Production function estimation in Stata using the Ackerberg–Caves–Frazer method

• Manjón, Miguel ^[1] ; Mañez, Juan ^[2]
  1. [1] Research Centre on Industrial and Public Economics Department of Economics Rovira i Virgili University
  2. [2] Department of Applied Economics II and Estructura de Recerca Interdisciplinar Comportament Econòmic i Social
• Localización: The Stata journal, ISSN 1536-867X, Vol. 16, Nº. 4, 2016, págs. 900-916
• Idioma: inglés
• DOI: 10.1177/1536867x1601600406
• Enlaces
  • Texto completo
• Resumen

  • Authors present a new e-class command, acfest, that implements the method of Ackerberg, Caves, and Frazer (2015, Econometrica 83: 2411–2451) to estimate production functions. This method deals with the functional dependence problems that may arise in the methods proposed by Olley and Pakes (1996, Econometrica 64: 1263–1297) and, particularly, Levinsohn and Petrin (2003, Review of Economic Studies 70: 317–341) (implemented in Stata by Yasar, Raciborski, and Poi [2008, Stata Journal 8: 221–231] and Petrin, Poi, and Levinsohn [2004, Stata Journal 4: 113–123], respectively). In particular, the acfest command yields (nonlinear, robust) generalized method of moments estimates using a Mata function and two specification tests (Wald and Sargan–Hansen). After estimation, predict provides the estimated productivity of the firms in the sample.