Are grassroots Non-governmental organization actors’ policies pro-poor? Questioning inclusive beneficiary targeting and people-centred policy rhetoric in Ethiopia

The study analysed the interface among pro-poor targeting, gender inequality and women beneficiaries of programme-based grassroots level intervention in urban setting in Ethiopia. Using feminist lenses of service delivery, social capital, intersectionality and change theories, a quasi-experimental time design was employed to assess the intervention impacts. From desk review, questionnaire and interviews data, the impact of the much-advocated pro-poor grassroots urban agriculture programme prove mixed results. The study result shows that grassroots interventions are neither inclusive in targeting nor it prove pro-poor purpose. The evidence shows a widening gender gap even among beneficiaries. The study thus concludes that interventions by grassroots non-governmental organizations are neither pro-poor in purpose nor the actors are inclusive in targeting. The study revealed doubts on people-centeredness of grassroots interventions and actors. Thus, a feminist rights-based inclusion framework should inform grassroots interventions since such framework is based on intersectional perspectives, gender-sensitive indices and critical of categorical targeting of beneficiaries. The grassroots evidence finally brings new insights on agriculture – programme-based urban agriculture and women empowerment in agriculture index and brings new insight on intervention logic of women rights in policy design and evaluation.


INTRODUCTION
Women empowerment approach evolved from debates spearheaded by feminist researches, advocacy and policy practice in the last seven decades (Muchtar et al., 2019). Evolving as an integral component of poverty studies, the approach attracted researchers, policy practitioners and advocacy groups since the late 1960s (Yntiso, 2015;UN Women, 2019). Its relevance gained momentum in explaining the gender gaps in political, economic, social and collective measures of women as individual and collective agents of change in the last two decades (Muchtar et al., 2019). It also involves inclusion and exclusion debates regarding development contexts, processes and outcomes (UN Women, 2019). Policy practitioners, researchers and gender activists were *Corresponding author. E-mail: filmon.hadaro@aau.edu.et.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License interested in women"s collective agency (as group and individuals) (Bradshaw et al., 2019). Women"s self-help associations and movements were argued as mechanisms of pooling assets capabilities (Smith, 2014).
Grassroots non-government organizations, individual or consortiums, were assumed as actors that manifest inclusion in targeting and pro-poor and people-centred in addressing the needs of the poor (UNDP, 2020;UN Women, 2019).
These organizations are seen as proactive in transformative discriminatory contexts, processes and benefits of the poor; and are seen as tools of empowerment (Yntiso, 2015;Kabeer, 2016). The scholars argue that collective agency gives the poor, including women, the capacity to negotiate with powers of exclusion (Smith, 2014). In Ethiopia, from Asian experience, women collectives were seen as tool for gender equality; and argue that woman"s collectives and grassroots organizations as pro-poor and inclusive in targeting (Mindaye, 2014;Gizaw, 2009).
Ethiopian grassroots community institutions are rigid, exclusionary; also limit women"s economic, social and political empowerment; not as such, reciprocal social capital as argued by Mindaye (2014). Thus, both grassroots collectives of poor women and interventions by non-governmental organizations need to be questioned; that is, whether these collectives are venues to fighting rigid institutions and discriminatory practices is subject to further inquiry (Daniel, 2021).
Also, whether grassroots actors are inclusive in targeting poor women (having intersectional gender barriers) and interventions are pro-poor and peoplecentred as argued by scholars is up for further inquiry (Narang, 2012;Smith, 2014). Mindaye (2014) argues that grassroots women collectives such as self-help groups not only empower poor women members but also give them the power to fight rigid and exclusionary institutions and actor in Ethiopia.
As empowerment has several dimensions (Dijkstra and Hanmer, 2000;UN Women, 2019), and empirical studies are scant and limited; inquiries on whether women"s membership into grassroots collectives and support interventions to these collectives improve beneficiaries": (i) economic, social, political and personal status; (ii) inclusive targeting to intervention benefits; and (iii) whether the interventions and actors are pro-poor and people-centred; and (iv) whether their collective agency influence institutional powers.
