Impact of microfinance on rural household diet diversity and expenditure in India

Thesis Abstract

In India's rural areas, microfinance initiatives have expanded financial access for low-income households, yet their impact on diet diversity and household expenditure remains contested due to heterogeneous loan utilization, repayment pressures, and differing regional food environments. This study addresses the problem of understanding how microfinance participation influences nutritional diversity and expenditure patterns among rural households, with a view to informing policy design and microfinance program tailoring. The aim is to quantify the effect of microfinance participation on diet diversity scores and food expenditure shares, and to identify mediating pathways such as credit use for productive inputs, income stabilization, and women's empowerment. The specific objectives are (i) to estimate the association between microfinance membership and household diet diversity, (ii) to assess the impact of microfinance on total food expenditure and its composition across food groups, (iii) to examine whether income shocks, productive investment, and intrahousehold time allocation mediate these relationships, (iv) to explore heterogeneity of effects across landholding sizes, gender of the loan holder, and regional agricultural productivity, and (v) to test robustness across alternative model specifications and endogeneity controls. A multi-stage stratified sampling design is employed in the study. The population comprises rural households in four states representing diverse agro-ecologies and microfinance penetration Bihar, Madhya Pradesh, Tamil Nadu, and Maharashtra. A sample of 2,400 households is drawn using probability proportional to size from district-level listings, with equal representation of microfinance participants and non-participants within strata. Data collection combines a structured household survey (diet diversity measured via the Household Diet Diversity Score and the Minimum Dietary Diversity for Women indicator where applicable) and a 12-month expenditure diary, complemented by module modifications to capture microfinance exposure (loan size, frequency, purpose, repayment status) and asset ownership. Instrument validity is ensured through pre-testing, translation-back translation procedures, and Cronbach’s alpha testing for multi-item scales. Data collection also integrates qualitative components with 40 in-depth interviews to contextualize quantitative findings. The analysis proceeds in three stages. First, descriptive statistics summarize household characteristics, microfinance exposure, diet diversity, and expenditure patterns. Second, multivariate regression models estimate the association between microfinance participation and diet diversity and food expenditure, controlling for household socioeconomic status, agricultural productivity, market access, and household assets. Regression specifications include OLS for continuous diet diversity scores, and generalized linear models for expenditure shares with appropriate link functions. Third, structural equation modeling (SEM) tests mediated pathways from microfinance participation to diet diversity via income smoothing, credit for productive investments, and women’s decision-making autonomy, while propensity score matching (PSM) and instrumental variable (IV) approaches address potential endogeneity arising from self-selection into microfinance programs. Regional fixed effects and clustering at the district level are applied to account for unobserved heterogeneity. Key expected findings include (i) a positive association between microfinance participation and diet diversity, particularly among households that use loans for income-stabilizing activities and women-led loan arrangements; (ii) increased total food expenditure with a reallocation towards nutrient-dense staple and micronutrient-rich foods in positively empowered households, and mixed effects where repaid loans are diverted to non-food uses; (iii) mediation effects confirming that income smoothing and enhanced women’s control over resources partially explain diet diversity improvements; (iv) significant heterogeneity with larger effects in resource-poor districts and among smallholders with limited off-farm income sources. The study contributes to knowledge by linking microfinance participation to nutritional outcomes through a detailed examination of expenditure structure and mediating mechanisms within a rigorous causal inference framework. It integrates theories of financial inclusion, the Livelihoods and Empowerment framework, and the Food Security and Dietary Diversity perspectives to interpret findings. Policy implications include targeted microfinance product design to emphasize productive investment, women-centric loan features, and integrated nutrition-sensitive financial services. The study advises scaling region-specific support services, incorporating nutrition education into microfinance programs, and strengthening monitoring systems to track dietary impacts alongside credit performance.