Optimal local property tax for redistribution and public good provision and the estimation of demand for public education: An application to a model of discrete community choice

The dissertation consists of two essays. The first examines optimal local property tax for redistribution and public good provision. Suppose that local government finances a public good or a poll subsidy by imposing a property tax on housing consumption which is produced by a private good and land owned by absentee landlords. This paper shows that there exists a taxation rule which makes the property tax on housing non-distortionary. When a median voter decides a level of property tax for redistribution, it is shown that the median voter chooses a non-distortionary level of property tax when the income distribution is symmetric and mobility is absent. With mobility, the median voter sets the property tax rate above the non-distortionary level if out-migration of rich households is expected. Therefore, out-migration of rich households is not a constraint for the median voter when property tax is exported to absentee landlords.;The second essay considers the issue of estimating the demand for local public education. As a theoretical model for the estimation of demand for local public education, a discrete community choice model popular in local public finance literature is considered. Applying nested-logit model, consistent estimates of demand parameters are obtained from the estimates of the parameters of indirect utility function. The result in my paper shows that the price and income elasticities of demand for public education increase steadily as income increases. It is also found that the estimated distribution of households' benefit, net of local taxes, is pro-poor.