representation at the local government level. We use a factorial analysis to extract the primary and most significant factors that undermine the effective political representation of women as local government authorities in the selected MMDAs in the Ashanti Region of Ghana. In order to ensure that the components causing variation are 70% or more to arrive at a reliable conclusion, all factors with initial communality extraction values less than 0.50 were taken out as less significant factors. Consequently, those with extraction values above 0.50 were retained for further analysis. As a result, 8 out of the 12 factors based on their extraction values greater than 0.50 as shown in Table 2 were used for the next level of analysis.

Table 2 Initial communality extraction of the factors undermining the participation of women in local governance
Full size table

Further analysis of the explained variance for the extracted factors presented in Table 3 showed that six out of the eight factors with Eigen extraction values greater than 0.5 retained in the factor analysis had a cumulative percentage of 73.59%. From the rotated component matrix shown in Table 4, the six factors identified are lack of self-confidence, low educational levels, lack of affirmative action policies, cultural context, fear of losing elections and religious beliefs of men as heads. These six factors out of all the direct and indirect factors identified and included in the factor analysis accounted for almost 74% of why women’s representation as local government authorities in the study areas is relatively lower. Hence, those factors occur as the most significant or direct factors that explain assemblywomen’s low representation in the study areas. Thus, a component matrix is shown in Table 4 to identify the six significant factors.

Table 3 Explained variance for extracted factors
Full size table
Table 4 Rotated component matrix for the factor analysis process
Full size table

The six significant factors influencing women’s representation in local government as district assembly members were determined by the rotated component matrix table because the factors were perpendicular to one another. Therefore, the corresponding factor to the highest value in each component was drawn amongst the six significant factors affecting women’s political representation as assembly members in the Kumasi Metropolitan Area, Ejisu-Juaben Municipality and Bosomtwe District in Ghana as presented in Table 4.

Among the six explanatory factors, fear of losing elections and lack of affirmative action policies constitute political factors whereas lack of self-confidence, the cultural context, low levels of education and religious beliefs of men as heads accounted for socio-cultural factors for women’s lower representation in local government compared to men. In general, women’s low self-confidence, low educational levels and lack of affirmative policies were found to be the most crucial factors inhibiting the representation of women in local government. As emphasized by a traditional leader at Nyameani the selected communities during an interview:

Women need adequate education and training to effectively participate in local government as assembly members because the work of the assembly is about understanding of communities, and effective communication in English. If you cannot read and write, what business do you have standing for elections? (February 16, 2016).

The statement by the traditional leader sheds light on the reality that given women’s relatively low educational levels, they are mostly unable to communicate effectively with their electorates. In view of the fact that, English remains the primary means of conducting (local) government business and communication, women may tend to shy away from such roles due to their low educational attainment. This subsequently feeds into their lack of self-confidence to take up local government political roles or perform well at their chosen roles, particularly given the lack of affirmative policies at the local government level. As such, some participants expressed concerns for addressing various barriers to women’s active involvement in local government. For instance, during a focus group discussion at Konkoma, a unit committee member opined that, ‘the roles of women in local government and development decision-making at all levels are very important if we are to develop as a country. Therefore, factors that limit us from participating in local governance must be addressed’ (February 26, 2016).

Discussion of results

With increasing calls for increasing women’s representation in politics in both academic and policy spheres (Chattopadhyay and Duflo 2004; Dovi 2007), there has also been strong voices to consider female representatives’ performance in political leadership positions. This latter position argues that, while efforts at women’s descriptive representation ought to enhance women’s participation in politics, female representatives also need to exercise good performance in their chosen and elected roles (Bari 2005; Graff 2003). Rather than just having ‘to be there’, their performance is likely to engender more women into political leadership based on their capabilities as well as improving female citizens’ participation in politics. Thus, the performance of female representatives is noted as crucial for women’s descriptive and substantive representation (McGregor and Webster 2017). As shown in our findings, assemblywomen in the selected study communities performed well in their accountability and responsive representation to those they represented, including women. As McGregor and Webster (2017) have previously emphasized, female elected representatives may not necessarily see themselves as representing only women, but their entire communities, of which women form an integral part.

Concerning accountable representation, female representatives performed well in their awareness of decentralized roles and their effectiveness for delivering community projects. Awareness of their roles as assemblywomen provides a basis for their answerability to their communities, for which the latter can sanction them for non-performance. Thus, assemblywomen’s awareness and understanding of their local government roles and the enforcement of sanctions by their electorates for not representing their best interests frames the conduct of their relations. This is, therefore, likely to enhance the delivery of local community projects as found in the study areas. The ability of assemblywomen to mobilize resources to deliver effective projects holds promise for accountable local government, should women have an active political participation and strengthened leadership capacities. As the findings of Chattopadhyay and Duflo (2004) in India have shown, when women have greater voice and participation in public administration and policymaking, public resources are more likely to be allocated accountably towards investments in human development priorities such as child health and access to employment. This further enhances the practice of democratic decentralization while attending to the needs of their communities, especially women-friendly policies.

