Scaling up Women's Agripreneurship Through Public-private Linkages in Rural Ghana

This project includes quantitative cross-sectional surveys, qualitative data collection, and a quasi-experimental behavioural trial. The description of the study is shown below by specific objective.

Specific objective 1. Identify (i) individual, (ii) organizational, and (iii) social/community/institutional factors that facilitate or impede women's participation in farming entrepreneur associations, and that can be modified through intervention.

Quantitative hypothesis Women's enrolment and active participation in women farmer entrepreneur associations is associated with individual characteristics (e.g., self-confidence) as well as organizational (e.g., having a functional association) and social/community/institutional factors (e.g., having community leadership support to join) that may be modifiable through interventions.

Qualitative research questions (i) What gender inequalities exist within the context of the local organizations (e.g., meetings held in locations and times that are inconvenient for women) and community/institutions (e.g., norms that prevent women from doing business in the community) that have implications for the entrepreneurial activities and well-being of women farmers? (ii) What organizational and community/institutional factors contributing to gender inequalities are modifiable through interventions to improve women's entrepreneurial activities and well-being?

Activity 1.1. Cross-sectional survey in contiguous districts. Data collection will be carried out in the two districts contiguous to Upper Manya Krobo District (UMKD) to collect data among members of four associations identified (20-30 women/group, total of approximately 100 women) as well as from their matched non-member neighbours (approximately 100 women of similar non-modifiable characteristics).

Specific questions on facilitators and barriers about joining a farmers' association and specific questions about perceived gender differences in access to resources and services will be included. The results will identify modifiable factors that can then influence how district-level service providers interact with women farmers.

Analysis of data. The analysis will test for the cluster effect of the association. Unadjusted bivariate analyses will test the relationship between the outcome and possible covariates using independent Student's t-test for continuous variables and Pearson's Goodness-of-fit chi-squared for categorical variables and those with a p\<0.20 will be included initially in the model. The investigators will use a logistic regression model with standard errors corrected for clustering, using a backward elimination stepwise variable selection procedure. Variables with a p-value of \< 0.10 will remain in the model.

Activity 1.2. Focus group discussions and in-depth interviews. Qualitative methods will be used to examine gender inequalities at the institutional level. Of interest is the programs and services that institutions provide and their influence on the lives and work of women agripreneurs as well as institutional staff. The number of interviews and focus groups will depend on reaching saturation; however, the investigators expect to conduct at least 6 in-depth interviews with institutional staff (both men and women) in each of the three districts. The staff will be key informants selected based on their role or personal knowledge about the existing associations and institutional responsibilities. The women will be selected from the association membership and survey lists to represent different levels of entrepreneurship. The focus group discussion and interviews will focus on (1) identifying gender gaps in services and access to resources; (2) analyzing existing programs and how they address the different needs of women and men; and (3) discussing policies and resource allocation decisions and whether they support gender equality. The study interest is in the perception of the staff on how their role with the programs and services for women is assessed and valued within the institution and the barriers they face in supporting women.

The interviews will be followed up with four focus groups (\~10 /group) with women farmer entrepreneurs (two in UMKD and two in the two surrounding districts; total = 40 women) to examine the gender institutional inequalities from the viewpoint of women agripreneurs. In addition, women's perception of the role of their own empowerment (how they define empowerment, understand its importance for their participation and entrepreneurial success, view limitations to their or others' empowerment) and how this perception is linked to their interactions with the institutions is of interest.

Analysis of data. The transcripts (about 28 hours of data) will be translated, coded and analyzed using a content analysis approach with a priori and emerging themes.

Specific Objective 2. Develop approaches that enhance both the improved outcomes of women agripreneurs (knowledge, skills, and empowerment; entrepreneurial success; nutrition/health) as well as improved outcomes for the institutional stakeholders (program participation/clientele reached, effective/efficient service provided, staff knowledge and skills, job satisfaction).

Activity 2.1. Focus group discussions and in-depth interviews. Qualitative methods will be used to discuss with women and staff the value of incentives and their importance for participation in associations. The selection of participants and the number of interviews and focus groups have been described above in Study 1.2. The discussion for these questions with both women and staff will focus on different types of incentives, their value to women, and their value compared to the cost of participation at association meetings. With only staff, the investigators will discuss different types of incentives that encourage staff to support meetings.

