Estimating the preferences of the households for urban amenities in Delhi

The object of the research is to examine the robustness of a result, obtained in a paper written with S. Thoron, which provides a condition on households preferences which is necessary and sufficient for a process of endogenous formation of jurisdictions to give rise to stratifications. An assumption that clearly needs some weakening is the absence of a housing, or land, market. In the Gravel and Thoron model, households can live in a jurisdiction, and participate in the collective choice of the tax rate without buying or renting housings. It appears therefore important to enlarge the decision space of the household from the public-private spending one to the public spending, housing space and private spending and to see how the conditions identi-fied in the Gravel and Thoron paper can be reformulated in this enlarged decision space. Once housing is incorporated into the analysis, one can use data on the housing market to test, using hedonic methodology, whether or not the condition on households’ preference is satisfied. Hedonic methodology consists in using cross-section data on housing prices, housing characteristics and households’ characteristics to estimate the households’ preferences.

The socio-spatial dynamics of Indian cities, especially in Delhi, appears quite suitable for this sort of test. The aim of this case study will be, therefore, to test whether or not households’ preferences in Delhi, as they can be estimated by hedonic methods, satisfy the condition identified in Gravel and Thoron.

We plan to construct a data set on of some 5000 housing units in Delhi, based on sample surveys. The sample will be representative of the various segments of the housing market and of the geographical spread of the urban agglomeration. The data set will then be completed by matching the information on individual housing prices and characteristics with aggregate data on the local environment of the housing.

Besides its use for testing the particular hypothesis of the Gravel and Thoron model, the availability of such a rich data set will be of invaluable help to better understand the urban dynamics in Delhi, as well as the behaviour of housing price, and the implicit valuation of the various urban amenities that these prices reveal.

  • Participants :
    Véronique DUPONT