The “State” of Persons with Disabilities in India

Abstract

Among countries with comparable levels of income, India has one of the more progressive disability policy frameworks. However, people with disabilities in India are still subject to multiple disadvantages. This paper focuses on state-level variations in outcomes for people with disabilities to provide an explanation for the contrast between the liberal laws on paper and the challenges faced by people with disabilities in practice. Using average monthly per-capita expenditure as an indicator of economic well-being, instrumental-variable Wald estimator results indicate that households with members with disabilities have expenditures that are 14% lower compared with households with able members. This effect is most pronounced among families with male adults and children with disabilities, and in states that are relatively poor, relatively more urban, those that experience extremes in annual rainfall and temperature, and those that have low to medium levels of inequality.

Keywords:

• Disability
• India
• Monthly expenditure
• State variations
• Poverty
• Inequality

JEL codes::

• O12
• I15
• I18

Acknowledgements

Thanks to Sarita Bhalotra, Alka Bhargava, Jeanine Braithwaite, Deon Filmer, Marty Krauss, P. S. Meena, Philip O'Keefe and Aleksandra Posarac for helpful suggestions and comments. We thank two anonymous referees whose comments have greatly improved the paper. N. Menon is grateful for a faculty research grant from Brandeis University. The usual disclaimer applies.

About the Author

Nidhiya Menon is Associate Professor of Economics at Brandeis University, Waltham, MA, USA. Her research interests are development economics, health, labor and gender issues. She has worked in several areas including microfinance, occupational diversification, child health and women's labor choices in countries of South Asia, East Asia and Africa. She teaches classes in statistics, development economics and international economics. She holds a PhD from Brown University, Providence, Rhode Island, USA.

Susan L. Parish is Nancy Lurie Marks Professor of Disability Policy at Brandeis University, Waltham, MA, USA. Her research examines the health and financial well-being of children and adults with disabilities, as well as their caregiving families. She is particularly interested in the health of women with disabilities and the impact of poverty policies on people with disabilities. She teaches classes in disability policy that emphasizes both qualitative and quantitative methods. She holds a PhD from the University of Illinois, Chicago, IL, USA.

Roderick A. Rose is Research Associate and Project Director at the Carolina Institute for Public Policy at the University of North Carolina, Chapel Hill, NC, USA. He has a background in economics and is versatile with advanced econometric methods. He specializes in research methods for intervention and human services evaluation, with interest in missing data imputation and hierarchical linear modeling. He holds an MSc in Economics from the University of North Carolina at Chapel Hill.

Notes

1. Under India's federal set-up, the State List consists of 66 topics that state governments may legislate on independent of the central government. “Relief of people with disabilities and unemployable” is one of these 66 topics.

2. Measuring the institutional capacity of a state as the sum of planned and non-planned revenue expenditures between 2002 and 2003 normalized by state population, pair-wise correlations between this variable and indicators of state support in 2002 and 2003 for persons with disabilities, such as central government grants to NGOs to promote voluntary action for persons with disabilities, central assistance to persons with disabilities for purchase/fitting of aids and appliances and the number of special schools for disabled children, reveals no correlation that is significant at the 5% level. Data on state-wise revenue expenditures and total population were obtained from the Reserve Bank of India's Database on the Indian Economy. Sources for state-level measures of support for persons with disabilities are discussed below.

3. We think of these as exploratory results.

4. However, note that such under-reporting leads to a conservative bias in our estimates; that is, with the inclusion of such members, our results should become stronger.

5. For example, if states provided more for disabled populations because residents had lower average monthly expenditures to begin with.

6. An example of an NSS disability question that is asked when the demographic and other particulars of household members are being compiled is “whether (a particular member) has a disability?” The (coded) responses allowed are “yes: single: mental-1, visual-2, hearing-3, speech-4, locomotor-5; multiple-6; no-7.” Note that disability status is self-reported and beyond a broad question outlining the extent of physical disability (“cannot take self-care even with aid/appliance, can take self-care only with aid/appliance, can take self-care without aid/appliance or aid/appliance not tried/not available”), there is no distinction made between minor and major disabilities.

7. “Reservation” implies that jobs or seats in educational institutions are only to be filled by persons with disabilities (are reserved for them). Hence, for example, 1% of vacancies in government jobs in certain posts may only be filled by those with visual, orthopedic or hearing impairments.

8. The household mean of an indicator for those who do not work is 0.281 in households with disabled members versus 0.252 in households without disabled members. The annual mean medical expenditure (medicines, tests, doctors’ fees and hospital and nursing home charges) in households with disabled members is about Rupees 9000 versus Rupees 8400 in households with able members.

9. The full set of results for all controls in the Wald estimator models are presented in Appendix Table A3. Note that these results indicate that state spending on disability programs increases household expenditures, suggesting that in these specifications the institutional capacity of the state that may be reflected in such spending has a positive impact on the welfare of households with persons with disabilities.

10. To further check the validity of the identifying instrument and these results, we re-estimated the model using the standard two-stage least squares (TSLS) technique. The identifying instrument in the first stage is the same as in the Wald estimator models: district mean of the disability indicator at different levels. Results from the TSLS models were broadly consistent with those of the Wald estimator models: the expenditure penalty is highest in households with a child (children) with disabilities and male disabled members; the penalty is not evident in households with females with disabilities. The TSLS models have the further advantage of reporting tests of instrument validity, and in all cases the tests reject the null hypothesis that the models are weakly identified (Kleibergen–Paap statistics) or under-identified (Craig–Donald statistic). Since the equations are exactly identified (one identifying instrument in each case), no test of over-identification (Hansen's J) is reported. Further, tests of endogeneity (equivalent to a Durbin–Wu–Hausman test) strongly reject the null that the disability indicator at the household level, for males, for females and for child (children), is exogenous. Since we already report results from three alternate specifications in the paper (random coefficients, OLS and Wald estimator), the TSLS results are not reported but are available on request.

11. Appendix Table A4 shows the location of states by percentiles of the economic and climate-related state characteristics considered in this section.

12. Unlike the previous tables, the F-statistics and p values for the null test of equality of state effects are not reported in Table 3 because of space constraints. These are available on request.

13. We acknowledge that it is not always the case that central government-backed schemes have equally propitious impacts across states; rather, disparities may still remain as effects depend on the state's effectiveness in administration. However, we do find that these state-level measures of centrally provided grants have positive impacts on welfare.