Linear models for panel data
Random-effects regression for binary, ordinal, categorical, and count-dependent variables
- Probit
*
- Logistic regression
*
- Complementary log-log regression
*
- Ordered logistic regression
*
- Ordered probit regression
*
- Multinomial logistic regression
*
- Interval regression
![](/features/i/pdfdoc-icon.png)
- Tobit
![](/features/i/pdfdoc-icon.png)
- Poisson regression (Gaussian or gamma random-effects)
*
- Negative binomial regression
![](/features/i/pdfdoc-icon.png)
- Robust standard errors with (*) regressions
![](/features/i/pdfdoc-icon.png)
- Bayesian estimation
Conditional fixed-effects regression for binary, categorical, and count-dependent variables
Two-stage least-squares panel-data estimators ![](/features/i/pdfdoc-icon.png)
- Between-2SLS estimator
- Within-2SLS estimator
- Balestra–Varadharajan–Krishnakumar G2SLS estimator
- Baltagi EC2SLS estimator
- All with balanced or exogenously balanced panels
- Robust and cluster–robust standard errors
Random-effects regression with sample selection ![](/features/i/pdfdoc-icon.png)
Random-effects extended regression models
- Combine endogeneity, Heckman-style selection, and treatment effects
- Linear regression
![](/features/i/pdfdoc-icon.png)
- Interval regression, including tobit
![](/features/i/pdfdoc-icon.png)
- Probit regression
![](/features/i/pdfdoc-icon.png)
- Ordered probit regression
![](/features/i/pdfdoc-icon.png)
- Exogenous or endogenous treatment assignment
- Binary treatment–untreated/treated
- Ordinal treatment levels–0 doses, 1 dose, 2 doses, etc.
- Endogenous selection using probit or tobit
- All standard postestimation commands available, including predict and margins
Regressors correlated with individual-level effects ![](/features/i/pdfdoc-icon.png)
- Hausman–Taylor instrumental-variables estimators
- Amemiya–MaCurdy instrumental-variables estimators
- Robust and cluster–robust standard errors
Panel-corrected standard errors (PCSE) for linear cross-sectional models ![](/features/i/pdfdoc-icon.png)
Swamy’s random-coefficients regression ![](/features/i/pdfdoc-icon.png)
Stochastic frontier models ![](/features/i/pdfdoc-icon.png)
- Time-invariant model
- Time-varying decay model
- Battese–Coelli parameterization of time effects
- Estimates of technical efficiency and inefficiency
Specification tests
Panel-data unit-root tests ![](/features/i/pdfdoc-icon.png)
- Im–Pesaran–Shin
- Levin–Lin–Chu
- Hadri
- Breitung
- Fisher-type (combining p-values)
- Harris–Tzavalis
Summary statistics and tabulations
- Statistics within and between panels
![](/features/i/pdfdoc-icon.png)
- Pattern of panel participation
![](/features/i/pdfdoc-icon.png)
Panel-data line plots ![](/features/i/pdfdoc-icon.png)
- Graphs by panel
- Overlaid panels