Speaker
Description
Introduction
Ensuring food security and nutrition remains a primary concern for many countries in the Global South, while frequent climatic (and non-climatic) shocks are creating a serious challenge for sustaining current levels of food security. Agroforestry has been praised for its potential as an adaptation strategy to climate change, in particular for addressing deforestation, biodiversity loss, and environmental degradation, but also for its potential to improve wellbeing and nutrition. Agroforestry can affect nutrition directly through the provision of food (products from trees, shrubs and crops) but also indirectly, by improving yield (through fertility-enhancing synergies), resilience (shade, reduced erosion, reduced pests), and income (through the sale of food products but also timber and medicinal plants), in turn affecting food security and nutrition. Yet, rigorous studies on the link between agroforestry and household nutrition are lacking, and little is known about the role of gender norms in shaping this relationship. On one hand, women are responsible for preparing food, on the other hand, they lack control on the planting choices and on the use of the harvest and of income, ultimately affecting the sign of the relation between agroforestry and nutrition.
To bridge this gap, we focus on Madagascar, whose rural population is largely food insecure (65% prevalence of moderate or severe food insecurity in 2021, World Bank (2022), while its biodiversity-rich environmental resources are subject to quick depletion. Moreover, the frequency of cyclones and droughts urge farmers to adapt their agriculture. Agroforestry is not new to Madagascar, as farmers have traditionally cultivated food and cash crop perennials in agroforestry systems (tsabo). Lately, agroecological measures such as agroforestry have been widely promoted in the country. Moreover, Madagascar is an interesting case study for its traditional division of gender roles. For instance, women in Madagascar can own and inherit land, however, compared to men, they are less engaged with cash crop production, have lower access to extension services, and are less likely to own titled land (Widman & Hart, 2019). The separation of roles is such that men are responsible for the preparation of soil and cattle, while women take care of the weeding, transplanting, household chores and care of the children (Jarosz, 1997).
Objective
In this paper, we aim to bridge this gap by analysing the causal link between agroforestry and food security and nutrition of the household, while exploring the mediating impact of the gender of the household head and the decision maker for agricultural decisions. To this end, we use novel representative data from three regions in southern Madagascar (Anosy, Androy and Atsimo Atsinanana) collected by the Potsdam Institute for Climate Impact Research from 2023. The survey contains information on farm production, agricultural practices and food security and nutrition, amongst others. In particular, we focus on the direct impact of having trees on the parcel on households’ nutrition and food insecurity.
Methodology
Methodologically, combining trees with other crops is endogenous to household choices, which in turn affect nutrition. Therefore, an Instrumental Variables approach is recommended. As instrument, we use the previous ownership of land by the mother of the respondent and wind speed during cyclones. As trees are a long-term investment, the adoption decision strongly depends on the (perceived) tenure security of land and on a long enough presence on the farmland by the household's family. Data on wind speed comes from ISIMIP cyclone events maximum wind speed. Specifically, we count the number of times during cyclones in which the maximum wind speed exceeded a certain threshold. We construct these variables for the period 2008-2018, excluding explicitly the last 5 years’ events which might still have repercussions on the nutrition and food security at time 0. The assumption here is that strong winds in the past affected perennials in the past but not current food crops (hence nutrition). The exclusion restriction would be violated if alternative pathways exist from mother owning land or strong winds to nutrition, that do not run through agroforestry. To limit these concerns and those related to the use of climate IV (Mellon, 2023), we conduct several robustness tests.
Findings
Preliminary results show that the effect of agroforestry on nutrition and food security outcomes is mostly not significant or is negative. This is consistent in the OLS and IV specifications. Moreover, the role played by the gender of the decision maker partially makes this relation more explicit, signalling some important differences.
The two instruments are not weak according to the test of Montiel-Pflueger (2013), and Kleibergen-Paap (2006). The Anderson-Rubin indicates a well-specified model (cannot reject the joint hypothesis) in the cases of mild or no food insecurity, no hunger, frequent protein intake, food expenditure (only first wave), and diversity of food produced.
Conclusions
Climate change urges effective solutions in adaptation for agricultural systems in the Global South, which are dominated by smallholder agriculture. Agroforestry presents a multi-purpose adaptation strategy which not only improves environmental conditions, but can also (potentially) contribute substantially to reduce food insecurity.
Yet, establishing the direct impact of agroforestry on household nutrition is not an easy task. The difficulties are shaped by not only the endogeneity of agroforestry strategy adoption but also by the existence of different linkages connecting agroforestry and nutrition. Also, the availability of the right data is fundamental. Moreover, analysing the relationship between agroforestry and nutrition with the intermediation of gender is relevant for the interpretation of this relationship, and it can also shed light on possible obstacles which limit the efficacy of policies aiming to promote agroforestry. Indeed, the results from this study, while creating supporting evidence of this relationship, will also inform on the design of policies aimed at strengthening household adaptation in agriculture and household resilience.
References
- Jarosz, L. (1997). Women as Rice Sharecroppers in Madagascar. In C. E. Sachs (Ed.), Women Working in the Environment. Routledge.
- Mellon, J. (2023). Rain, Rain, Go Away: 195 Potential Exclusion-Restriction Violations for Studies Using Weather as an Instrumental Variable.
- Widman, M., & Hart, R. (2019). Joint Land Titling and Household Bargaining in Madagascar. Feminist Economics, 25(4), 211–239.
