IFPRI Gender Methods Seminar, May 28, 2015: Women's Empowerment in Agriculture: Implications for Agricultural Productivity in Rural Bangladesh

Editor's Notes

• #3: Begin with background on women and agriculture in rural Bangladesh. Despite steady growth in agricultural production over the past 40 years, large portions of the population in Bangladesh, particularly women and children, continue to struggle against chronic food insecurity. How to address this? Hoddinott and Haddad (1995): Increasing women’s share of cash income significantly increases the share of the household budget allocated to food. Indeed, this is a key part of what it means for women in rural Bangladesh to be “empowered.” To paraphrase Naila Kabeer’s definition of empowerment: Greater engagement in agriculture increases women’s ability to make strategic life choices in a context where that ability was previously denied.
• #4: Significant obstacles exist to women’s empowerment in Bangladesh: Traditional gender division of labor: women spend the vast majority of their time in domestic work (e.g., preparing meals for her family, caring for her children, cleaning the family’s dwelling, etc.) and men spend most of their time in agricultural labor outside of the home. This means that women’s agricultural labor contributions tend to be limited to tasks that can be accomplished within or near their homesteads (e.g., post-harvest activities such as drying, sorting, and packaging of crops). Women in Bangladesh, particularly those from poorer households, are increasingly willing to take up socially “unacceptable” work, things like engaging in agricultural wage labor alongside men.
• #5: Specifically, I look at gender gaps in terms of women’s role in the management of agricultural plots of land operated by their households. I define plots as either female-inclusive or male-exclusive, based on women’s participation in decision-making and ownership for a plot of land.
• #6: Specifically, I look at gender gaps in terms of women’s role in the management of agricultural plots of land operated by their households. I define plots as either female-inclusive or male-exclusive, based on women’s participation in decision-making and ownership for a plot of land.
• #7: I operationalize women’s empowerment in two ways: Due to some econometric concerns: primarily the possibility that the empowerment score may be endogenous with production. I also operationalize women’s empowerment in terms of one of the indicators that make up the WEAI that is less likely to be endogenous: group membership. Specifically, membership in credit, microfinance, and informal savings group.
• #8: I operationalize women’s empowerment in two ways: Due to some econometric concerns: primarily the possibility that the empowerment score may be endogenous with production. I also operationalize women’s empowerment in terms of one of the indicators that make up the WEAI that is less likely to be endogenous: group membership. Specifically, membership in credit, microfinance, and informal savings group. I’ll talk about why I choose all of these indicators and provide greater details on their definitions later on in the presentation.
• #9: So why might we expect women’s empowerment to be associated with increased agricultural productivity? There are a number of channels to consider. Social capital: social relationships that women may develop from belonging to a group may provide them with a means of gaining access to information about new technologies and farming practices that they might otherwise be excluded from. In times of hardship, social relationship may be accessed to acquire needed resources or other assistance. Access to credit: increased access to credit allows for a greater ability to invest in infrastructure---improvement on plots, such as the building of mechanized irrigation---and offers protection against consumption and production shocks.
• #10: I frame the analysis of my questions in terms of gender differences in agricultural productivity. I demonstrate what that means in this slide. There are a number of different ways to measure agricultural productivity… … I’ve populated the space below the production frontier with observations for a hypothetical sample of exclusively male-managed farms and female-inclusive farms. Farms A, B, and C produce very different levels of output, while using the same level of input. This makes it very easy to compare the farms in terms of technical efficiency: we simply look at how far each is below the production frontier. Clearly, Farm A is more efficient than Farm B which is more efficient than Farm C. When we talk about gender productivity gaps, we’re referring to the gaps between Farms A and B and between Farms A and C. In practice, this is the sort of distribution often seen in less developed countries. Female-managed farms achieving lower levels of technical efficiency than male-managed farms, often because of a reliance on traditional technologies, due to women’s limited access to physical and human capital.
