The Transformative Potential of High-level Gender Equality: The relationship between gendered laws and perceptions in Rwanda

As part of its reconstruction process following the 1994 genocide against the Tutsi, Rwanda introduced several laws and policies protecting gender equality, which contradicted traditional patriarchal norms. This study focuses on the relationship between institutional gender reform and local perceptions. Specifically, it explores the extent to which perceptions of gender have caught up to legal changes and identifies where the greatest gaps exist across political, social, and economic spheres. Data collection occurred in two parts: quantitatively, a multiple-choice survey was distributed to 76 Rwandan adults investigating their perceptions of gender in political, social, and economic contexts. Qualitatively, structured interviews were conducted with three Rwandan gender experts in order to supplement and better explain survey findings. Both quantitative and qualitative results indicated that the greatest gaps between gender-related laws and perceptions exist at the social level, with the smallest in economic contexts. Analysis also revealed that overall, level of education was associated with more congruity between gendered laws and perceptions, while both sex and age were independent of responses. Based on these findings, this study provides recommendations for aligning gendered laws and perceptions.


Acknowledgments
I owe great thanks to many individuals who helped make this study possible. Firstly, I would like to thank my Academic Director, Celine Mukamurenzi, for her extraordinary support throughout all stages of the study. Celine was always willing to dedicate her time, effort, and wisdom to helping me develop my research. I am truly thankful for all her guidance.
Secondly, I would like to thank Dr. Desha Girod from Georgetown University for her support and thoughtful advice on data analysis, presentation, and organization. I feel extremely lucky to have had a mentor with such expertise in social sciences and conflict studies as I I also owe special thanks to Ninette Umurerwa and Fidéle Rutayisire, who also agreed to be interviewed, for their time, expertise, and thoughtful answers that enriched my data analysis and deepened my understanding of gender in Rwanda. Lastly, I owe thanks to the 76 survey respondents who graciously volunteered to participate in my study.

Chapter I: General Introduction and Background of the Study
Since the end of the 1994 genocide against the Tutsi, Rwanda has introduced several laws protecting gender equality across political, social, and economic spaces that stand in stark contradiction to traditional patriarchal norms. Through quantitative survey data and qualitative interview analysis, this study identifies and explores the areas where gaps exist between gendered laws and perceptions in Rwandan society.

Background of the Study
After the 1994 genocide against the Tutsi, Rwanda embarked on a reconstruction process that completely renovated the country's social, political, and economic institutions. Rwanda's project of post-conflict development focused not only on physical infrastructure, but also on rebuilding a national identity after polarized ethnicities drove Rwanda into civil war and ultimately, genocide. A strong emphasis on gender equality was incorporated into Rwanda's longterm reconstruction vision, standing in stark contrast to the highly patriarchal society that dominated Rwanda before the genocide. Goals for gender equality and women's empowerment across state-level institutions represent a tremendous leap forward. For example, Rwanda leads the world in women's representation of government, with 61% of the lower house of parliament being female ("Women in National Parliaments", 2019), while women's employment is the highest in Sub-Saharan Africa ("Women, Peace and Security Index", 2019).
Instituted during Rwanda's reconstruction process, provisions safeguarding gender equality are enshrined in the country's formal legal code. In 2003, Rwanda ratified a new constitution (revised again in 2015) guaranteeing gender equality and equal protection before the law. A 30% quota of female representation in government and gender mainstreaming throughout all state institutions was also mandated (The Republic of Rwanda, 2015, art. 75 and art. 80).
Several laws passed since the 2003 constitution have furthered the agenda of legal gender equality.
For example, organic laws 22/1999, 08/2005, and 43/2013 concerning land gave women equal rights to inheritance, land ownership, and co-ownership of joint property shared with their husbands (Abbott et al., 2018). Other post-genocide laws concerning women's rights include equal pay, the prohibition of workplace harassment, protection and punishment for gender-based violence, gender-sensitive budgeting across the public sector, and maternity leave.
In 1999, the Ministry of Gender and Women in Development was reorganized and launched (Burnet, 2008), and the Gender Monitoring Office was established by the 2003 constitution to oversee institutional compliance with gender mainstreaming and standards, further reinforcing state-level efforts towards gender equality.

Research Problem
Women's legal possession of rights does not necessarily equate to the ability to exercise them.
Although gender equality in Rwanda underwent dramatic legal reforms after the 1994 genocide against the Tutsi, the extent to which perceptions around gender have developed is less clear. It is thus important to measure the consistencies or inconsistencies that exist between gendered laws and perceptions because visions for full gender equality cannot be realized without congruity between the two. Although Rwanda has made remarkable progress with respect to gender equality at the institutional level, gaps within norms and perceptions of gender roles endure, creating demand for research into where and on what scale these incongruities exist.

1.3.Research Questions
This study seeks to answer the following questions: 1. To what extent are legal provisions protecting gender in Rwanda consistent with Rwandan perceptions of gender norms? 2. Which political, social, and economic-related laws in Rwanda are most consistent and inconsistent with local perceptions concerning gender?

Objectives of the Study
1. To explore the relationship and identify gaps between legal provisions and common perceptions of gender equality in Rwanda. 2. To identify key political, social, and economic areas where gaps exist between legal provisions and Rwandan perceptions of gender equality.

