Room with a View or Rear Window? Hedonic prices of the Parisian real-estate

Real-estate market is characterized by a high price variability. Under some theoretical hypotheses, the hedonic approach permits to determine which attributes are valued by the buyers and to quantify the corresponding marginal willingness to pay at the equilibrium. Thanks to a semi parametric method, I estimate the hedonic price function which relates the price to the bundle of house characteristics. My data, geocoded at the address, describe very precisely the real-estate property and its neighborhood and explain more than 90% of the variance of prices. For a robust estimation of the marginal willingness to pay for the different attributes, I introduce fixed-effects at a very local level (grand-quartier or IRIS). I find a substantial positive marginal willingness to pay for job accessibility (of approximately 2%) and middle school quality (of approximately 1%) and a weaker but significant negative marginal willingness to pay for a higher crime rate in the area (of approximately 0.01%). By contrast, noise level or public transport accessibility seem to have less influence on housing prices.