Previous studies considered positive impacts of grassroots interventions, actors and membership-based associations and do not address the negative aspects nor question the mainstream models. This study thus questions the exclusionary aspects of grassroots interventions, actors and women"s collectives and the widening gender inequalities.
The gaps tip several inquiries; (1) how categorical and objectivist indices hide gender gaps of beneficiaries and shape intervention targeting biases? The temporal design in this study is employed to measure and observe the impact before and after the intervention. (2) What existing assumptions about pro-poor interventions, peoplecentred targeting and inclusion require critical gender analysis to document biases? The grassroots actors and women collectives were selected to observe and document the barriers.
(3) How does existing frameworks hide the drift in and claims about grassroots actors" and interventions" pro-poor and inclusion "rhetoric"? Evidence for these three inquiries is substantiated from the field and analysed using feminist lenses of community-based asset and social capital theories, dynamic service delivery, continuum change and proactive empowerment theories.
Regarding the existing literature and analysis frameworks, in development policy practice, intervention beneficiaries (poor women in this case) is often targeted as homogenous groups, at micro-level institution (household), needing external support (Sen, 1987;McCall, 2005). In development research, micro-level evidence is seen as context-specific and cannot serve generalization to the impacts of wider interventions (Kabeer, 2016). In terms of intervention arrangements and systems, grassroots institutions are assumed to be "people-centred" with mutually reciprocal social arrangements (Chambers, 2005) and interventions are pro-poor and inclusive (Narang, 2012;Mindaye, 2014).
In terms of development actors, scholar argued that grassroots non-governmental organizations are, inclusive in targeting and "right-fit" to serve the purpose of pro-poor policy objectives (Chambers, 2005;UNDP, 2020). However, feminist critical scholars dispute the aforementioned notions; and call for specific merits of interventions and actors to groups that are systematically marginalized and invisible in existing theory-policy matrices and advocacy messages (Kabeer, 2005;Bradshaw et al., 2019). This study disputes the pro-poor and inclusive notions of grassroots non-governmental organizations and (mutual) reciprocity of women"s grassroots collective associations and movements based on the feminist critical theories; questioned the inclusiveness of the targeting and impact of the interventions on women"s empowerment.
The study further argues that grassroots actors and institutions and beneficiaries such as individuals, associations, households and organizations are not of homogenous. From gender perspective, individual women"s intersectional barriers are embedded in these institutional systems and manifest pervasive discriminations, inequalities and social disparities (Collins and Bilge, 2020). Women"s collective agency empowers individual poor women (Yintiso, 2015;Dessiye, 2014) but can also widen the existing inequalities and reshape the intersectional discriminations (Crenshaw, 1989). A feminist perspective of grassroots institutions, interventions and poor women beneficiary targeting was thus employed as themes of analyses (Daniel, 2021) in this study.
Regarding theories that inform women empowerment and women"s collective agency, feminist scholars identify women"s political, social, economic and collective agency as individual and group measures of women empowerment (Hennink et al., 2012;Kabeer, 2016). Specifically, women"s political empowerment includes women"s ability to engage in decisions affecting their lives as individuals and groups (Brody et al., 2017). Women"s political empowerment result indicators include awareness of rights (laws), actions (voting and being elected), assets/property inheritance and rise to leadership (Brody et al., 2017;AWARD, 2021). Women's political empowerment shapes women"s voice and its inclusion in policies that influence their lives, welfare and wellbeing (Babington, 1999;Yntiso, 2015). Women"s social empowerment refers to women"s freedom of mobility, freedom from violence, negotiation powers, control over their sexuality and reproductive health, choosing spouse, marriage age and access to education and services (Brody et al., 2017). Women"s social empowerment shapes their private and public lives and is diverse (Chant and Sweetman, 2012) and intersectional and inhibits women"s attainment in development and constraints their individual and collective agency (Yintiso, 2015).