The Scope of the Study
This study took place in Kigali, Rwanda over a 4-week period beginning on November 2 nd , 2019 and ending on December 3 rd , 2019. Falling under the domain of sociology, this study specifically focuses on gender, social institutions, and their relationship to the law. Data collection occurred in two parts, on quantitative and one qualitative, each targeting a different population of interest. Quantitative data was collected through a survey given to a random sample of 76 Rwandan adults who met the following three criteria: they were citizens of Rwanda, they were at least 25 years old, and they had lived in Rwanda since 1994. The first criterion was important because the aim of this study is to examine the relationship between legal gender equality in the law and Rwandan perceptions of gender. Thus, my sample needed to be entirely Rwandan. The second two criteria were important because, in order to best measure the extent to which norms have caught up to institutional changes made since the end of the 1994 genocide, the most relevant information comes from Rwandans who have witnessed the entire period of change. All surveys were anonymous and completed on a volunteer basis. See section 2.2 for further details on survey data collection techniques.
The qualitative data for this study was collected through structured interviews with 3 gender experts in Rwanda. The purpose of these interviews was to gain complementary insight into the social dynamics of gender in Rwanda to help better understand quantitative survey results. While survey results identified where gaps exist, interviews with experts helped explain why they may exist where they do. All experts gave permission for their identities to be published in this study.

Data Collection Technics
Data was gathered in two parts: the first was quantitative and collected through a survey, and the second was qualitative and collected through structured interviews with Rwandan gender experts. The study and data collection processes were approved by a Local Review Board in Rwanda.

Survey Design
The survey was originally written in English and then translated into Kinyarwanda. All informants took the survey in hard copy, each numbered uniquely. The survey was split into three sections: the first was a coversheet providing general information about the study, including the contact information of myself and my Academic Director, criteria for taking the survey, instructions for taking the survey, guarantee of anonymity, and assurance of Local Review Board approval. The second section collected demographic information on each informant's sex, age, and highest level of education earned. Informants' names were not asked in order to protect their identities and increase the likelihood that they answered. The third section was comprised of 16 survey questions (15 questions, one of which had two parts, making a total of 16). Each question was written as a statement, to which respondents answered by selecting one of five possible answers. Nine questions were on a Likert scale, and informants checked one of the following boxes for each question: "strongly agree", "agree", "neither agree nor disagree", "disagree, or "strongly disagree". The remaining four questions provided five multiple choice answers for different ranges of time. Each statement corresponds to a specific law in Rwanda's legal code. 1 All questions on a Likert scale were formatted so that agreement indicated the most congruity with the law, while disagreement indicated the least. All questions with possible answers as ranges of time included one answer option that reflected exactly the law. Thus, survey answers helped identify where gaps exist between legal provisions and perceptions; questions with more agreement indicated smaller gaps between laws and perceptions, while questions with more disagreement identified greater gaps.
In order to identify specific areas where gaps exist between legal provisions and Rwandan perceptions of gender equality, survey questions were categorized as either political, social, or economic; the first four questions were political, the next six (seven including the question with two parts) were social, and the final five were economic. Definitions and criteria for these categories are outlined below: Political questions: Consider gender in the context of national decision-making power, including voting, government representation, participation in peacebuilding processes, and the role of gender-focused state institutions.
Social questions: Address gender at the household and community level, focusing on issues such as education, marital relationships, and land.
Economic questions: Center around gender in the workforce, looking at issues of compensation, maternity leave, and the importance of equality in sustainable national development.

Survey Distribution
Survey data was collected for 76 respondents, contacted through various means. Some surveys were given to members within my own network in Kigali, such as my host family and their friends, and employees of places I frequently visited. However, most survey responses came from large organizations and offices in Kigali. Using contacts given to me by Celine Mukamurenzi, my Academic Director, and those of guest speakers who lectured at SIT, I was able to drop several copies of my survey off at once to large office buildings in order to increase the volume and diversity of responses. Office staff offered a good population for a random sample, as members came from different backgrounds with different levels of education, and were of different ages and sex. All contacts were aware of my study and agreed to help distribute the survey prior to meeting. In those cases, I took 20-30 copies with me to meet the individual I was in contact with, left the surveys with them in their office for distribution, and collected them several days later. In all cases of survey distribution, I ensured a multiple-day gap between delivery and collection in order to allow participants to complete the answers on their own time.
Once surveys were collected, all data was entered into a Microsoft Excel spreadsheet.

Interview Processes and Participants
All 3 interviews with Rwandan gender experts were structured and conducted in English. I met each interviewee at their primary workplace in Kigali or another convenient location, provided an overview of myself and my study, and asked five standard questions. 2 At the end of each interview, I asked each expert if they would be comfortable with their name being mentioned in the study report or if they preferred to remain anonymous; all agreed to have their 2 For a full list of interview questions, please see appendix 4 identities known. The first interviewee was Arlette Iyakaremye, a gender consultant in Rwanda with former work experience in Rwanda's Gender Monitoring Office. Her current work as an independent gender consultant includes gender mainstreaming projects, training, and program coordination. Iyakaremye was my host mother in Kigali, and accepted a request to be interviewed on a volunteer basis. The second interviewee was Ninette Umurerwa, the National Executive Secretary of Haguruka, a Rwandan NGO providing legal aid to and defending the rights of women and children. Umurerwa is a lawyer by profession with expertise in gender. Dr.
Aggée Shyaka, an advisor for this study, connected me with Ms. Umurerwa. The third interviewee was Fidéle Rutayisire, Executive Director of RWAMREC (The Rwandan Men's Resource Center), an organization dedicated to promoting positive masculinity and engaging men in the fight for gender equality. Rutayisire is a lawyer by education and avid gender activist.