• #11: Focus on Bangladesh; whereas the literature tends to be dominated by studies on sub-Saharan Africa. Joint cultivation is the norm in South Asia (and Bangladesh): men and women tend to specialize in different agricultural activities according to traditional gender norms---makes it more difficult to attribute managerial responsibility to a single household member. This is why in my study I tend to focus on women’s joint participation in plot management Many studies also fail to properly deal with the problem of simultaneity of productivity and input choice. Since both the quantities of output and variable inputs are simultaneously determined by the conditions of profit maximization, single-equation estimation of production functions can produce inconsistent parameter estimates unless proper steps are taken. For this reason, in my essay I use stochastic frontier analysis, which solves the simultaneity problem through specific distributional assumptions.
• #12: A production frontier is estimated based on the levels of output and input associated with each plot within the sample. This gives us an estimate of the level of technical inefficiency achieved on each plot. (How far it falls below the production frontier.) The second part of the model involves the estimation of a technical inefficiency function. Using the information on technical inefficiency obtained from the production frontier, technical inefficiency is modeled as a “function” of a set of exogenous variables. This generates parameter estimates which allow us to evaluate the relationship of each variable to technical inefficiency. The results that I’ll discuss later come from this portion of the model.
• #13: For those of you familiar with the BIHS, I use the entire sample (including the over-sampling in the FTF zone-of-influence) to maximize my sample size.
• #14: Due to econometric concerns about the possibility of the empowerment score being endogenous (which I won’t go into now, but I’m more than happy to come back to later)… Includes decisions about the type of crop to plant, the use of inputs, the marketing of the crop, and the spending of revenue generated by the plot. Women much more likely to be included in decision-making than in ownership (though both remain very rare among plots in Bangladesh) I also include a standard set of control variables based on other stochastic frontier analyses conducted on Bangladesh.
• #15: This isn’t quite as depressing as it sounds. Although it does imply that there may not be any benefit (in terms of technical efficiency) associated with women’s inclusion in plot management, that’s not the same as saying there’s no benefit at all to women’s inclusion. Indeed, as we talked about at the beginning of the presentation, simply being more engaged in agriculture is empowering for women in rural Bangladesh and may improves women’s resource position within the household.
• #16: Do these relationships depend on the primary female decision-maker being directly involved in plot management (in terms of female-inclusive ownership and decision-making)? In other words, are the effects of the empowerment score and group membership (on technical efficiency) conditioned by female-inclusive ownership and decision-making? To answer this question…I re-estimate both models and include a full spectrum of interaction effects (between empowerment indicators and plot management indicators). KEY FINDING: Women’s empowerment benefits not only women but also other member of their households. Positive spillover effects: if we think about group membership, information gained from being a member of a group may be shared among household members or credit obtained through a group might be used to fund household consumption and production.
• #17: The accumulation of social capital may depend on actively participating, rather than just belonging to a group---a woman who only rarely attends meetings or tends not to engage in group discussions or activities may not develop the same social relationships as a more active participant. In all cases these again refer to credit, microfinance, or informal savings group Compared to the 26% of women in the sample who belong to a credit, microfinance, or informal savings group… Attendance also appears low---roughly half of women have attended only one of or have attended none of the last five meetings.
• #18: Thus, it may be that the relationship we observe between group membership and technical efficiency is, in fact, driven by the relationship of group leadership and technical efficiency. Though this is a difficult assertion to make given the very small number of women in leadership positions. Larger implication makes this an interesting subject for future research…
• #19: Also important to be aware that women’s contributions to agricultural production may not always be immediately visible and that to find them you may need to look beyond indicators of household headship or even beyond plot ownership and decision-making. These indicators only told part of the story. It was only when I brought women’s empowerment into the picture, that women’s real contributions became evident. KEY FINDING: My results suggest that women’s empowerment may benefit not only women but also other member of their households (at least in terms of agricultural productivity).