Mahbub Hossain
Department of Agricultural Economics, Bangladesh Agricultural University, Mymensingh
Nutritional status, Rural children, Agricultural households, Bangladesh
Quality and Nutrition
Livelihood, Social status
Source of data and the sample: The Bangladesh Integrated Household Survey (henceforth BIHS), which is one of the most comprehensive household surveys available in Bangladesh until now (Ahmed, 2016), provides the data for this research. Notably the BIHS is representative of the rural areas of all administrative divisions of Bangladesh. It has collected particularly comprehensive child health-related information from those households in which a child up to the age of 24 months resides. Although the BIHS surveyed 6,500 households in 2012 and 6,436 households in the second wave in 2015; the numbers of households in which a young child up to 24 months resides are 1,136 in the first wave and 1,011 in the second wave. Thus, young children’s health and care-related information is available for 2,174 households in both waves; however, the sample of this study consists of 1,444 households: 866 households from the first wave and 578 households from the second wave. Such a reduction of the sample size occurs mainly because of the exclusion of households in which the young child is not the biological child of the household head. For instance, households in which the surveyed young child is a grandchild, nephew, or niece of the household head have been excluded. Furthermore, the sample size reduces because of the absence of relevant information as well as inappropriate anthropometric data. Eventually, the sample of this study comprises 1,444 children, one child from each household. Measurement of the nutritional status of the children: The key variable in this research is children’s nutritional status, which is generally denoted by the anthropometric indicators following the guideline of the World Health Organization (WHO). In this study, two anthropometric indicators, namely weight-for-age Z score (WAZ) and height-for-age Z score (HAZ) have been considered, respectively, to denote whether a child is underweight and stunted. Interested readers are referred to the WHO manual (WHO, 2006), stated in the reference, for more information on how the anthropometric indicators are calculated. The low height-for-age indicates cumulative deprivation of adequate nutrition over a long period of time, hence it is considered as an indicator of long-term nutritional status (NIPORT et al., 2016). A value of HAZ less than -2 and -3, respectively, means chronically stunted and severely stunted. Therefore, a binary variable ‘stunted’ is defined such that it takes the value 1 if the HAZ is less than or equal to -2 and 0 otherwise. On the other hand, low weight-for-age could mean either a child has low weight-for-height or low height-for-age. The WAZ score below -2 and -3, respectively, means chronically underweight and severely underweight. Following this, another indicator variable ‘underweight’ is defined such that it takes 1 if the WAZ is less than or equal to -2 and 0 otherwise. These two nutritional status markers ‘underweight’ and ‘stunted’ are used as the dependent variable in this paper. Estimation strategy: The conceptual framework of malnutrition has depicted that the causes of child under-nutrition can be multisectoral extending to individual, parental, household and societal level factors (UNICEF, 1990). In addition to this conceptual framework, several previous studies (Asfaw et al., 2015; Choudhury et al., 2017; M. M. Hossain et al., 2015; Rahman et al., 2009) have guided the estimation strategy adopted here which involves regressing indicator of child nutrition on a set of observable child, parents, household, and community level variables. The main intention here is to find variables that have significant association with the probability of child undernutrition.
The Bangladesh Journal of Agricultural Economics, 41(2): 63-80, 2020
Journal