Children (under the age of 15 years) growing up in poor and/or nutritionally deprived households also live with a number of layers of deprivations that stifle their freedom to actively participate in and benefit from elementary school education. Lack of healthcare, limited access to quality schooling and opportunity cost of participation in education are some of these layers. The Human Development Report 2010, using Oxford University’s newly developed Multidimensional Poverty Index (MPI), adds more dimensions to poverty measures over and above those of the Indian Planning Commission’s (2009) new measure or absolute poverty used in this paper. These enrich our understanding but do not directly deal with children growing up in absolute poverty and non-participation in schooling. This issue can be meaningfully explored by using the household as the unit of analysis. The paper uses household level data for 2004–05 (NSS 61st Round) and 1993–94 (NSS 50th Round) for India and also major states to analyse these issues. We start with the size of child population, changing share of states and uneven demographic transition in India (particularly the movement in Total Fertility Rates across Indian states during the period 1961–2001.

Changes in the number of children and the household size in very poor, poor, non-poor low-income and non-poor high income households from 1993–94 to 2004–05 are analysed within the cross-sections and also between the two cross-sections. Participation in education, and non-participation separated as child labour and ‘nowhere’ (neither in schools nor in labour force) by poverty status at the all-India and the state levels are reported and commented upon. Changes in magnitudes and proportions of children in poverty in India and across states during the years 1993–94 and 2004–05 are presented and the share of some states in these magnitudes is highlighted. The determinants of non-attendance in schools (that is, a child being in the labour force or ‘nowhere’) for 5–14 year olds are analysed by using formal econometric models—Probit with binary variables and also Multinomial Logit Models. The results are robust and confirm our descriptive analysis. Finally, the broad features of The Free and Compulsory Elementary Education Act, 2009 (Law w.e.f. April 2010) are reported and linked to the policy implications of our empirical findings for meaningful implementation of the Elementary Education Law. The potential usefulness of the Unique ID in the delivery of child-focused services and monitoring is also highlighted.