If You Clothe Them, They Will Come: The Provision of School Uniforms and Educational Outcomes

• In Event: Gender and Class Inequality in India and Nepal

Abstract

In order to increase access to education, many countries have eliminated tuition fees for students attending public schools. Despite widespread efforts to eliminate this cost barrier to schooling, the expense of school uniforms remains. For example, in rural India, the acquisition of school uniforms can represent over five percent of a family's annual income. Thus, some governments and non-governmental organizations have begun offering free school uniforms. Several small-scale studies have shown a positive relationship between such policies and educational attainment, but this study provides evidence on the impact of an ongoing, large-scale government-run school uniform program in Uttar Pradesh, India. In the spring of 2005, the Indian government announced a program to provide free uniforms to girls and disadvantaged-caste students in “educationally-backward” village groups for academic year 2005-06. The provision of a clothing-based schooling incentive within particular village groups provides a quasi-experimental setting to understand how reducing this cost of schooling can influence educational decisions for all children.

In order to increase access to schooling, many countries have eliminated tuition fees for students wanting to attend government (or public) schools. Despite widespread efforts to eliminate this cost barrier to schooling, the expense of school uniforms remains. While mandatory school uniforms may generate benefits by removing a visible symbol of socioeconomic status, the cost of school uniforms may be a prohibitive barrier for the poorest students. This additional cost can lead families to underinvest in the education for their children. In particular, parents under severe resource constraints may keep their children out of school or attempt to maximize productivity in a productivity-equity trade-off. Accordingly, they may determine intra-household allocations of resources based on who may be the most productive, therefore leaving the more vulnerable members of the family (children, and girls in particular) at a greater risk of underinvestment.

In order to reduce the cost of schooling and thus encourage school participation, especially for girls, uniform-distribution programs have been implemented in various developing countries by government agencies and non-governmental organizations (NGOs). The primary benefit of offering school uniforms free-of-charge to families is that it effectively reduces the cost of schooling by removing an expense that families would otherwise have to bear: purchasing school-specific clothing for their children. By removing a non-trivial cost, it removes one barrier that might otherwise discourage families from enrolling their children in school.

My paper adds to this literature by examining a large-scale, ongoing government program in the 951 blocks of Uttar Pradesh, India. During the spring of 2005, the Indian government announced a program to provide free school uniforms to girls and disadvantaged-caste students in “educationally-backward'' village groups in academic year (AY) 2005-06. Thus, recipient schools were chosen based on whether they were in a village that was defined as being “educationally-backward”: having a female literacy rate below the national average and having a gender gap in literacy above the national average. The introduction of free-uniform provision within particular village groups provides a quasi-experimental setting to estimate the causal impact of providing uniforms on educational outcomes of females and students of lower socioeconomic status.

In this paper, I investigate the impact of free-uniform provision on school enrollment, measured three months into the academic year. For my empirical strategy, I adopt a differences-in-differences framework to analyze school-level data from the District Information System for Education (DISE) dataset on Indian primary schools. DISE is an annual census of registered schools, reporting data on provision of school incentives, school infrastructure, and student educational outcomes by student sex, caste, and grade. I estimate the “first difference” by comparing the post-treatment school enrollment of a group of schools that offered free uniforms in AY 2005-06 (i.e. the “treatment” group) to their pre-treatment outcomes. To control for any secular trends in enrollment that may have occurred regardless of the intervention, I subtract a “second difference” estimated from the difference between pre- and post-treatment enrollment of another group of “comparison” schools that did not offer free uniforms any time between AY 2002-03 and AY 2006-07.

I find that the policy generally increased the enrollment of female students and students from historically-disadvantaged castes but decreased the enrollment of students from higher socioeconomic-status backgrounds. Additional research is needed to determine if the increase of students from lower socioeconomic-status backgrounds crowded out students from higher socioeconomic-status backgrounds or caused families to resort across schools.

Author

• Anjali Adukia, University of Chicago