Women"s economic empowerment refers to having access to (use) and control over (own) key economic assets such as finance, land, and community assets such as mutual and reciprocal grassroots institutions (Brody et al., 2017;Collins and Bilge, 2020). Women"s economic empowerment indices include (wage) employment, income and materials and non-material assets including leisure (Crenshaw, 1989;Daniel, 2021). Chant and Sweetman (2012), Hennink et al. (2012) and Daniel (2021) argue that grassroots institutions, associations and interventions supporting them create the foundations for political, social and economic empowerment of beneficiaries. By pooling collective agency, the process gives the beneficiaries to evolve above constraints inhibiting their empowerment and challenge the institutional and individual constraints and institutions mediating access to and control over assets, freedoms and rightsas individual woman and group (Babington, 1999). The collectivization of women"s agency enhances beneficiary women to be above power of discrimination and exclusion as beneficiaries (Hennink et al., 2012;Nussbaum, 2003). However, from feminist point of argument, the impact of grassroots intervention needs gender statistics. This study inquired the approaches and tools as well as the mixed results of impact on the beneficiary women.
Regarding the explanatory theories and approaches of women empowerment, four theories from the review of extant literature were selected as lenses of this study; which are asset-based theory (Dessiye, 2014;Keeble, 2006;Mindaye, 2014), social capital theory (Babington, 1999;Muchtar et al., 2019), intersectionality theory (Crenshaw, 1989;Hancock, 2007;Smith, 2014) and empowerment approach (Hennink et al., 2012;UN Women, 2019;AWARD, 2021). These theories differ by discipline and sector; as well as ignorance to feminist views of development research, policy practice and advocacy (Muchtar et al., 2019). For instance, Crenshaw (1989), Hancock (2007) and Smith (2014) argue that invisibility of intersectional gender barriers at households, communities and broader institutions constrain women"s empowerment; and unintentionally widen the existing gaps. Muchtar et al. (2019) and Hennink et al. (2012) argue in favour of reconceptualization of development assumptions and gender gaps. The asset-based theory argues on the need to begin from existing assets and build alternative capacities of beneficiary persons, associations, institutions and proactive transformation towards resilience (Keeble, 2006;Mindaye, 2014). This theory promotes grassroots existing assets including women"s associations as instruments of awareness, need prioritization and institutionalization (Crenshaw, 1989;Yintiso, 2015).
The integrated and composite use of these theories constituted a holistic approach to women"s barriers, discrimination and oppression at specific individual, group, community and broader levels (Collins and Bilge, 2020). Inclusive and transformed grassroots reciprocity and addressing intersectional discriminations improve political, social, economic and institutional statuses of the poor; positively enhance women empowerment (Collins and Bilge, 2020). Though reciprocal institutions build social capital, they also enhance gender disparities.
Women empowerment approach evolved from academic, policy practice and advocacy debates (Mindaye, 2014;Dessiye, 2014;EWLA, 2021) informed by the broader framework of feminist thoughts and women studies in the mid-1980s. This approach emphasizes on women"s intersectional subordination and documents the holistic causes including the historical injustices, social, political, economic, and cultural structures (Chant and Sweetman, 2012). This approach tacitly challenges the established conventional perspectives; incorporates the views that constituted the lens to gender gap indices in the last several decades; shifted the understanding and solutions to address gender inequalities in holistic manner (Narang, 2012).

Economic empowerment indices (N = 3850)my economic status improvement in … Frequency & % (Yes) Frequency & % (No)
Access to collective assets and benefits 2837 ( Hence, this approach tallies with feminist view of development research, policy practice and advocacy as well as includes indices of intersectional, social capital and community-based asset development theories to assess self-help group approach to grassroots inclusive intervention.

