Identifying Regression Parameters When Variables are Measured with Error
The paper proposes an approach for identifying and estimating the economic parameters of interest when all the variables are measured with errors and these are correlated. Two propositions show how the parameters of interest and the bias are identified. Three Monte Carlo simulations illustrate the results. The empirical application estimates returns to scale and technological progress in US manufacturing sectors. The results can be linked to previous works in the literature to demonstrate the ambiguous bias in least squares estimates of returns to scale parameters and to compare estimates of trends in technological change using two alternative identification approaches.
|Date of creation:||08 Apr 2016|
|Contact details of provider:|| Postal: St. Lucia, Qld. 4072|
Phone: +61 7 3365 6570
Fax: +61 7 3365 7299
Web page: http://www.uq.edu.au/economics/
More information through EDIRC
When requesting a correction, please mention this item's handle: RePEc:qld:uq2004:557. See general information about how to correct material in RePEc.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (SOE IT)
If references are entirely missing, you can add them using this form.