Ethical Values Observed during the Study
For the quantitative aspect of data collection, the gathering and publishing of personal information was the main ethical concern. Two strategies were applied to mitigate this concern.
The first was guaranteed anonymity to all survey participants. This was emphasized on paper on the survey's coversheet and in person upon survey delivery. Secondly, surveys were mostly delivered in bulk to be distributed among workplaces and networks, and respondents were given multiple days to take the survey before collection. In those cases, the survey was administered through a single contact at each office and not from myself directly. This helped ensure that informants felt comfortable taking the survey and felt confident that their answers would not be traceable to them. It was also emphasized to respondents that their participation was entirely voluntary.
A second ethical concern arose regarding survey design. In order to make respondents feel comfortable completing the survey, no questions or statements asked about personal experiences. Rather, all statements were phrased objectively and depersonalized, so that respondents didn't feel that that their privacy was being breached.
For the qualitative aspect of data collection, ensuring participant comfort was also a main ethical concern. Before all interviews, details surrounding the study and Local Review Board approval were shared to ensure transparency. I also told all interviewees before the interview that, after the interview, I would ask if they felt comfortable with their name being published in this study given their answers during the interview. All interviews took place at the primary workplace of the interviewees or another convenient location in order to facilitate a comfortable conversation and accommodate their schedules.

Limitations of the Study
Several practical and content-driven limitations affected this study. With only four weeks to complete the study, time for survey distribution and collection as well as interviews was limited to around three weeks. Given these time restraints, the potential volume of survey responses and expert interviews were limited. Further, time restraints contained the geographic scope of the study to mostly within Kigali, with the exception of one individual who distributed some surveys in Huye, a town in southern Rwanda. Time and geographic constraints limited the measurement of gaps between legal provisions and perceptions of gender across all parts of Rwanda.
Geographic constraints also limited the diversity of the study's final sample. Nearly all of survey respondents originated from Kigali. Thus, survey results lacked representation of rural-based Rwandans. Other sampling biases also existed given the network used for survey distribution. As described in section 2.2.2, surveys were mostly distributed among large offices through contacts from SIT. As a result, the majority of responses came from well-educated individuals. Thus, diversity of educational background was limited.
These sampling biases and geographic constraints have restrictive implications for the internal and external validity of this study, which was able to measure gaps in gender equality among only a small part of Rwanda's population.

Chapter III: Literature Review
A large selection of literature exists around both general theories of gendered social norms and relationships between men and women in Rwanda specifically, providing abundant context for this study. This chapter begins by examining general theories of social norms, continues by outlining gender in the case of Rwanda both before and after the 1994 genocide against the Tutsi, and concludes with an exploration of the relationship between gendered laws and perceptions, ultimately combined to generate expected findings for this study. Key terms are underlined and defined throughout.

Theory of Social and Gender Norms
Gender norms refer to patterns of behavior to which men and women are socially expected to conform because of their biologically determined sex. Gender norms are a subsection of social norms, which will first be explored before more closely examining gender norms. Bicchieri defines social norms as informal rules that dictate how members of a society ought to act. Conformity to social norms is derived from both empirical and normative reasons. Empirically, these informal rules are followed because individuals believe that most people in their reference network will also conform. Normatively, people conform to social norms because they believe that the majority of others in their reference network believe those rules should be followed (Bicchieri, 2017). Social norms are powerfully grounded and sustained through people's desire to belong in a given society, and they are enforced by social pressure or punishment against those who do not conform (Church and Chigas, 2019). The creation of social norms depends on the emergence of normative expectations first, which are then followed by empirical expectations, together solidifying into widespread obedience to informal rules.
Changing social norms can be a sluggish process because it requires a change in collective expectations; deviating as an individual is costly. Often, the process of norm change occurs generationally, as younger members of society are more likely to adopt new and different moral standards, initiating the emergence of new normative expectations (Bicchieri, 2017). However, the process of changing personal beliefs, especially those inherited by tradition, is slow and can be further weighed down when entire social structures support and enable the endurance of certain social norms.
Gender norms are a type of social norm and are exceptionally resilient. Patriarchal values imbued within fundamental social structures and institutions universally relegate women to inferior and less powerful positions in society compared to men. The structural dimension of gender norms reinforces empirical and normative conformity while reflecting enduring traditions of male privilege. York traces the historical root of patriarchal hierarchies to a biologically-driven division of labor by concluding that, due to the reproductive and nurturing roles of women, they undertook the role of primary caretaker and often stayed at home due to their nursing responsibilities (2011). This division laid the foundation of separate roles for men and women, which became associated with different levels of power. Gender norms are taught at the family level and reproduced generationally, perpetuating and deepening patriarchal values across all levels of society (Boudet et al., 2013). Patriarchal structures also create an enabling environment for physical, emotional, and psychological violence against women which becomes somewhat normalized as a way of preserving established cultures and protecting patriarchal norms (York 2011). Cultural violence, or "the day-to-day use of overt or implicit force to obtain one's ends in social relations" (Hudson et. al., 2008/09), is what creates structural violence, penetrating a host of social relations, far beyond just the family unit. The threat of violence reinforces the status quo and discourages efforts to challenge existing norms (Boudet et al., 2013); thus, patriarchal norms are self-reinforcing and exceptionally resilient.
The process of transitioning to new gender norms is not linear; new and old can co-exist and overlap, while segments of the population progress and different rates than others. For example, York finds a correlation between social class and the malleability of gender norms. While middle and upper-class individuals are more likely to be exposed to alternative viewpoints through education, more access to recreation, and more social mobility, working class families are more likely to enforce strong gender stereotypes of labor division, particularly after several generations of low social mobility (York, 2011). Boudet et al. (2013) find that the most powerful accelerating agent in normalizing gender equality is education for girls, which expands opportunities for social mobility and economic empowerment. Overall, however, normalizing gender equality requires "a constant dialogue between social norms, empowerment, capacity to aspire, and the structure of opportunities" for women across political, social, and economic dimensions (Boudet et al., 2013).