METHODS AND MATERIALS
The study adopted a mixed-methods approach and temporal design to measure the impact of interventions by grassroots nongovernmental organizations; which is advised by Creswell (2014) for such themes and levels of analysis. Both mixed qualitative and quantitative approaches was employed to capture data on the indices of empowerment of women, the pro-poor, inclusive and people-centeredness of grassroots organizations and interventions. A descriptive and inferential data and analysis technique was used to interpret the data results and present the findings.
A beneficiary questionnaire was administered for three thousand one hundred programme beneficiaries of consortium of faith-based grassroots organizations across the country, randomly selected from the aggregate national list. The semi-structured interview was held with programme leaders and experts from fifty members of faith-based consortium and the consortium leaders and government policy-makers that work on women affairs at federal, regional and local levels. These methods were used to collect primary data. The consortium targets 548,060 households directly with 2,740,300 indirect beneficiaries in these households. From this total beneficiary population, about 7% (3850 beneficiaries by rule of thumb) were selected randomly.
The secondary data were gathered from policy documents, programme implementation reports and scholarly findings on the theme of women empowerment. Since the beneficiaries constitute programme beneficiaries of the consortium, each population unit is given equal chance of being selected and sample frame was the consortium"s beneficiary list.
The Statistical Package for Social Sciences (SPSS) software 25 was used to summarize both descriptive and inferential statistics. The descriptive statistics employed percentages while the inferential statistic employed a Pearson"s Correlation and multiple (linear) regression results. Using temporal quasi-experimental design (observing before and after the support), the intervention impact on beneficiary women was observed and measured. After the presentation of the impact of the intervention on women beneficiary before and after the support, the study questioned beyond the observed impacts on beneficiaries. The study inquired the established programme intervention assumptions and reason out the widening inequalities and exclusion. It thus inquired programme approach to women empowerment from feminist perspectives of targeting and inclusion as well as the relevance of micro-level research to broader level generalizations.

RESULTS AND DISCUSSION
Women empowerment, in line with grassroots interventions, refer to individual and collective changes to beneficiaries at one hand and continuity and progress of these improvements in institutionalizing gender equality and resilience of individual women and groups in household, community and societal levels (Crenshaw, 1989;Chant and Sweetman, 2012). Women"s individual or collective level empowerment can be measured using political, social and economic empowerment indices (Eyben, 2011;Collins and Bilge, 2020) and grassroots level of collective women"s agency such as self-help groups add a positive boost to the impact of interventions by grassroots actors such as consortium synergies. The descriptive data results and interpretation is presented as follows. From the responses of questionnaires on specific indices of women"s empowerment, the first category of data result was summarized on the indices of economic empowerment of target women after the intervention by the consortium. The sample respondents took seven questions on the specific indices and the result in Table 1 shows that there were considerable improvements in access to collective assets and benefits (73.7%), personal savings and owning bank accounts (60.2%), personal and family incomes (73.6%), diversity in personal and family income sources (73.4%), skills and labour productivity (71.9%) and collective assets and management systems (75.7%). In terms of programme intervention outcomes, the results imply that the majority of the programme intervention beneficiary women considerably improved their economic status in incomes, assets, skills, income source and systems of asset management at one hand and indicates the performance of the intervention by the consortium. However, the intervention has relatively failed in terms of improving beneficiary women"s access to loan and credit finance; which indicates 43.1% performance.  However, in terms of feminist analysis of beneficiary women"s empowerment, the improvement has mixed results. On the one hand, the economic status of beneficiary women has not improved for 26.3% in access to collective assets and benefits, 39.8% in personal savings and owning bank accounts, 56.9% in access to loan and credit, 26.4% in diversifying personal and family income sources, 28.1% in skills and labour productivity and 24.3% in collective assets and management systems. These mixed results confirm the feminist arguments on gender gap assessments, prioritization, inclusion and targeting at beneficiary levels. First, it confirms the ignorance to intersectional vulnerabilities of women (Hancock 2007;Collins and Bilge, 2020). Second, it confirms the categorical targeting of diverse women groups into a single intervention category (McCall, 2005). Third, it confirms the invisibility of women"s strategic and practical economic needs (Crenshaw, 1989) and ignorance to the dynamism of women"s needs in interventions. Therefore gender-blind interventions, that aim at economic empowerment of poor women, further widens the existing economic inequalities of the specific category of women targets at one hand and excludes the nonbeneficiary women category that live in similar contexts. Also, without gender sensitive intervention priorities and targeting, grassroots level actors that were assumed to be pro-poor and socially inclusive are never true and practically widening the gender gaps.
