## Abstract

This note studies the Lee and Yu (2009) spurious regression model for the special case where the weight matrix is normalized and has equal elements, and where the nonstationarity is caused by near unit roots. It shows that spurious spatial regression will not occur in a spatially autoregressive (SAR) model when the spatial weight matrix is row-normalized and has equal weights. In fact, the asymptotic distribution of the OLS estimate will always converge to its true value zero. The only condition required is that the spatial coefficients of the dependent and independent variables be both less than 1, which is a requirement for the SAR model to be an equilibrium model.

Original language | English (US) |
---|---|

Pages (from-to) | 1640-1642 |

Number of pages | 3 |

Journal | Statistics and Probability Letters |

Volume | 80 |

Issue number | 21-22 |

DOIs | |

State | Published - Nov 2010 |

## Keywords

- Equal weights
- Spatial autocorrelation
- Spurious spatial regression

## ASJC Scopus subject areas

- Statistics and Probability
- Statistics, Probability and Uncertainty