Gender in Pre-Genocide Rwanda
Pre-genocide Rwanda was defined by strong patriarchal norms and systems, confining women to inferior societal positions both privately and publically (Kagaba, 2015). At the household level, a strong gendered division of labor was starkly visible, with women commonly working informally within the domestic sphere to take care of children, cooking, and cleaning while men embodied the traditional breadwinner role. Rwandan women often married at a young age, inhibiting their full access to education (Doan, 2010). Kagaba notes that husbands beating their wife was not uncommon and that it was a practice nobody questioned, reflecting strong male authority (2015). The threat of violence paired with strong norms of male power disabled women from challenging their husband's decisions, leaving little latitude for autonomous agency in the home (Brown, 2019).
Norms of patriarchal dominance were reinforced by laws marginalizing women, reflecting both public and private oppression. Legally, women had no claim to inherit property and lacked the right to seek paid employment or open a bank account without the permission of their husbands (Berry, 2015). In cases where women did work formal, paying jobs, it was common for their husbands to control their income (Burnet, 2008). The 1992 family code formally recognized men as the heads of Rwandan households, reinforcing patriarchal power within family units and in greater society (Brown, 2016).
Politically, women represented just 5 percent of the top branch of the MRNDD, the Hutuled ruling party of Rwanda during the years leading up to genocide. The party also actively repressed women's rights movements, which were perceived as threats to the country's patriarchal society and non-democratic leadership (Brown, 2018).
This illustration of pre-genocide Rwanda is steeped in patriarchal norms, attitudes, and structures. From both the top and bottom levels of society, men enjoyed privileged positions of power that relegated women to an inferior status.

The Effect of Conflict on Gender: The Case of Rwanda
Cases of conflict often dissolve socially constructed gender roles alongside institutions and state order. The transition out of war and women's increasing agency are not merely cotemporaneous, but are causally related (Brown, 2016). In Rwanda, genocide disrupted traditional gender relations and created space for greater inclusion of women in political and social capacities. Brown traces this to the period during genocide, when processes of militarization and mobilization amplified women's agency, triggering the disruption of traditional gender hierarchies (2016). Politically, post-genocide Rwanda underwent a host of state-level changes during the early phases of reconstruction, which offered opportunities for gender reform. Women's advocacy groups, such as PROFEMMES, fought to ensure gender equality was on the reconstruction agenda, and played a large role in the passage of the 1999 land law granting equal ownership rights, the 2003 constitution protecting equality before the law and mandating a 30% quota of women in government, and the 2009 law against gender-based violence (Debusscher and Ansoms, 2013).
Rwanda's vision 2020, rolled out by President Paul Kagame in 2000, offered a blueprint for gender inclusivity as a cross-cutting theme throughout national development and growth objectives.
Rwanda's post-genocide political transition created opportunities for both grassroots, civilian-led activism that was inaccessible before 1994, and female leadership at high levels.
Socially, the genocide killed a disproportionate number of men and skewed national demographics to 70% female in mid-1994 (Gervais, Ubaligoro, & Nyirabega, 2009). While many husbands were either killed or imprisoned, women often became the heads of households (Burnet, 2008), reflecting a family and community-level shift in authority.
Brown concludes that Rwanda would not have witnessed such a large paradigm shift in such a short period of time without the breakdown of gender roles traced to genocide (2016).
However, these shifts in gender dynamics were products of necessity, a large absence of men, and post-conflict situational instability that challenged the entire state system of Rwanda. They were not caused by organic changes in values and beliefs, foreshadowing the existence of gaps between institutionally-led reform towards equality and grassroots perceptions of gender roles.

