Household Determinants of Multi-Saving Behaviour in East Africa: Evidence from Uganda, Kenya, and Tanzania
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Household saving is central to financial resilience and long-term development, yet most empirical work treats saving as a binary outcome whether a household saves or not without recognising the diversity of saving channels available in African financial ecosystems. This paper models saving as a multi-tiered behaviour, capturing the number of distinct mechanisms a household uses to save. Using nationally representative data from Uganda, Kenya, and Tanzania, the study applies multinomial logistic regression to estimate the determinants of saving through one, two, three, and four or more mechanisms, relative to non-saving.The analysis incorporates demographic, socioeconomic, financial inclusion, infrastructural, and housing variables, including tenure, floor and roof materials, water and energy access, borrowing history, mobile money and bank account usage, household decision-making, and indicators of economic stability. Results show that multi-saving behaviour is strongly associated with housing quality and infrastructure, financial inclusion, borrowing status, urban residence, and decision-making autonomy. In Uganda, saving diversification is closely linked to housing quality and basic infrastructure, with informal and semi-formal channels (SACCOs, VSLAs, ROSCAs, and asset-based saving) remaining central. In Kenya, digital finance especially mobile money combined with economic stability and joint financial decision-making, drives progressive saving diversification. In Tanzania, urbanisation, modern energy access, and formal employment structures are key, while bank account ownership remains an indispensable gateway to multi-channel saving.Across the three countries, the findings highlight consistent structural drivers of saving diversification and reveal non-linear patterns in gender and borrowing effects as households transition from credit-supported to asset-based saving at higher tiers. Policy implications include the need to deepen financial inclusion beyond account opening, link formal and informal saving systems, invest in infrastructure and modern energy, and promote gender-responsive and autonomy-enhancing financial interventions.