A further analysis was conducted on the impact of the intervention in terms of access and frequency to finance and the amounts of loan and credit received by women beneficiaries.
The mean frequency access to the amount of loan received since the percentage of beneficiaries was low (43.1%) and those who did not access were (58.9%). The result in Table 2 shows that the loan frequency of access by the beneficiaries is 12 times at the maximum and at least 1 time at the minimum with standard deviation and mean points of 1.63 and 2.51 respectively. The amount of loan received by the beneficiary is 21,350.00 birr at the maximum and 200.00 birr at the minimum; with standard deviation of 3.833.00 birr and mean of 3,780.00 birr was found to be the largest loan taken while 240.00 birr was found to be the lowest loan taken with an average loan of 3,780.00 birr. Though the intervention provided loan and credit services, the arrangement implies exclusion of beneficiary women. From feminist point of view and practical evidence on ground, women"s access to finance and supply always manifest short of demand and imply widening further the existing inequalities (Collins and Bilge, 2020).
The second category of data result was summarized on political empowerment of target women before and after the intervention on the basis of five indices. The data results obtained, as presented in Table 3, shows the  figures on the indices before the intervention. In this respect, before the intervention, respondents" decision power is low; for instance, decision of wife over family expense (18.4%), family health care (22.9%), children education (23.5%), personal savings and loan (12.2%) and involving in committee (20.2%). However, decisions made by spouse on these issues are considerably high; for instance, the spouse decision over family expense (57.4%), family health care (62.5%), children education (59.8%), personal savings and loan (65.7%) and involving in committee (75.4%). On these issues, decisions made by spouse and wife over family expense (24.3%), family health care (14.6%), children education (17.2%), personal savings and loan (22.1%) and involving in committee (4.3%).

Measurement indices (N = 3850) … Before support (Frequency and %) Frequency and % (Yes) Frequency and % (No)
The data results obtained, as presented in Table 3, shows the figures on the indices after the intervention. In this respect, after the intervention, respondents" decision power has improved; for instance, decision of wife over family expense (57.0%), family health care (54.0%), children education (54.5%), personal savings and loan (62.1%) and involving in committee (56.6%). However, decisions made by spouse on these issues are still significant; for instance, the spouse decision over family expense (20.7%), family health care (20.9%), children education (19.6%), personal savings and loan (14.1%) and involving in committee (18.7%). On these issues, decisions made is, for instance, by both spouse and wife over family expense (22.3%), family health care (25.1%), children education (25.9%), personal savings and loan (23.9%) and involving in committee (24.7%). The results, comparing Table 3 above, show that for considerable number women beneficiaries, still their spouses decide over the behalf of the wife or both decide on issues concerning the wife"s personal choices. Thus, the results of improvement are mixed; confirming the feminist points of argument that beneficiaries tacitly cement power differences in households and further deny beneficiary women from their rights and benefits due to targeting errors (Hennink et al., 2012;Collins and Bilge, 2020).
The third category of data result was summarized on social empowerment of target women before and after the intervention. The respondent"s status of social empowerment was assessed on three social empowerment measurement indices (questions). The results of social empowerment indices obtained from the questionnaire response are presented in Table 4. In terms of respondent"s status before the intervention, for instance, the figures in Table 4 shows that beneficiaries that participate in social gathering (7.6%), frequent participants (7.9%) and those beneficiaries that are facing discrimination and ill-treatment (66.2%). However, the majority of women do not have low rating in social empowerment indices; for instance, beneficiaries that did not participate in social gathering (92.4%) and nonfrequenters (92.1%).