Analysis of data. The analysis will be carried out as described under Study 1.2.

Activity 2.2. The project will develop approaches to support the training of women farmers and facilitate continued education training of district staff to strengthen associations and to enhance staff knowledge and skills to address local challenges that members face. The investigators will carry out a formative assessment with interviews with staff and administration on local training needs.

The investigators expect that some of the approaches may include activities to address the following:

(1) How can the district staff be incentivized to work with the women's association and create demand for their services?; (2) How can the institution empower private sector value-chain entrepreneurs to provide high quality information to women agripreneurs?; (4) How can the bank better reach women agripreneurs?

Data collection will focus highly on the process evaluation to document what happens in each tested approach, perceptions of what works/does not work, and the perception of sustainability of the specific activity. The method of evaluation will depend on the specific approach. Whenever possible, indicators that are regularly collected by the institution will be used to assess the results (e.g., number of women agripreneurs assisted by the district veterinarian). Both quantitative and qualitative data will be collected, as is appropriate. The specific approaches will be determined after the cross-sectional survey is completed.

Activity 2.3. Two-arm quasi-experimental trial. The focus is to test an intervention that is feasible within the context of permanent institutions, thus the package must be self-sustaining. The primary outcomes of interest are (i) participation in the association, (ii) small business economic outcomes, and (iii) women's well-being. The study includes women who are members of women farmer associations in the three designated districts. The intervention will use Heifer's Passing-on-the-Gift approach that could be replicated in the future without project intervention. Women's groups will be randomly assigned to treatment (intervention, control) and to a group cycle (1, 2). The weekly project activities (starting with Group 1) are repeated over 4 months. At the end of the cycle of 4 months, Group 2 initiates the intervention, using resources from the first group's "passing on the gift®" for the financial base. The two groups alternate cycles so that over a 16-month period, group 1 and group 2 both receive the intervention package for 8 months.

In addition to the quantitative analysis, a case study approach will be used to examine women agripreneurs and the role that the association had in their success. The first goal is to develop a consensus on what is "success". This will be obtained through focus group discussions, examining the question: "What would one expect to see in the life of a woman who is doing well in her poultry business?" The investigators will direct the discussion to include women's empowerment, financial stability, and maternal and child well-being. Based on the criteria developed on "success", as well as the project's own criteria, the investigators will identify among the association members the ten most successful and the ten "less successful" entrepreneurs. The researchers will carry out in-depth interviews with the 20 women on life histories, experience with the association, and perceived successes, failures, and obstacles. Additional interviews will be carried out with two household members and three business colleagues per case to examine further the woman's experiences to have a complete picture. Direct observation (\~ 2 hr/visit) of each case at home and at her business location will be carried out to have a full picture of the woman's activities, environment, and social interactions in each site.

Finally, there is interest in understanding the pathways by which the outcomes of interest are attained for women participating in associations. The project includes repeated surveys (baseline, 8 months, 16 months) to analyze changes that occur with participation in associations and allow us to estimate the means by which the associations influence lives.

Analysis of data. The outcomes of interest from the trial will all be continuous. The data will come from a cluster-assigned, quasi-experimental trial. Unadjusted bivariate analyses will test the relationship between the outcomes and possible covariates using independent Student's t-test for continuous variables and Pearson's Goodness-of-fit chi-squared for categorical variables and those with a p\<0.20 will be included initially in the model. The analysis will use a multiple linear regression model with standard errors corrected for clustering, using a backward elimination stepwise variable selection. Variables with a p-value of \< 0.10 will remain in the model.

The analysis of the case studies will follow the process used by the team previously. The focus groups and interviews will be translated and transcribed. There will be a systematic review of the focus groups (and confirmed with the individual interviews) to identify individual characteristics and factors that are associated with success. The next step is to generate theoretical explanations of how success is achieved or not achieved.

Content analysis of transcripts and observations will be employed to generate themes around experiences that will be grouped into generalizations about successful women. While the focus is on 'successful' women, data from the 'less successful' women will be used as a counterfactual to cross-examine the data. The case studies will provide a rich understanding of the factors that are needed to support women at different times of developing their small businesses as well as the consequences for the woman and the family.

The results from the repeated surveys on pathways will use repeated measures analysis. The approach will be similar to that described above for the randomized trial outcome analysis.