Introduction | The Mroz baseline | LIML, Fuller, and k-class estimation | LIML | Fuller modification | k-class estimation | COVIV | Choosing among the estimators | The Klein consumption function | LIML with COVIV equals CUE | Instrument-validity tests | Overidentification | Orthogonality (the C-statistic) | Endogeneity of a regressor | Testing an included regressor: the empty-endogenous form | Anderson-Rubin LIML overidentification | Weak-instrument-robust inference | The Stock-Wright S statistic | When instruments are genuinely weak | Redundant instruments | Building the S statistic by hand with b0 | What the package does and does not offer | Two-way clustering | The nlswork panel | One-way versus two-way clustering | Two-way clustering with instrumental variables | Reduced-form regression | Single-equation reduced form | System reduced form | Degrees-of-freedom adjustments | Estimator equivalences | Publication tables with modelsummary | A single model | Comparing estimators side by side | Exponentiated coefficients | Customizing output | Stata migration | References

Introduction | OLS baseline | Two-stage least squares | Understanding the diagnostics | Weak identification | Underidentification | Overidentification | Endogeneity test (C-statistic) | Anderson-Rubin test | First-stage diagnostics | Robust and clustered inference | Heteroskedasticity-robust standard errors | Cluster-robust standard errors | VCE option summary | Weighted estimation | Sampling weights (pweight) | Analytic weights (aweight) | The other weight types | Beyond 2SLS | Tidyverse workflows | Coefficient table | Model summary | Diagnostic tests as data | Augmented data | Stata migration quick reference | Estimation basics and accessors | Estimators | Variance estimation and clustering | Weights, tests, and other options | Intentional differences from Stata | References

Introduction | HAC and AC standard errors | Choosing a kernel | Automatic bandwidth | Kernel-based VCE with IV | Lags and differences in the formula | Panel variance estimators | Large N, small T: the Arellano-Bond firm panel | Small N, large T: the Grunfeld investment panel | Large N, moderate T: the NLS young-women panel | Fixed-effects panels: the Stock-Watson correction | When to use which | Generalized method of moments | When GMM helps | A HAC-GMM arc on the Stock-Watson data | Reusing the weighting matrix | Centering the moment conditions | Partialling out regressors | PSD corrections | Suppressing identification statistics | Estimator equivalences | References

Introduction | The Mroz baseline | LIML, Fuller, and k-class estimation | LIML | Fuller modification | k-class estimation | COVIV | Choosing among the estimators | The Klein consumption function | LIML with COVIV equals CUE | Instrument-validity tests | Overidentification | Orthogonality (the C-statistic) | Endogeneity of a regressor | Testing an included regressor: the empty-endogenous form | Anderson-Rubin LIML overidentification | Weak-instrument-robust inference | The Stock-Wright S statistic | When instruments are genuinely weak | Redundant instruments | Building the S statistic by hand with b0 | What the package does and does not offer | Two-way clustering | The nlswork panel | One-way versus two-way clustering | Two-way clustering with instrumental variables | Reduced-form regression | Single-equation reduced form | System reduced form | Degrees-of-freedom adjustments | Estimator equivalences | Publication tables with modelsummary | A single model | Comparing estimators side by side | Exponentiated coefficients | Customizing output | Stata migration | References

Introduction | OLS baseline | Two-stage least squares | Understanding the diagnostics | Weak identification | Underidentification | Overidentification | Endogeneity test (C-statistic) | Anderson-Rubin test | First-stage diagnostics | Robust and clustered inference | Heteroskedasticity-robust standard errors | Cluster-robust standard errors | VCE option summary | Weighted estimation | Sampling weights (pweight) | Analytic weights (aweight) | The other weight types | Beyond 2SLS | Tidyverse workflows | Coefficient table | Model summary | Diagnostic tests as data | Augmented data | Stata migration quick reference | Estimation basics and accessors | Estimators | Variance estimation and clustering | Weights, tests, and other options | Intentional differences from Stata | References

Introduction | HAC and AC standard errors | Choosing a kernel | Automatic bandwidth | Kernel-based VCE with IV | Lags and differences in the formula | Panel variance estimators | Large N, small T: the Arellano-Bond firm panel | Small N, large T: the Grunfeld investment panel | Large N, moderate T: the NLS young-women panel | Fixed-effects panels: the Stock-Watson correction | When to use which | Generalized method of moments | When GMM helps | A HAC-GMM arc on the Stock-Watson data | Reusing the weighting matrix | Centering the moment conditions | Partialling out regressors | PSD corrections | Suppressing identification statistics | Estimator equivalences | References