kinkyreg: Instrument-free inference for linear regression models with endogenous regressors

In models with endogenous regressors, a standard regression approach is to exploit just-identifying or overidentifying orthogonality conditions by using instrumental variables. just-identified models, the identifying assumptions cannot be tested without imposition of other nontestable assumptions. While formal testing restrictions possible, its interpretation still hinges on validity an initial set untestable conditions. We present kinkyreg command for kinky least-squares inference, which adopts alternative identification. By exploiting nonorthogonality in form bounds admissible degree endogeneity, feasible test procedures can constructed that do not require The confidence bands more informative than intervals obtained from instrumental-variables estimation, especially when instruments are weak. Moreover, facilitates sensitivity analysis inference. particular, it allows user assess previously exclusion restrictions. Further instrument-free tests include linear hypotheses, functional form, heteroskedasticity, and serial correlation tests.