The Relationship between Laws and Norms in Rwanda
While, on one hand, women obtained greater agency and political power in post-conflict Rwanda, a number of limitations interfered with the full realization of gender equality at top and bottom levels. Transformative potential measures the capacity to exercise existing rights (Debusscher & Ansoms, 2013), and is relatively low in many sectors of Rwandan society. These incongruities are consistent with Boudet et al.'s observations that contexts change must faster than corresponding norms (2013), inhibiting full transformative potential. Overwhelming evidence finds that women's empowerment in public and legal spheres has not translated equally into domestic, private spaces. Deeply embedded cultural values reportedly continue to construct women as inferior, presenting paralyzing barriers to the implementation of gender equality laws (Abbott et al., 2018).
For example, even though land laws protect the equal ownership rights of women, land is still frequently regarded as 'really' belonging to men (Abbott et al., 2018). In practice, women's co-ownership does not always translate into equal say in land management, highlighting gaps between top-level, official changes and implementation in practice. Additionally, lack of knowledge among women concerning their land rights presents a significant barrier to exercising them; Abbot et al. (2018) find that only 9 percent of women had been trained on their rights, with most relying on community meetings and media for information, which can often be incomplete or inaccurate. Within education, although access is equal, gendered expectations still restrain girls in terms of how many years they attend school and the subjects they study (Boudet et al., 2018).
Furthermore, the opportunity cost of education for girls is relatively high because despite attending school, their expected quantity of household work doesn't lessen (Boudet et al., 2018).
These inconsistencies display a collision between top-level gender equality and entrenched patriarchal perceptions. Intersectional gaps are also evident: Wallace (2008) notes that gains have disproportionately benefited elite women while Brown (2016) finds a stark gap in impact between urban and rural areas. This shows that institutional-level reforms do not reach the population symmetrically; different groups respond differently and progress may be seen on some fronts while not on others.
Beyond just gaps in gender equality implementation, some reports find that institutional changes have even backfired to produce unintended, negative consequences for women. For example, within marital relations, reports from Rwandan women find that in some cases men retaliate against changing gender policies through physical or psychological abuse, with the most resistance coming from older men (Kagaba, 2015). Reporting marital abuse continues to be a challenge for women, despite strict laws against domestic violence. Women report persisting stigma and shame for reporting their husbands in addition to violent retaliation from their husbands (Kagaba, 2015). Thus, exercising their rights in cases of domestic abuse often worsens women's individual security by producing unintended suffering, incentivizing many women to opt for silence (Kagaba, 2015) and exemplifying again the obstructing role that norms can play in the transformative potential of women's rights and equality. Complementary to enduring patriarchal perceptions, Debusscher & Ansoms (2013) identify five categorical trends that threaten a fully transformative gender equality policy in Rwanda. The first is the dominance of an underlying economic rationale behind gender equality: as articulated through Rwanda's Vision 2020 and National Gender Policy, equality is encouraged as a means to maximizing economic output and productivity; its emphasis is extrinsic as opposed to intrinsically valuing equality as a human right.
Secondly, national policies neglect 'invisible labor' that is disproportionately dominated by women in the areas of agriculture and childcare; lack of policies here legitimizes unequal divisions of labor. The third, fourth, and fifth trends criticize the macro-level operationalization of gender equality reform, limited space for civilians to influence policy, and a lack of grassroots participation in decision-making.
Despite advances in women's legal protections and rights, assessments of implementation are mixed and ridden with gaps, generating conflicting images of gender equality. Legal rights and female political participation are necessary, but insufficient in fully realizing transformative potential; they do not eliminate the masculine regime, norms, and structures that produce gender equality (Berry, 2015). Research finds that structurally equalizing gender must begin with youth and is particularly effective when integrated into education frameworks (ICRW et al., 2011).

Expected Findings
Based on literature around norms, gender, and the social climate of Rwanda, expected findings include the existence of significant gaps between gendered laws and perceptions.
Discrepancies are expected to be particularly wide within the social dimension, given previous evidence found of enduring patriarchal power within the home (Abbott et al., 2018;Boudet et al., 2018;Kagaba, 2015) Table 1 summarizes demographic information collected from respondents.

Trends in Political, Social, and Economic Perceptions of Gender
When broken down by politically, socially, and economically-categorized survey questions, economic statements yielded the most agreement on average while social questions yielded the least. 4 Figure 1 shows the percent distribution of Likert answers across categories, found by summing the total number of answers for each option on the scale of agreement. Economically-categorized questions yielded most agreement on average, with more than 96 percent of answers being either "strongly agree" or "agree" and less than 3 percent of answers reflecting disagreement. Politically-grouped questions received an average of about 88 percent agreement or strong agreement, while about three quarters of socially-oriented questions reflected agreement.
The overall consistent agreement seen in economic survey responses reflects Rwanda's overall approach to gender equality reform, which has been mostly macroscopic and economically motivated. Debusscher & Ansoms (2013) criticize this top-level approach for limiting the full transformative potential of top-level initiatives. As outlined in Rwanda's National Gender Policy and Vision 2020, language around gender equality reflects a strong underlying economic rationale, justifying reform based on productive potential. This narrative rationalizing gender equality for its national economic benefits reflects survey results: the transformative potential of economic-based perceptions of gender seems to be high, reflecting overall acceptance to the idea that equality in the workforce produces positive effects on national development and growth, increasing the wellbeing of everyone in society.
On the other hand, survey results indicate less transformative potential within the social dimension. This is also consistent with the economically-focused national framework for gender equality, suggesting that top-down efforts have not been sufficient in fundamentally dismantling patriarchal norms at their roots within households and communities. Deeper analysis of politically, socially, and economically-driven survey data is further explored in the following subsections.

