Please use this identifier to cite or link to this item: http://hdl.handle.net/10419/171835
Authors: 
Buck, Andrew J.
Lady, George M.
Year of Publication: 
2015
Citation: 
[Journal:] Econometrics [ISSN:] 2225-1146 [Volume:] 3 [Year:] 2015 [Issue:] 3 [Pages:] 466-493
Abstract: 
This paper shows that a qualitative analysis, i.e., an assessment of the consistency of a hypothesized sign pattern for structural arrays with the sign pattern of the estimated reduced form, can always provide decisive insight into a model's validity both in general and compared to other models. Qualitative analysis can show that it is impossible for some models to have generated the data used to estimate the reduced form, even though standard specification tests might show the model to be adequate. A partially specified structural hypothesis can be falsified by estimating as few as one reduced form equation. Zero restrictions in the structure can themselves be falsified. It is further shown how the information content of the hypothesized structural sign patterns can be measured using a commonly applied concept of statistical entropy. The lower the hypothesized structural sign pattern's entropy, the more a priori information it proposes about the sign pattern of the estimated reduced form. As an hypothesized structural sign pattern has a lower entropy, it is more subject to type 1 error and less subject to type 2 error. Three cases illustrate the approach taken here.
Subjects: 
verification
falsification
specification
qualitative analysis
entropy
inherited restrictions
JEL: 
C15
C18
C51
C52
Persistent Identifier of the first edition: 
Creative Commons License: 
http://creativecommons.org/licenses/by/4.0/
Document Type: 
Article
Social Media Mentions:

10



Files in This Item:
File
Size





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.