Isabel Lambrecht is an Associate Research Fellow at IFPRI’s office in Ghana and a former visiting scholar at Cornell’s Dyson School.
As I embarked on a PhD, gender was the topic my advisor wanted me to include in my research. I was less keen to work on it. By then, gender seemed the preferred focus of so many projects and researchers. What could be left for a researcher that was trying to contribute something new?
Fortunately, a lot.
My research focus was on gender and agricultural technology adoption in South-Kivu (Eastern DRC), specifically related to a research-for-development project called the Consortium for Improving Agricultural-based Livelihoods in Central-Africa (CIALCA) aimed at introducing integrated soil fertility management (ISFM) practices. The project did not specifically target women, but it did work with both male and female farmers. Among others, I tried to answer the question: do we get higher uptake of agricultural technologies if agricultural extension programs work with only female farmers from a household, only male farmers, or both spouses jointly?
Before conducting my field research, I had meticulously explored the general literature on gender and agricultural technology adoption in Sub-Saharan Africa. Scientific information on the specific case study region was scarce, as recurring conflict in the past decades had inhibited researchers to visit the area. Most recent information came from reports from humanitarian organizations, painting a grim picture of extreme poverty and severe violence towards women. I started the field research with group discussions and stakeholder interviews. In a second stage, I gathered my own quantitative gender-disaggregated household data. Finally, I validated several hypotheses concerning gender and technology adoption using state-of-the-art econometric methods.
During each stage of the research, I encountered small and large violations of my expectations of gender in agriculture. For example, most well-known articles on gender differences in technology adoption base their analysis on comparisons between male- and female-specific plots, crops, or tasks. However, households in South-Kivu do not systematically differentiate on gender dimensions with respect to these categories. These insights required adaptation of the research hypotheses and strategies.
There are different reasons to suspect that adoption rates in farm households are either higher or lower if women participate in an extension program than they would be if only men participate. On the one hand, female farmers are significantly less aware of new agricultural technologies despite being major contributors to on-farm labor and plot management decision making. On the other hand, women generally have less access to credit or agricultural (complementary) inputs required for technology adoption. Women in South-Kivu rarely own land, especially in male-headed households. They are not allowed to take financial decisions without consultation with their spouse, earn significantly less off-farm income, and are less likely to borrow money compared to men. Could agricultural extension overcome these gaps?
My econometric analysis, available in the European Review of Agricultural Economics, resulted in three main findings.
- Joint participation in agricultural extension by male and female farmers from the same household led to the highest technology adoption rates.
- Female program participation has a larger impact on technology adoption in female headed households than male headed households.
- Female program participation is not conducive to the adoption of capital-intensive technologies, such as mineral fertilizer, while it is for technologies that increase the labor-intensity of specific female activities, such as row planting, or specific female crops, such as legume varieties.
The implications of these findings warrant rethinking some of the current mainstream gender development strategies. For example, despite the general perception that addressing women is a successful strategy in development projects, our conclusions show that in some cases female program participation may conflict with program objectives. Extension programs could be more successful if they engage female farmers together with, and not instead of, their husbands. Moreover, women should still be considered a diverse group. In our study we found that targeting female-headed households, rather than female farmers in general, could be a valid strategy. Finally, extension participation by itself may not be enough for female farmers. Successful extension participation by female farmers requires addressing additional constraints to access to productive resources.
This research project also taught me about gender research more generally. It became clear that there are many opportunities to do more and better in gender research. Two things stand out to me.
First, we have not characterized the wide diversity of gender issues across societies. Are the farmers in South-Kivu really so different, or is the current sample of case studies on gender not representative for Sub-Saharan Africa? Could there be self-selection in research (publications) of specific cases that lend themselves better for the analysis of gender in agriculture than others?
Second, there exists an unsubstantiated perception that working with women is a best-bet strategy for development. To be able to claim that female participation leads to better outcomes than does male participation, a thorough analysis must be done that actually compares the impact of female versus male program participation on the anticipated project objectives. Only a handful of studies do so (e.g. Pitt and Khandker 1998, Maertens and Verhofstadt 2013). Several studies look at projects specifically targeted at women and show that these projects have led to beneficial outcomes. The counterfactual, however, is not male participation, but no project at all. In addition, these studies are mostly done in situations where women’s access to cash increases, which is a specific setting that is likely very different from other interventions like agricultural extension.
While I was reluctant to work on gender initially, it has become my preferred research topic. Some case study settings, research questions, or even the interpretation of results can be more complex than others to analyze in a rigorous way that is accepted by economists, and importantly so, in high-ranked economics journals. And the research findings may be less straightforward than we had hoped to convey in a simple message. But for those keen on understanding and explaining, why not just go for another challenge?