In terms of respondent"s status after the intervention, for instance, the figures in Table 4 shows that beneficiaries that participate in social gathering (72.8%), frequent participants (72.2%) and those beneficiaries facing discrimination and ill-treatment (72.8%). Also, considerable number of women have low social empowerment status; for instance, beneficiaries that did not participate in social gathering (27.2%), non-frequent participants (27.8%) and those beneficiaries that facing discrimination and ill-treatment are still (27.2%).
The results demonstrate the feminist point of argument that as husband"s feel their wives are empowered, they become threatened of losing their power and new forms of violence and conflict emerge at household level at one hand; and constrain their wives from exercising their rights, choices and benefits (Crenshaw, 1989). The concluding remark here is that beneficiary targeting, follow up and maintaining continuity of improvement in women"s social empowerment demands policy and legal support as well as revisiting existing institutional arrangements and systems (Collins and Bilge, 2020).
The fourth category of data result was summarized on collective agency (empowerment) of target women as individuals and groups due to the intervention support.  The improvements from the collectivization of the agency of women was measured based on self-help group as proxy measure and the programme support as intervention synergy towards holistic results. The collective agency (measurement) was observed using five indices that indicate synergy and results of the programme support to beneficiary women who are organized into self-help groups. The results on the observations of each of the indices are presented in Table 5. The respondent"s status of empowerment from collective agency and programme support"s synergy indices shows that there are improvements. For instance, being in the group and benefiting from the programme support, the beneficiary women, for instance, improved engagement in peer-education, access to information and enhanced awareness (71.1%), change in life-choices based on peer-learning (65.5%), change in gender roles in the community (33.0%), no change in confidence and self-esteem (72.5% and achieved expected life-choices (72.9%).

Measurement indices on programme support (N = 3850) -improved… Frequency and % (Yes) Frequency and % (No)
However, from the data results in Table 5, considerable number of women beneficiaries did not improve in these indices of collective agency and synergy from intervention support results; for instance, no engagement in peer-education, access to information and enhanced awareness (28.9%), no change in life-choices based on peer-learning (34.5%), no change in gender roles in the community (67.0%), confidence and self-esteem (27.5%) and no achievements in expected and wished-for livechoices (27.1%). From the feminist points of argument, interventions that focus on collective agency of women can improve women"s collective and individual empowerment status; but, if intersectional and invisible aspects of difference among women are not identified and inclusive institutional and targeting arrangements are not put in place, intervention can widen existing gaps within women groups and institutionalize inequalities among women groups and discriminatory systems (Hancock, 2007;Hennink et al., 2012;Muchtar et al., 2019, Collins andBilge, 2020). Therefore, dynamic and continuous assessment of positive and negative results should be made to proactively address emerging gaps and scaling up grassroots best practices.
The inferential data results were also interpreted subsequently. In order to further substantiate the evidence and support the findings of descriptive data results, an inferential statistical summary was produced; and a step-by-step analysis was made. Also, further inquiry was made using different methods of data summary and analysis. In this respect, first, a correlation analysis was conducted on the four dimensions of women"s empowerment; as presented in Table 6. The correlational result, in Table 6 shows that the explanatory variables positively and significantly relate with the dependent variable. In specific extent and direction of association among the explanatory and dependent variables, the Pearson correlation coefficients are; for instance, 0.811 for economic empowerment, 0.631 for social empowerment and 0.322 for political empowerment. From the correlation results, political empowerment status is low as compared to other empowerment dimension; in terms of explaining the impact of the intervention. This result confirms the feminist point of argument that although improvements are observed in social (such as education and health) and economic (such as incomes) for women, political and institutional discriminatory systems remained intact and the exclusionary systems of power continue reducing the gains of the past decades at one hand and constrain the continuity of progress of women"s rights, freedoms and agency (at collective and individual levels). In order to further substantiate the correlation results, a multiple regression was conducted and the results were obtained as follows in Table 7. The model summary results of the predictor variables and their correlation coefficient with the dependent variable (programme) is 0.811 (81%). This result indicates the extent of correlation between the explained and the explanatory variables in general. Also, the combined result of the explanatory variables is 82%; since the R square value is 0.821. The adjusted R-Square also manifest 82% with standard deviation of 0.16 and Durbin Watson coefficient of 1.352 points. The evidence above is further substantiated using ANOVA tests.
