1983
DOI: 10.2307/1912288
|View full text |Cite
|
Sign up to set email alerts
|

Identification in Linear Simultaneous Equations Models with Covariance Restrictions: An Instrumental Variables Interpretation

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

2
59
1

Year Published

1987
1987
2021
2021

Publication Types

Select...
4
4

Relationship

1
7

Authors

Journals

citations
Cited by 58 publications
(62 citation statements)
references
References 10 publications
2
59
1
Order By: Relevance
“…4-5;Goldberger I 991, 361-62;Johnston 1963, 248-49;Maddala 1977, 226-28;Malinvaud 1966, 528-38; see also Heise 1975, 181-82;Hanushek and Jackson 1977, 271-76. ) For a general discussion, see Hausman and Taylor (1983) and Hausman,Newey,and Taylor (I 987). For a political science application, see Erikson I 982. Meanwhile, unmeasured variables that might affect spending directly are not likely to affect the vote directly.…”
Section: The Uncorrelated Errors Solutionmentioning
confidence: 99%
“…4-5;Goldberger I 991, 361-62;Johnston 1963, 248-49;Maddala 1977, 226-28;Malinvaud 1966, 528-38; see also Heise 1975, 181-82;Hanushek and Jackson 1977, 271-76. ) For a general discussion, see Hausman and Taylor (1983) and Hausman,Newey,and Taylor (I 987). For a political science application, see Erikson I 982. Meanwhile, unmeasured variables that might affect spending directly are not likely to affect the vote directly.…”
Section: The Uncorrelated Errors Solutionmentioning
confidence: 99%
“…Although this is a strong assumption, the fact that we use two-way fixed effects reduces the problem considerably. Under this restriction, Hausman and Taylor (1983) have shown that a valid instrument for C it in equation (5b) is the estimated residuals ε 1 it obtained from a 2SLS regression on equation (5a) with W it as instruments for opium. Due to the restriction on the covariance matrix of the errors, ε 1 it will be correlated with C it (from equation (5a)), but will be uncorrelated with ε 2 it , i.e.…”
Section: Correcting Simultaneity Bias By Instrumentingmentioning
confidence: 99%
“…We see that bothb andβ are positive and highly significant. In columns (3) and (4) we control for the simultaneity bias by estimating the parameters using the Hausman and Taylor (1983) approach. We here see that the effect of conflict on opium (β) is still highly significant, and in numerical value almost equal to the ones presented in Table 3.…”
Section: Correcting Simultaneity Bias By Instrumentingmentioning
confidence: 99%
See 1 more Smart Citation
“…To analyze the limiting distribution of the A3SLS estimator in the general case, it will be helpful to rewrite the FIML estimator from equations (3-13) in an instrumental variables form, following the derivation in Hausman (1975) for the unrestricted^case. Since we have one equation in equation (3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13) for each unknown 6, we take each equation from the first block of the gradient (3' 13) corresponding to the system of equations (3*1 …”
Section: Fiml Estimation In the M-equation Casementioning
confidence: 99%