IVoverid: Test of overidentifying restrictions

In kolesarm/ManyIV: Inference in Instrumental Variables Models with Many Instruments

Description

Report the Sargan and modified Cragg-Donald test statistics and p-values for testing of overidentifying restrictions, assuming homoskedasticity of the reduced form. The Sargan test is valid under few instruments. The Modified Cragg-Donald test (Modified-CD) corresponds to a test due to Cragg and Donald (1993), with a modified critical value. The modification was suggested in Kolesár (2018) to make it robust to many instruments and many exogenous regressors.

Usage

IVoverid(r)

Arguments

r	An object of class RDResults

References

Kolesár, Michal. Minimum Distance Approach to Inference with Many Instruments.” Journal of Econometrics 204 (1): 86–100. doi: 10.1016/j.jeconom.2018.01.004.

Cragg, John G., and Stephen G. Donald. 1993. "Testing Identifiability and Specification in Instrumental Variable Models." Econometric Theory 9 (2): 222–40. doi: 10.1017/S0266466600007519.

Sargan, John Denis. 1958. "The Estimation of Economic Relationships Using Instrumental Variables." Econometrica 26 (3): 393–415. doi: 10.2307/1907619.

Examples

r1 <- IVreg(lwage~education+black+married | as.factor(qob), data=ak80,
            inference="standard")
IVoverid(r1)

kolesarm/ManyIV documentation built on March 13, 2021, 6:31 a.m.