acreg: Arbitrary correlation regression

• Autores: Fabrizio Colella, Rafael Lalive, Seyhun Orcan Sakalli, Mathias Thoenig
• Localización: The Stata journal, ISSN 1536-867X, Vol. 23, Nº. 1, 2023, págs. 119-147
• Idioma: inglés
• Enlaces
  • Texto completo
• Resumen

  • We present acreg, a new command that implements the arbitrary clustering correction of standard errors proposed in Colella et al. (2019, IZA dis- cussion paper 12584). Arbitrary here refers to the way observational units are correlated with each other: we impose no restrictions so that our approach can be used with a wide range of data. The command accommodates both cross-sectional and panel databases and allows the estimation of ordinary least-squares and two- stage least-squares coefficients, correcting standard errors in three environments:

    in a spatial setting using units’ coordinates or distance between units, in a network setting starting from the adjacency matrix, and in a multiway clustering frame- work taking multiple clustering variables as input. Distance and time cutoffs can be specified by the user, and linear decays in time and space are also optional.