Analysis of political data
Overall, data from politically-categorized survey questions and interviews suggests that political perceptions of gender are mostly aligned with Rwandan law, but not completely; gaps were significantly wider for some questions than for others. However, the individual question with the most agreement came from this category; more than 98 percent of total respondents agreed that "men and women should equally participate in official conflict prevention, peacebuilding, and reconciliation processes". Figure 2 shows the distribution of answers for this question.

Figure 2: Percent Distribution of Answers to Question 3: "Men and women should equally participate in official conflict prevention, peacebuilding, and reconciliation processes"
Strong support for gender equality in peacebuilding efforts also reflects the dominating narrative that gender equality is important for macroscopic development and prosperity. During Although strong support is evident for female participation in post-conflict peace processes, relatively high resistance was exhibited toward the 30 percent quota of female representation in government. Overall, nearly 1 in 5 respondents disagreed with the 30 percent quota. Figure 3 shows the distribution of responses to the statement, "women should occupy at least 30% of decision-making positions in government, including in the Chamber of Deputies and the Senate". Although agreement still outweighs disagreement towards the 30 percent quota, disagreement here is much higher than it is towards women in peacebuilding initiatives, as shown in figure 2. Interestingly, twelve respondents agreed that women should participate in peacebuilding initiatives in equal proportions to men, but disagreed that women should represent at least 30 percent of parliamentary seats. This, in tandem with the nearly 20 percent disagreement with the quota, highlights that respondents show resistance to women in positions of political influence in some areas more than others.

Interview insights from Ninette Umurerwa, an expert in both legal and gender affairs in
Rwanda, highlights that women still hesitate to run for office due to traditional patriarchal norms that discourage women from assuming positions of power. 5 Iyakaremye adds that the gender quota is enforced by the government because it's relatively low cost and is emblematic of Rwanda's top-level narrative of gender equality and female empowerment, but doesn't necessarily reflect changed norms concerning women in power. 6 Relatively high resistance was also seen towards equal suffrage, with nearly 1 In 5 respondents disagreeing that men and women should be able to equally vote for government representatives. This disparity with the law is also likely derived from social roots that dissociate women with important decisionmaking. As seen in the cases of women's parliamentary representation and voting powers, social norms penetrate the political sphere. The next section will more thoroughly examine data collected on perceptions of gender within social spaces, which are least congruent with the law.

Analysis of Social Data
Social survey responses yielded the most disagreement of all three categories. The three questions within this group demonstrating the greatest discrepancies between laws and perceptions were all focused around marital relationships; specifically concerning a husband's ability to determine his wife's access to employment and the principal of marital rape.
As shown in Figure 4, nearly half of total respondents disagreed with the statement that husbands should not be able to determine their wife's access to employment, indicating that they believed husbands are entitled to make that decision. norms and laws endure. Iyakaremye noted that men still mainly decide how money is allocated and spent and make the majority of household decisions, 7 further contextualizing the relatively high support for the role of husbands in determining their wife's access to employment, as seen above in Figure 4.
Marital rape reinforces the persistence of strong patriarchal norms within the family unit.
As figure 5 shows, disagreement was relatively high towards the notion that marital rape should be a matter of legal concern; less than half of responses indicated agreement while nearly 36 percent of respondents answered with "disagree" or "strongly disagree". Even among respondents who agreed that marital rape should be a legal concern, opinions on appropriate punishment for perpetrators were significantly less extreme than the law, which mandates that marital rapists can face between 6 months and 2 years in prison (Law no. 59/2008, art.5). Of the 37 respondents who agreed that martial rape should be legally forbidden, about three quarters indicated that they thought perpetrators should receive either less than 6 months in prison or no jail time at all. Figure 6 shows the full distribution of what respondents (only those who agreed that marital rape was a legal crime) perceived to be just punishments for marital rape.  marital rape is particularly interesting. Relative to the 88 percent of respondents whose perceptions on marital rape far underestimated the law, the much higher support for severe prison punishment for rape suggests that the two forms of sexual violence are thought of differently. Marital rape is either not considered rape, or it is somehow perceived as significantly less problematic than when rape occurs outside the home. Iyakaremye clarified this discrepancy by commenting on the fact that the concept of marital rape is not widely accepted as wrongdoing, and that women are taught before marriage that they should never refuse sex from their husbands. 8 This reiterates the enduring patriarchal climate within the home, with men assuming decision-making power and greater control over marital relations. Rutayisire adds that the 2009 Gender-Based Violence law, in which the punishment for marital rape is stated, "threatened the manhood" of many Rwandan men, reflecting a culture of "negative masculinity" that is particularly powerful within the home. 9 During the period immediately following genocide, rape was considered and punished as a category one crime of genocide, the most extreme category. Thus, it is possible that the severity of rape (outside the home) has been normalized as a grave offence. Marital rape, on the other hand, was only recently criminalized in the 2009 Gender-based Violence law. Thus, either the recent nature of its criminalization or the greater threat that punishing marital rape poses to masculine power within the home, or perhaps a combination of the two, could explain why perceptions are so different for these two types of sexual violence, and why a significantly larger gap exists between laws and perceptions around marital rape.
Both experts emphasize that the most powerful persistence of patriarchal norms, and largest gaps between laws and perceptions, exist at the household level. 10 This idea is reflected by the fact that the three survey questions with responses most incongruent to the law all pertain to marital relationships.
Interestingly, survey answers for both questions on rape were independent of respondents' age, sex, or level of education (further explored in section 4.4), suggesting that these cultural norms are so deeply engrained in the cultural conscience of Rwanda that they cut across demographics.
Survey results indicate that the issue of land and gender is far more aligned with Rwandan law. Specifically, 91 percent of respondents agree that women should be able to own and manage their own land regardless of marital status, while approximately 95 percent of answers conveyed agreement toward equal land inheritance for sons and daughters. However, information gathered from Iyakaremye, Umurerwa, and Rutayisire emphasized that incongruities between laws and perceptions around land are some of the most visible. Rutayisire highlighted that there is a problem of awareness around the 1999 land law, which outlines equal ownership and inheritance rights for men and women. 11 Thus, many Rwandans still operate according to traditional norms of male ownership and management of land because they are not properly informed about changed laws. Umurerwa adds that, in her professional experience with legal disputes, the most common spousal conflicts arise from land. 12 She emphasized that the land law faces a huge implementation gap because of cultural norms endorsing male authority and decision-making power. According to Umurerwa, issues of land management "create a dilemma for women between their rights and their marriage". In other words, women often fear challenging their husbands' authority regarding land, even if they have a legal right to, because it may jeopardize her marriage and relationship with her family, who would likely condemn her challenging behavior. This dilemma highlights a critical barrier to the transformative power of women's land rights; although the law formally protects these rights, the ability to exercise them is hindered by enduring patriarchal norms that endorse male authority in the home.
As seen in the cases of household decision-making, marital rape, and matters of land, large gaps exist at the family and community levels between perceptions of gender and gender equality in the law. The location of these gaps within the social dimension corroborate a key criticism of Rwanda's approach to reform: it has been too driven toward macro-level objectives around productivity and growth, but hasn't focused on dismantling the fundamental norms of patriarchal power at their core. The extrinsically-motivated narrative and justification for gender equality has glossed over negative masculinity and patriarchal power embedded deeply within Rwandan social norms, offering one potential explanation for why gaps between laws and perceptions are widest at the social level.