Furthermore, our analysis showed that the performance indicators for assemblywomen’s responsive representation were high concerning community engagements and the delivery of development projects in their electoral areas. Female representatives’ ability to actively involve community members in the delivery of development projects and sometimes carrying out needs assessment regarding where projects were better served highlights their willingness to execute projects signaled as preferred by and in the best interests of their community members. The delivery of short-term development projects and policy outputs deemed necessary by citizens and what they receive from political leaders highlights the responsiveness of political leaders (Kang and Powell 2010). Consequently, having responsive assemblywomen plays a pivotal role towards the realization of democratic decentralization. Moreover, building on their community engagements, assemblywomen were able to respond to diverse community needs in the implementation of local development policies and projects in their electoral areas. It is therefore, along this route, that arguments for substantive representation of women becomes beneficial. This is because, female local representatives as part of their broader community roles can address general societal concerns, including that of women, due to the diversity of felt needs they respond to in their local areas.

Despite the promising performance of assemblywomen, their overall performance compared to men was lower. This situation was more pronounced in the rural electoral areas where assemblywomen’s performance was merely satisfactory. From a comparative viewpoint, community members’ low perception of assemblywomen’s performance points to gender issues in Ghana’s local government system. For instance, the gender distribution in the selected MMDAs revealed a limited involvement of female representatives as assembly members despite the general female population being higher than the male population. This is particularly evident by the very low representation of elected assemblywomen compared to men in 2010 and 2015 for the predominantly rural electoral areas in the Bosomtwe District. The performance gap in the representation of women in local government also underscores the various barriers to women’s political participation under Ghana’s decentralized system, which is further discussed.

As found earlier, six main factors accounted for the relatively low representation of women in Ghana’s local government system compared to men—lack of self-confidence, low educational levels, lack of affirmative policies, cultural factors, fear of losing elections and religious beliefs of men as heads. These multifaceted socio-cultural, political and behavioral factors combine to hinder women’s representation and performance as local government authorities in the Ashanti region of Ghana. As previous research has shown, women in both advanced and fledgling democracies face considerable political, socio-cultural, economic and behavioral constraints that limit their political representation, and subsequently their participation in politics and public life (Tsikata 2009; Pini and McDonald 2011; McGregor and Webster 2017). As Bari (2005) further contends, the common pattern of women’s political exclusion stems from socio-political discourses, political structures and institutions, cultural norms and functional constrains that restrict women’s individual and collective agency.

Concerning women’s lack of self-confidence, the historicity of women’s background roles in democratic governance and formal politics partly explains this situation. Having been confined and socialized into the family domain and private life, most women tend to have doubts over their performance as political leaders (Odame 2010). As such, female local government representatives may not be very confident and assured of themselves to take up political leadership positions. As Odame (2010) emphasizes, this situation is likely to feed into their fear of losing elections. When women feel less confident about occupying political leadership roles, they may be held back by the very idea of putting themselves up for elections for fear of losing. This could therefore be a contributory factor for women’s relative low representation in local government.

Paradoxically, while gender based affirmative policies have over the years been instituted to address women’s under-representation, especially at the national level, the local government sphere reveals a contrasting picture. The lack of such affirmative policies at the local government level to engender women’s representation highlights the failure of affirmative policies to ‘touch’ the roots of governance espoused by democratic decentralization. Gender based affirmative action policies in many developing world contexts appear effective on paper but are often poorly implemented in practice (Tsikata 2009). As such, though Ghana’s affirmative policies such as the Affirmative Action Policy formulated in 1998 and the National Gender Policy introduced in 2015, emphasize women empowerment, mainstreaming gender equality, and reducing women’s political under-representation at all levels, this is yet to be fully achieved. For instance, the third principal objective of Ghana’s National Gender Policy seeks to promote women’s leadership and accountable governance (MoGCSP 2015), however, the policy is yet to make a mark on women’s political representation in local government. Owing to this, the preoccupation of such affirmative action policies with national level politics in areas such as increasing women’s representation in national legislature, executive and judiciary positions means that the local government sector is grossly overlooked. This situation likely corresponds to the declining political representation of elected female local government authorities since 2010 (Electoral Commission of Ghana cited in MoGCSP 2015: p. 10) and the failure to achieve the declared goal of 40% female representation in politics and the public sector (Tsikata 2009).

Furthermore, educational attainment has been highlighted as an important ingredient for achieving women empowerment, improving women’s functional capacities and engendering their active involvement in politics (Kabeer 1999). This is because education has the potential to enhance women’s understanding of politics and decision-making processes that impact their lives. This becomes more crucial in local government where women are often disproportionately disadvantaged in terms of political representation. However, women tend to have lower educational levels compared to men due to various setbacks such as total denial of formal education for patriarchal reasons, financial constraints, priority of male children towards formal education and socialization processes (MoGCSP 2015; Tsikata 2009). In view of these hindrances, very few women are able to attain adequate levels of education required to occupy political leadership positions and to address women’s concerns in a democratic manner. This situation further shadows women’s political representation in local government, hence, the relatively low numbers of assemblywomen.

