Household diversification and market dependence: understanding vulnerability in rural West Africa

ABSTRACT

Understanding the empirical relationships among three parameters is necessary to reduce household vulnerability: diversification, market dependence and economic security. A cross-sectional study of these relationships was conducted in five areas with histories of cash cropping in Mali, Burkina Faso, Niger, and northern Nigeria. Household diversification was found to be positively associated with economic security while negatively associated with household dependence on markets. Moreover, individuals in market oriented households are more likely to adopt individualistic orientations. These findings confirm a positive role of diversification but raise concerns about how markets affect the prospects for maintaining or reducing household vulnerability.

KEYWORDS:

• Climate change
• poverty
• cash cropping
• commodification
• livelihoods

Acknowledgements

This research was funded by by the Research Program on Dryland Systems of the Consortium of International Agricultural Research Centres. We are indebted to our research informants who showed both patience and thoughtful engagement with our questions. We thank Kalilou Adamou, Sunday Elaigwu, Oumarou Moumouni, Clarisse Umutoni, and Viviane Yameogo for their contributions to the research. We also thank two anonymous reviewers for their useful comments and suggestions. Any deficiencies of this work are our own.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Matthew Turner is the Vilas Distinguished Achievement Professor of Geography at the University of Wisconsin–Madison. Taking a political ecology approach, he has worked on a range of topics in the Sudano-Sahelian region of West Africa over the past 30 years including environmental governance, climate change vulnerability, farmer-herder conflict, gender and household economy, land tenure, food security and nutrition, grazing ecology, pastoral mobility, social roots of soil fertility variation, and land-cover change. He is co-editor (with Mara Goldman and Paul Nadasdy) of Knowing Nature: Conversations at the Intersection of Political Ecology and Science Studies.

Augustine Ayantunde holds a PhD degree in ruminant nutrition from the University of Wageningen, The Netherlands. He is currently a principal scientist in sustainable livestock systems at International Livestock Research Institute (ILRI) and the regional representative for West Africa. He is based at ILRI office in Ouagadougou, Burkina Faso. He has more than 20 years of experience in participatory testing and evaluation of livestock-related strategies for sustainable intensification of crop-livestock systems in West Africa, evaluation of feed resources in West African Sahel, and evaluation and monitoring of natural resource use in (agro)-pastoral systems including conflict management. He has been responsible for the coordination of many bilateral projects. Ayantunde has authored and co-authored more than 100 scientific papers, book chapters, conference contributions and technical reports.

Notes

1 Please note that two rural communes were chosen in the Koutiala Cercle of Mali with one representative village chosen from each commune.

2 The question as posed may have caused some confusion since it was worded as: Do you hold outstanding debts incurred to buy grain to feed your family? In some cases, it is possible that the interviewee may have responded in reference to whether he or she individually had outstanding debts rather than the intended meaning of the question of whether there were any outstanding debts incurred to feed all family members. Given that the interviewees were the household head and one of his wives, this error may likely not be large since it is the household head who is responsible for feeding the family and who is thus, most likely to incur such debt.

3 The method of Smith, Barrett, and Box (2001) is to normalize ordinal ranks using the formula: $1-{\displaystyle \frac{\left(rank-1\right)}{Numberofitemsranked}}$. To avoid having the ranks of those respondents, who mention and rank more responses than others, to weigh more heavily in statistical analyses, the normalized ranks calculated in this way were further modified by dividing each normalized rank by the sum of the respondent’s normalized ranks. In this way, the ranks produced for each informant all sum to 1.

4 This is a binary variable with those individuals of equal seniority to the household head (household head or his mother, father, brother, sister, or wife) having the value of 1, otherwise, all other individuals have the value of 0.

5 Data of grain acquisitions over the previous year support the conclusion that the number of months not supported by grain harvests is a good measure of market dependence since 95% of grain acquisitions were purchases (using cash income or loans) while 5% represent grain acquired through barter, direct loans, or gifts.

Additional information

Funding

This work was supported by the Research Program on Dryland Systems of the Consortium of International Agricultural Research Centres.