Factors behind Access to Latrine in India:An Application of Multinomial Logistic Regression Model

Abstract

This paper attempts to find out the factors that influence the owning of a latrine. For this purpose, unit level NSSO data of 69 th round, collected exclusively to get an idea of this, has been used. On the extracted unit data, Multinomial Logistic Regression Model has been applied. Different sets of socio-economic and governance indicators are chosen as the predictors. Based on the available questions, the type of latrine has been divided into three categories-latrine exclusively use for household. Second one clubs different types of latrines to form as fixed point latrines. The third one is categorised as no latrine. In running MLR, no latrine is considered as reference category. All the predictors, starting from religion, region, location, caste and economic conditions have positive and significant impact on owning a latrine. Therefore, the latrine problem, usually being typed as only traditional and cultural problem, needs to look from holistic angle and it is recommended that while designing public policy all these factors are considered to mitigate the problem.