The ANOVA Table 7 indicates that the regression model significantly predicted the dependent variable based on the explanatory variable identified; with the F statistics of 501.005 and P-value less than 0.01 the summary of coefficients, presented in Table 7 also shows  the significant association between the explanatory and the prediction variables. Based on the coefficient table the following regression formula (model) is constructed in an effort to predict: self-help group Programme support fruits on woman empowerment. Economic, social and political empowerment indices identified were found to be robust in predicting the impact of the explanatory variables over the dependent variable (with P-value < 0.05).
Finally, the inferential statistics results show the positive impacts of the intervention on social, economic and political empowerment of beneficiary women; but the figures, as in the case of mainstream debates on gender statistics, the categorical, intersectional and objectivistsubjectivist dilemma, the status of the empowerment of those who could not benefit from the intervention were galvanized by improvement evidence supported by percentages, correlation, regression, ANOVA tests and model summary including coefficients. These findings therefore alarm the longstanding feminist view that development research theories, methods, policy practice systems and advocacy need revisiting and should be critically questioned; specifically, from the perspectives of intersectionality, pro-poor policies and inclusive targeting "rhetoric".
After the discussion of findings (both from descriptive and inferential data results), questions for further research are identified. In the broader development process, gender perspectives, gaps and demand for innovative, flexible and dynamically continuous frameworks of targeting have become highly relevant; but gender insensitivity remained pervasive at expert, institutional and systemic levels (Babington, 1999). As gender perspectives evolved from interface of theory, policy practice and advocacy over the last sixty years, issues such as awareness over the gender gaps in crisis and normal settings, demand for gender-responsive outreach and user-friendliness of interventions and targeting has become critical demands of liberating the development context, process, outcomes and institutions from gender biases (Muchtar et al., 2019).
From the inquiry and the evidence substantiated from the field, the findings of this study shows that dynamically context-fitting interventions, gender statistics and information and tailored response to the barriers of gender equality, women empowerment and inclusive development targeting is essential; and therefore, mainstream assumptions on institutions, targeting frameworks, pro-poor policy indices and "inclusive" actors need revisiting and reconceptualization; informed by feminist arguments of gender-based exclusion (Kurgat and Ombui 2013;Eyben, 2011). To inform the new inquiry on intervention targeting, a holistic empowerment dimension of gender gaps, intervention processes and outcomes is required (AWARD, 2021). Though the lack of holistic approach plagued interventions on filling the pervasive gender gaps, women empowerment approach, which evolved from development interventions and policy practices over the last sixty decades, is serve as a starter and this study used this lens in addition to dynamic service delivery, social capital, community-based asset and change continuum theories (Kurgat and Ombui, 2013;Eyben, 2011;Nussbaum, 2003;Kabeer, 2016). The dynamic intervention and response theory focuses on assessing context and fixing interventions to dynamic and flexible needs of beneficiaries (Kurgat and Ombui, 2013). The intervention changes and continuum theory focus on fitting intervention goals to changing demands of beneficiaries, actors and continuity towards resilience (Narang, 2012;Muchtar et al., 2019). The women empowerment continuum approach focuses on shaping and reshaping conditions for continuous individual, group and systemic level outcomes; leading to building transformative capacities of women as individuals and groups as well as systems as organizations and networks of partners towards a synergy (Nussbaum, 2003;Kabeer, 2016).