Analysis of Economic Data
Economically-categorized survey questions received by far the most overall agreement; 96 percent of all answers to economic questions were either "strongly agree" or "agree", highlighting particularly strong alignment between laws and perceptions for this group. For example, more than 97 percent of respondents agreed that men and women should be paid equally for the same work. Further, 95 percent agreed that women on maternity leave should be compensated in benefits equal to their normal salary while nearly 90 percent believed the length of maternity leave should be either as long as or longer than what the law mandates, which is 3 months of paid leave. Figure 8a and 8b show the distribution of agreement for paid maternity leave and opinions on duration of maternity leave. As further explored in section 4.3, agreement for paid maternity leave and chosen length of leave were independent of respondents' age, sex, and level of education, suggesting that strong support for paid maternity leave is universal.
The vast majority of respondents also agreed that gender equality is important for the sustainable socio-economic development of everyone in society, with only about 1 percent disagreeing. Data for this question is particularly important because it sheds light on overall survey trends between political, social, and economic questions on perceptions of gender. The distribution for this question on gender equality and national development is shown in Figure 9. As has been previously emphasized, Rwanda's framework for gender equality reform in the law has been driven mostly by national growth and productivity-related objectives.
Responses to question 15 epitomize the core of this framework, and show strong, universal support for this nationally propagated rationale. On the other hand, the lack of nationally propagated language and efforts towards redefining gender norms at the local and family levels is reflected by significantly greater gaps between laws and perceptions within the social sphere.

Relationships between Demographic Information and Perceptions of Gender
Interestingly, demographic information collected from survey informants was mostly independent of answers. Determined by the Chi Squared statistical test for independence, 13 the 13 See appendix 3 for a sample calculation of the Chi Squared test variation of agreement and disagreement across survey responses overall was independent of respondents' sex or age. The one exception was level of education; the Chi Squared test concluded that informants with more education were more likely to agree or strongly agree to survey questions overall. The corroborates the literature on changing social norms, which generally finds that individuals with more education are more exposed to diverse ideas and thus more openminded to non-traditional norms (Boudet et al., 2018;ICRW et al.,2011).
Individuals with access to university are also correlated with higher wealth and, as York (2011) finds, wealthier individuals with greater social mobility are also more likely to adopt newer, more progressive ideals.
However, when considering questions within the political, social, and economic categories, level of education was independent of agreement or disagreement. This shows that, although education overall is associated with greater congruity to Rwandan law around gender equality, it is not necessarily linked to all areas of gender. More specifically, education was completely independent of responses to the most disagreed-with questions concerning perceptions around marital rape, punishment for marital or general rape, and the power that husbands have to determine their wives' access to employment. Within the politically, socially, and economically-categorized survey questions, both age and sex were also entirely independent of responses.
For questions with most agreement, such as those about equal participation in peacebuilding mechanisms and the importance of gender equality for national development, neither sex nor age nor education were associated with agreement or disagreement.
What these trends show is that overall, patterns of agreement and disagreement cut across various demographics in Rwandan society. The gaps between gendered laws and perceptions seem to be wide in some areas, such as those concerning marital relationships, regardless of demographic background. This implies that certain cultural values are deeply embedded within a resilient cultural schema, and thus aren't easily affected by variations in sex, age, and education.
Rwanda's approach to gender equality, which has been branded according to the economic objectives equality can produce, reinforces this. The national framework of reform has focused far less at dismantling the fundamental patriarchal power structures that exist at the root of Rwandan society within homes and families. Thus, while perceptions have clearly evolved within economic and some political contexts, social norms have not caught up to the same extent.

