The research, published in BMC Agriculture, was led by Eucabeth Majiwa of Jomo Kenyatta University of Agriculture and Technology together with colleagues from the International Food Policy Research Institute and other institutions. The team analyzed three waves of the World Bank’s Living Standards Measurement Study surveys, covering 2012/2013, 2014/2015, and 2019/2020. After matching households across all three waves, the final balanced panel comprised 512 Tanzanian households tracked over seven years, allowing the researchers to observe how the same families changed their income strategies and their eating patterns over time.

The stakes are high. According to the 2025 State of Food Security and Nutrition report by the Food and Agriculture Organization, roughly 2.3 billion people worldwide were moderately or severely food insecure in 2024, and 22.3 percent of Sub-Saharan Africa’s population remains undernourished. East Africa carries a disproportionate share of the burden, with severe food insecurity in the region rising from 20.6 percent in 2021 to 24.4 percent in 2024. Tanzania, meanwhile, faces a triple burden of malnutrition: underweight, overweight, and micronutrient deficiencies coexist, and at least 36 percent of women aged 20 to 49 are obese or overweight, compared with 17 percent of men.

To measure diversification, the researchers used two complementary metrics. The first was a simple count of livelihood options, drawn from eight income sources captured in the survey data, including crop sales, livestock income, wages, non-farm profits, remittances, and pensions. The second was the Herfindahl-Hirschman Index, a concentration measure familiar from industrial economics, calculated as the sum of squared income shares across sources. A household relying entirely on one activity scores near 1, while income spread evenly across many activities pushes the index toward 0. Lower values therefore signal greater diversification.

Food and nutrition outcomes were assessed with three indicators. The Household Dietary Diversity Score counts how many of twelve food groups, from starchy staples and green leafy vegetables to eggs, fish, and milk, a household consumed over a seven-day recall period. The Food Consumption Score, developed by the World Food Programme, weights food groups by their nutritional importance. Total household food expenditure served as a third, monetary gauge of food access. Together, the three measures capture both the variety and the economic depth of what families actually eat.

The descriptive results revealed a paradox hiding in plain sight. Dietary diversity scores rose from a pooled mean of 7.89 in 2012/2013 to 8.22 in 2014/2015, then slipped back to 8.07 in 2019/2020, a statistically significant trajectory. Food expenditure peaked in 2014/2015 and declined thereafter. Meanwhile, the Herfindahl-Hirschman Index fell from 0.77 to 0.73 between the first two waves before rising to 0.75, indicating that households were more diversified in 2019/2020 than at the study’s start, yet less diversified than at the midpoint. In other words, diversification edged upward even as food security indicators deteriorated.

The econometric analysis sharpened the picture. Using a fractional panel regression with fixed effects, appropriate because the diversification index is bounded between zero and one, the team found that female-headed, younger, and wealthier households diversified more, while asset ownership, mobile money adoption, and exposure to severe shocks were associated with less diversification. Larger landholdings and more livestock also reduced the likelihood of diversifying, suggesting that households with productive assets can meet their needs from a single, higher-return activity. Notably, households hit by severe shocks diversified less, possibly because shocks such as the prolonged 2018/19 dry spell and Fall Armyworm infestations destroyed the very capital needed to launch new income activities.

The fixed-effects impact models delivered the study’s most striking result: as the Herfindahl-Hirschman Index decreased, meaning households diversified further, dietary diversity, food consumption scores, and food expenditure all declined significantly. The authors suggest two mechanisms. First, much diversification in poor rural settings is necessity-driven, a coping response to failed harvests or lost wages rather than a genuine opportunity, and it channels families into low-return informal work that generates too little income to improve diets. Second, cultural and contextual factors may limit how additional income translates into food choices, particularly for women who often have less control over household resources and less diverse diets than men in the same households.

Yet the study is not a brief against diversification itself. When measured as a count of livelihood options, adding one more income source was associated with improvements in food consumption of at least 1.5 times, a positive and significant effect. The divergence between the two metrics implies that the composition of diversification matters as much as its extent: spreading labor across many marginal activities can hollow out food security, while adding one substantial, remunerative activity can strengthen it. Livestock ownership, mobile money use, and larger household size were all positively associated with food expenditure and dietary quality, pointing to pathways that reliably nourish families.

The policy implications are concrete. The authors argue that Tanzania should promote high-return diversification rather than diversification at any cost, encouraging income-generating activities that build genuine resilience. Interventions should be gender- and wealth-sensitive, since female-headed and poorer households diversify out of survival and reap the least nutritional benefit. And to counter carbohydrate-heavy diets dominated by roots, tubers, and cereals, policymakers should support diversification into nutrient-rich production such as dairy livestock and fruit trees. As climate shocks intensify across East Africa, the study’s central lesson will only grow more urgent: more income streams are not automatically better diets, and the quality of each stream determines whether diversification feeds families or merely exhausts them.

Subject of Research: The impact of livelihood diversification on household food security and dietary diversity in Tanzania

Article Title: Impact of livelihood diversification on household food security and dietary diversity in Tanzania

Article References: Majiwa, E., Mwangi, C., Kiage, B., Obebo, F., Ateka, J., & Mbeche, R. (2025). Impact of livelihood diversification on household food security and dietary diversity in Tanzania. BMC Agriculture, 1(1), Article 17. https://doi.org/10.1186/s44399-025-00017-7

Image Credits: AI Generated

DOI: 10.1186/s44399-025-00017-7

Keywords: livelihood diversification, food security, dietary diversity, Tanzania, Sub-Saharan Africa, household panel data, Herfindahl-Hirschman Index, malnutrition, rural livelihoods, fixed-effects regression, gender, mobile money