On the basis of the above reflections on gender equality, women empowerment and transformative development at individuals, groups and systems levels, future gender research needs politicization of inequalities (Hancook, 2007;McCall, 2005;Chant, 2007;Boserup, 1970), liberating gender bias in research, interventions, documentation and lessons (McCall, 2005;Hennink et al., 2012), establishing continuous and progressive feminist argument against feminization of gender gaps, statistics and interventions (Chant, 2007;Collins and Bilge, 2020), leading by scientific argumentation on the arguments in literature, policy practice and advocacy against and in favour of universality of gender gaps (Crenshaw, 1989;Chant and Sweetman, 2012) and at the end, quantifying and qualifying, in fine balance, gender gaps, information and the changing theories, practice and advocacy at micro, meso and macro aspects of women empowerment in the development process and institutional systemic levels (Boserup, 1970;Collins and Bilge, 2020).

Conclusion
Along with the above concluding remarks and foundations for initiating new forms of enquiry, practice and advocacy, the feminist points of argument inform the following.; which are: (i) gender gap has remained enduring problem despite global feminist advocacy, policy commitments and legal commitments and movements demanding inquiry (Bradshaw et al., 2019;Kabeer, 2015;Folbre, 2006); (ii) gender gaps are similarly a global concern as well as a particular form of policy attention than the usual rhetoric, inquire its wider, specific and complexity frameworks (International Finance Corporation (IFC) 2014); (iii) broader and specific measures and gender statistics on gaps and barriers are still far from complete; (iv) discipline diversity and approach to gender gaps in theory, (multiple) feminist movements in policy and advocacy arena need to be ignited again as in 1960s and 1970s; that transformed development research, policy practice and advocacy; and, (v) gender statistics at micro-meso and macro levels should be generated and inform succinctly and serve as a basis of feminist argumentation in theory, influence in policy practice and shape advocacy messages; focusing on why gender gaps remain pervasive, widening and complex; while "rhetoric" remain floating at policy level.
Thus, the "rhetoric" pro-poor policies, pro-poor actors, inclusive institutions, impact of interventions and benefits to target beneficiaries need to be questioned; because from the fact that: (a) interventions in the development process, whether at micro, meso or macro-levels widen existing gender gaps; (b) the mainstream inclusion and targeting frameworks shape and reshape intersectional vulnerabilities and inequalities of women as individuals and groups; and (c) institute further institutions and systems of exclusion to further complicate the gender barriers. Therefore, based on feminist thoughts so far existing theories, policy frameworks, advocacy messages and movements need to be questioned.
The re-enquiry and revisiting suggested by the study is not only based on the findings from the field but also from the existing presentation of feminist critique as follows: (1) institutions and intervention mechanisms remained gender-biased and become pervasive challenges for creating enabling environment for women in the development process (Amarech, 2019); (2) gender perspectives alert on widening gaps and at recent times, related with diverse crisis contexts in the development process, the gains of the past decades is waning and research evidence, policy responses and advocacy messages are not exhaustive and holistic (Amarech, 2019;Cornwal and Rivas, 2015;Berii, 2019); (3) gender equality and women empowerment measurements, sectoral level indices and statistics if far from complete and fragmented on the basis of context, discipline and political economy of institutions; therefore, demands extensive research, evidence and policy information (Dijkstra and Hanmer, 2000;Muchtar et al., 2019;UN Woman, 2019); and, (4) gender, norms, benefits and care economy need more research, evidence and advocacy message to influence academic, intervention and advocacy actors (Folbre, 2006;Yntiso, 2015).

ACKNOWLEDGMENTS
The author would like to thank all respondents that participated in this research and also appreciate the agriculture offices and the experts that supported in providing materials (documents) during the study.