Can microcredit improve food security among the rural poor?

A core objective of microcredit in Bangladesh is to make the rural poor more food secure. To what extent has this been achieved? Analysing household data from Bangladesh, this column finds that participants of microcredit programmes are more food secure, with improved calorie availability, reduced child stunting and better maternal nutritional status. However, programme participation in itself does not improve dietary diversity.

Despite the growth in food production and availability, food insecurity is still a major problem in large parts of South Asia. In rural Bangladesh, most of the income of a poor household is derived from agriculture, which exposes households to seasonality in agricultural employment, poverty and consumption. According to the United States Department of Agriculture, about 33 million people, out of Bangladesh’s population of 165 million, lacked food security in 2010 and this number is expected to increase to 37 million by 2020. A survey conducted by the Economist Intelligence Unit in 2012 ranked Bangladesh 81st out of 105 countries in terms of the Global Food Security Index. This poor standing seems surprising, given that improving the food security situation of the rural poor had motivated the development of microcredit institutions in Bangladesh and that Bangladesh is home to one of the oldest microcredit programmes in the world. The Grameen Bank has clearly documented that the objective of microcredit is to improve food security of the rural poor (Yunus 1994). Yunus received the World Food Prize in 1994 for his original approach to promoting social, human and economic development of the most vulnerable in Bangladesh, by providing the impoverished and undernourished with access to adequate food and nutrition for the first time in their lives, through microcredit. However, evaluation studies of microcredit programmes have yet to document if and how much microcredit helps address food insecurity.

The microcredit sector in Bangladesh has grown phenomenally since the 1990s. Various economists have evaluated the impact of microcredit on income, consumption expenditure and poverty. Overall, it is fair to say that the findings have been mixed. Banerjee et al. (2015) found that microcredit has insignificant effect on income, consumption and child schooling but significant effects on business development among slum households in India. A number of studies using data from Bangladesh have found that microcredit leads to an improvement in consumption smoothing and asset building¹, reduction in household vulnerability in response to health and income shocks (Pitt and Khandker 1998, Islam and Maitra 2012, Islam 2015), and improved health and nutrition (Pitt et al. 2003). It seems clear that higher benefits from microcredit are likely to accrue from long-term participation in such programmes (Islam 2011).

Microcredit and food security

In the first study of its kind, we examine how microcredit could enable households to become food secure, by using a large household-level dataset, collected over eight years (Islam, Maitra, Pakrashi and Smyth 2015)². A notable feature of our study is that we attempt to define food security broadly using multiple indicators, with a view to capture its various dimensions – availability, access and utilisation. Specifically, we evaluate the impact of microcredit on three alternative indicators of household food security – absolute level of calorie availability and shortfall from standard cut-offs of calorie norms (referred to as food poverty), dietary diversity indicators, and anthropometric indicators for children below five years of age (incidence of stunting, wasting and underweight) and women of reproductive age (15-49 years) (focussing on Body Mass Index (BMI) and Mid-upper arm circumference (MUAC)³).

Our results suggest that households participating in microcredit programmes are able to increase their equivalised calorie availability per day by 2.75%. Participating households are also found to be more likely to meet the minimum energy dietary requirements. Using a calorie cut-off of 1,805 kcal per person per day to define “hard-core food poverty” and 2,122 kcal per person per day for “absolute food poverty”, we find that participation in microcredit programmes reduces the incidence of “hard-core food poverty” by 3.7% but has no significant effect on “absolute food poverty”. It appears that the underlying calorie-income relationship is non-linear in this case – the calorie intake of households at the lower end of the income distribution is found to be highly responsive to increased access to credit, but calorie intake does not increase with further increases in expenditure beyond a certain threshold.

However, when food security is measured using dietary diversity indicators, microcredit programme participation is found to have a mixed impact. Using the most common dietary diversity indicators; namely, the food variety score (FVS) and the dietary diversity score (DDS)⁴, we find that microcredit programme participation in itself has no effect on diversity in the diet of the households, but increased access to loans provided by the microcredit institutions significantly increases household dietary diversity. This finding suggests that the dietary diversity of participating households is not significantly different from non-participants as below a certain threshold level of expenditure, poor households tend to concentrate on calories rather than on dietary diversity; however, once that level is met, further increases in expenditure (through increased loan amounts) may lead to increases in both calorie and dietary diversity.

Microcredit is also found to decrease the incidence of severe stunting among children under the age of five by 16.6% and stunting by 22.4%, suggesting that it improves child health in the long run. On the other hand, the effect of microcredit participation on wasting and being underweight is somewhat mixed and therefore inconclusive. Membership in microcredit programmes is also found to reduce the prevalence of maternal underweight by 6.21%. Increased access to microcredit, however, is found to improve both maternal BMI and MUAC.

Short vs. long run

The impact of microcredit programme participation, measured in terms of the extent of loan amount, on household food security is found to vary by the duration of participation in the programme. Microcredit may have no significant effect on food security in the short run, but improves all indicators of food security in the long run – measured in terms of reducing the incidence of food poverty or improving the quality of the diet of households. Long-term participants in microcredit programmes with increased access to loans are found to consume more calories and are less likely to be food poor, reflected by a decline in hard-core and absolute food poverty.

Mixed effects

Overall, we find mixed evidence of microcredit participation on food security status of participating households. Results suggest that microcredit programme participants are relatively more food secure with improved calorie availability and reduced levels of food poverty. However, microcredit programme participation in itself does not increase dietary diversity. Additionally, we find that participation in such a programme has mixed effects on anthropometric measures for children and women of reproductive age. Microcredit participation reduces child stunting suggesting that microcredit improves child health in the long run; however, it has a mixed effect on child wasting and on being underweight. As far as women’s health is concerned, microcredit programme participation improves nutritional status by improving BMI.

Notes:

Increased access to loans provided by microcredit institutions allow credit-constrained households to diversify into more productive self-employment based activities, resulting in higher income and better livelihoods. It also helps to smooth income and consumption of poor rural households as such activities generate income streams that are not associated with income from agricultural production, which is seasonal.
Our sample includes about 3,000 rural households selected from 91 villages across 23 thanas (sub-districts) within 13 districts of Bangladesh. We conducted the survey over a period of eight years, via four rounds 1997-98, 1998-99, 1999-2000 and 2004-05.
Low BMI and MUAC indicate incidence of acute malnutrition among women of reproductive age.
FVS is a measure of the number of food items consumed by a household during the reference period while DDS captures the number of food groups consumed by a household. The DDS ranges from zero (“non-diverse”) to 12 (“diverse”) on the 12-scale Household Dietary Diversity Score (HDDS) developed by the Food and Nutrition Technical Assistance (FANTA) Project of the United States Agency of International Development (USAID). This has been used to aggregate food items consumed under the following 12 food groups — cereals, roots and tubers, vegetables, fruits, meats, eggs, fish and other seafood, pulses/legumes/nuts, milk and milk products, oils and fats, sugar and honey, and miscellaneous such as spices, condiments and beverages.