Overall Congruity between Gendered Laws and Perceptions
Evidence from quantitative and qualitative data suggests asymmetric progress on the alignment between gendered laws in Rwanda and local perceptions of gender; gaps are far wider in some areas than in others. Perceptions reflected the most similarity to the law economically, the second most similarity politically, and the least similarity socially.
Expected findings for this study were partially realized. 14 As predicted, gaps were widest within the social dimension, corroborating the criticism that the approach to gender reform was too focused on economic objectives and failed to dismantle patriarchal perceptions at their root, within family and communal spaces. However, although level of education was associated with greater congruity between laws and perceptions, respondents' sex and age were independent of answers, presenting findings different than expected. Inferred reasoning for this has been previously explored, but is also tied closely to the macroscopic and economically-tied motivation 14 See section 3.5 for gendered reform, that is criticized for neglecting structures of patriarchal powers within grassroots-level, social spaces.

Recommendations for Closing Gaps between Gendered Laws and Perceptions
Insights from both literature 15 and interviews revels that closing gaps between gendered laws and perceptions, particularly within the social sphere, requires a multidimensional approach. Realizing the transformative potential of equal rights demands systematic renovations; both top and bottom-level, formal and informal mechanisms must be activated to change collective beliefs at local levels and dismantle entire social structures that endorse patriarchal power. Thus, a number of recommendations arise: Firstly, raising awareness of existing laws protecting gender equality is a pre-requisite for attitude change; citizens, particularly women, must be aware of the rights they have in order to exercise them. 16 Secondly, men must be engaged in the journey toward equality; they must play an active role in advocating for change and bringing women forward. When men are engaged, they are less likely to feel that gender equality jeopardizes their power or produces loss for them. 17 Dialogue is an especially important mechanism in this process, as it brings members of the community together to openly talk about rights and injustices.

Appendix 2: Survey Questions and Corresponding Laws
The survey used in data collection contained 16 combined statements and questions. 12 of the 16 survey items were statements while 4 were questions. For each statement or question, respondents were given 5 multiple choice answers to select from. Survey statements, listed in Table X along with their corresponding laws, were answered on a Likert scale where respondents selected one of the following options: "strongly agree", "agree", "neither agree nor disagree", "disagree", or "strongly disagree". The survey questions, listed in Table Y along with their corresponding laws, were answered based on a selection of different ranges of time. Specific ranges for each question are mentioned below each question.

6A
If you selected "strongly agree" or "agree" to statement 6, how long should they go to jail for? (If you didn't answer with "strongly agree" or "agree", leave this question blank). For the purpose of data analysis, survey statements and questions were categorized as either political, social, or economic, as labeled in the second column of Table X and Table Y. Economic Maternity leave for women should be paid in benefits equal to their normal salary.
Article 11 of guarantees paid maternity leave and states that "maternity leave benefits are equal to the last woman's salary" (Law no. 003/2016) 15 Economic Gender equality is important for sustainable socio-economic development for everyone in society.
Rwanda's National Gender Policy states that "the Government of Rwanda attaches great importance to the promotion of gender equality and equity as a prerequisite for sustainable development" (Ministry of Gender and Family Planning, 2010).

6A Social
If you selected "strongly agree" or "agree" to statement 7, how long should they go to jail for? (If you didn't answer with "strongly agree" or "agree", leave this question blank). Possible answers: "no jail time", "less than 6 months", "6 months -2 years", "2 -5 years", or "more than 5 years" Article 19 of Law no. 59/2008 on the Prevention and Punishment of Gender-based Violence states that "any person who coerces his/her spouse to sexual intercourse shall be liable to imprisonment of six (6) months to two (2) years." (Law no. 59/2008, art.19) 8 Social How long should perpetrators of rape be imprisoned for? Possible answers: "less than 6 months", "6 months -5 years", "5 -10 years", "10 -20 years", "more than 20 years" Article 16 Law no. 59/2008 on the Prevention and Punishment of Gender-based Violence states that any person who is guilty with rape shall be liable to imprisonment of ten (10) years to fifteen (15) years. Where rape has resulted in a bodily or a mental illness, the person guilty with rape shall be liable to imprisonment of fifteen (15) years to twenty (20) years and medical care fees for the person Chi Squared (X 2 ) Formula: X 2 = Σ ( − ) 2 Calculating expected values of disagreement counts for each category of education: Total number of either "disagree" or "strongly disagree" answers survey-wide = 107 Expected value = (total number of disagree answers (106)