India’s Ease of Living Index Report 2018 – A data driven approach

Abstract

Government of India, Ministry of Housing and Urban Affairs released the first ever Ease of Living Index Report 2018. The report ranked 111 cities in India on the basis of multiple facets of urban living. The underlying data for each of the attributes was also provided. The ranking was based on a-priori weightages to different dimensions of the analysis. This paper attempts to refine the ranking and applies the technique of Partial Least Squares Path Modeling, a technique which may be an alternative for such analyses. The results show that the regression coefficients estimated through the analysis show some variation compared to the weights assigned in the report. Results prima facie indicate that emphasis of different facets of urban living is different. It is also different for different city sizes when the cities were considered as large and small. In addition to the PLSPM Model, additional models have been prepared including neural network models. The neural network and Random Forest models appear acceptable based on accuracy of fit of the models, as represented by Pseudo R2. The report for 2019 is yet to be released. This paper is an academic paper having relevance for research in urban planning with its boundaries touching upon economics and land use. With passage of time and with evolution of the concept of ease of living and forming an index, we may witness different variables coming into play and having different loadings on the outcome of ranking of cities. It could be useful for students of Real Estate Programs in the country and abroad as also for those pursuing Urban Planning and related academic disciplines. Familiarity with this concept and its variation will help them become better data science professionals as they seek gainful employment in various enterprises associated with urban planning and real estate development.