Spatial Analysis Using Big Data

5.3. Parameter estimation of the spatial econometric models

5.3.1. Ordinary least squares method

The most widely used parameter estimation method in the CLR model is OLS. However, it is well known that the OLS estimator is not the best linear unbiased estimator if spatial autocorrelation exists. In the following, after briefly discussing this, we explain a representative parameter estimation method of the spatial econometric models.

First, let us consider the following simple equation where spatial autocorrelation exists among observations. This model, for example in Anselin (1988), is called the first-order spatial autoregressive model.

y = ρ W y + ε

(5.3.1)

The OLS estimator of   ρ, say ρ ˆ , is obtained