As emphasized in previous research, socio-cultural norms remain a critical issue for women’s political representation, especially in contexts where patriarchal structures exist (MoGCSP 2015; Tsikata 2009). Due to existing discriminatory cultural practices and social norms politics is often regarded as a male affair unwelcomed to women. Hence, the cultural context, especially in rural electoral areas showed a disproportionately lower number of female assembly members despite women having a relatively high population than men. Again, traditional beliefs and political gender stereotypes also perpetuate ideologies and traits of men as more capable political leaders than women (Huddy and Terkildsen 1993). As previous studies have emphasized, political gender stereotypes often go against women’s political representation and legitimizes women’s political subordination and under-representation (Arat 2010; Blackman and Jackson 2019).

Besides, the double burden of women political leaders to balance politics and family duties, especially childcare, in the face of lack of gender-sensitive working policies, also deters most women from engaging in active politics at both local and national politics. As such, not only do women lose out on descriptive representation but also suffer from having a collective voice to speak up for themselves on their issues democratically, due to their low political representation in local government (MoGCSP 2015). As argued by some feminist political theorists, the shortage of women in political institutions may not only have serious consequences for the political agenda but may also limit the articulation of women’s interests (Phillips 1995; Lovenduski 2005).

An important factor tied to the cultural context is religious beliefs that entrench men as heads. Religious beliefs, ideologies and doctrines have a powerful influence on how women perceive themselves in relation to men as far as everyday life and politics is concerned. As shown by Arat (2010), the intertwining of Islam and politics in Turkey spreads patriarchal religious values and sanctions women’s secondary roles through the public bureaucracy, educational system and civil society organizations. Given that, a considerable number of African societies are highly religious, adherence to religious beliefs and doctrines shapes how women engage in politics. In some traditional Ghanaian societies such as the Akan and Dagomba, male-dominance in political decision-making is partly informed by their belief systems. Thus, where religion reinforces patriarchy and systemic male dominance in politics, women are often faced with enormous challenges towards the realization of descriptive and substantive political representation (Arat 2010; Blackman and Jackson 2019). Religious beliefs of men as heads could, therefore, justify, legitimize and entrench male superiority in politics which become a ‘normalized’ aspect of national politics, including the local government arena.

Conclusion and policy recommendations

Our study investigated women’s political representation and performance as local government authorities under Ghana’s district assembly system. Our findings show that assemblywomen perform well in their delivery of responsive and downwardly accountable representation in selected electoral areas in three district assemblies in the Ashanti Region, namely the Kumasi Metropolitan Assembly, Ejisu-Juaben Municipal Assembly and Bosomtwe District Assembly. However, they can do better with motivation and capacity building to improve women’s descriptive and substantive representation as well as the broader community representation. Moreover, we have shown that assemblywomen’s performance and involvement as local government leaders is relatively low compared to men. We therefore find a gender performance gap in Ghana’s local government system based on community members’ perceptive evaluations, which stems from the interplay of socio-cultural, political, and behavioral issues. Through the analysis of factors affecting women’s representation in sub-national governance, we find that six significant factors—lack of self-confidence, fear of losing elections, lack of gender-based affirmative policies, low educational level, the cultural context and religious beliefs of men as heads—account for the relatively lower representation of women in the study communities.

As a result of these barriers that restrict women’s political representation and performance in local government, appropriate capacity building interventions and streamlining of gender-based affirmative action policies are needed to encourage more women to show interest and actively participate in local government and other public sector positions in the country. An important step will be to expedite the passage of Ghana’s long-awaited Affirmative Action Bill, which inter-alia, seeks to promote a progressive increase in women’s active participation in politics and public life by 2030, in line with the Sustainable Development Goals 5 (Targets 5c and 5.5) and 16 (Targets 16.6 and 16.7). This way, there will be more women at the district assemblies to improve the legitimacy of such sub-national institutions, champion women’s interests and also fully represent their communities.

Also, as a response to the coordination and implementation deficiencies towards gender mainstreaming in local government, relevant ministries such as the MLGRD and the MoGCSP should work together to address budgetary, institutional and capacity constraints of local government units such as inadequate financial resource allocation, delays in releasing funds, rigid bureaucratic procedures and limited human resources which subvert the promotion of women’s political representation. Other key stakeholders such as non-governmental organizations, traditional authorities and civil society organizations should also play advocacy roles to promote equitable gender representation at all levels of government. Lastly, in order to avoid poor targeting, the National Commission for Civic Education (NCCE) and District Assemblies must prioritize community sensitization programs and media advertisements for women to improve their local government representation and also promote inclusive local governance in Ghana.