A Monte Carlo study of bias corrections for panel probit models

Alexander, Blair; Breunig, Robert

doi:10.1080/00949655.2014.994516

Cited by 5 publications

(10 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In line with other studies (e.g. Alexander & Breunig, 2016, and references therein), the ML estimates of the average marginal effect are less biased than the β coefficient. In contrast, the BRFE estimates are more biased than their corresponding β estimates.…”

Section: Results For Quantities Other Than ̂ Isupporting

confidence: 90%

“…The case of ρ = 0 corresponds to Panel (A) of Figure 3 shown in Section 3. F I G U R E A 2 Estimated versus true distributions of i for the baseline DGP employed by Heckman (1991), Newey (2004), Fernández-Val (2009), and Alexander and Breunig (2016).…”

Section: Discussionmentioning

confidence: 99%

“…Notes: Graphs show average estimates of 1 , …, 100 over 500 replications against their true values. 2009) and Alexander and Breunig (2016). Compared to our baseline with normally distributed i , this DGP features a regressor that is serially correlated and standard normally distributed i .…”

Section: F I G U R E 5 Mean Of Estimated Individual Effects By True I ( I ∼ Uniform)mentioning

confidence: 99%

“…, Tables 1 (upper panel, statistics for ̂ ) and 2 (lower panel, statistics for marginal effects of x it -indicated as ME(x)). The remaining columns are reproduced from Alexander and Breunig (2016) Tables 1-2). The true value of β is 1.…”

Section: Appendix B Details On the Brfe Estimator For General Binary Response Panel Modelsmentioning

confidence: 99%

See 3 more Smart Citations

Predicting Individual Effects in Fixed Effects Panel Probit Models

Kunz

Staub

Winkelmann

2021

Journal of the Royal Statistical Society Series A: Statistics in Society

View full text Add to dashboard Cite

Many applied settings in empirical economics require estimation of a large number of individual effects, like teacher effects or location effects; in health economics, prominent examples include patient effects, doctor effects or hospital effects. Increasingly, these effects are the object of interest of the estimation, and predicted effects are often used for further descriptive and regression analyses. To avoid imposing distributional assumptions on these effects, they are typically estimated via fixed effects methods. In short panels, the conventional maximum likelihood estimator for fixed effects binary response models provides poor estimates of these individual effects since the finite sample bias is typically substantial. We present a bias-reduced fixed effects estimator that provides better estimates of the individual effects in these models by removing the first-order asymptotic bias. An additional, practical advantage of the estimator is that it provides finite predictions for all individual effects in the sample, including those for which the corresponding dependent variable has identical outcomes in all time periods over time (either all zeros or ones); for these, the maximum likelihood prediction is infinite. We illustrate the approach in simulation experiments and in an application to health care utilization.

show abstract

Section: Results For Quantities Other Than ̂ Isupporting

confidence: 90%

Section: Discussionmentioning

confidence: 99%

Section: F I G U R E 5 Mean Of Estimated Individual Effects By True I ( I ∼ Uniform)mentioning

confidence: 99%

Section: Appendix B Details On the Brfe Estimator For General Binary Response Panel Modelsmentioning

confidence: 99%

See 2 more Smart Citations

Predicting Individual Effects in Fixed Effects Panel Probit Models

Kunz

Staub

Winkelmann

2021

Journal of the Royal Statistical Society Series A: Statistics in Society

View full text Add to dashboard Cite

show abstract

“…In the our example of the probit model using α i =α i +x i γ, these expected moments are equal to (see, e.g., Alexander and Breunig, 2016):…”

Section: Appendix a Details On Derivation Of The Bias In The Probit Mmentioning

confidence: 99%

Estimating Fixed Effects: Perfect Prediction and Bias in Binary Response Panel Models, with an Application to the Hospital Readmissions Reduction Program

2017

View full text Add to dashboard Cite

The Robustness of Conditional Logit for Binary Response Panel Data Models with Serial Correlation

Kwak

Martin

Wooldridge

2021

Journal of Econometric Methods

View full text Add to dashboard Cite

We examine the conditional logit estimator for binary panel data models with unobserved heterogeneity. A key assumption used to derive the conditional logit estimator is conditional serial independence (CI), which is problematic when the underlying innovations are serially correlated. A Monte Carlo experiment suggests that the conditional logit estimator is not robust to violation of the CI assumption. We find that higher persistence and smaller time dimension both increase the magnitude of the bias in slope parameter estimates. We also compare conditional logit to unconditional logit, bias corrected unconditional logit, and pooled correlated random effects logit.

show abstract

A Monte Carlo study of bias corrections for panel probit models

Cited by 5 publications

References 14 publications

Predicting Individual Effects in Fixed Effects Panel Probit Models

Predicting Individual Effects in Fixed Effects Panel Probit Models

Estimating Fixed Effects: Perfect Prediction and Bias in Binary Response Panel Models, with an Application to the Hospital Readmissions Reduction Program

The Robustness of Conditional Logit for Binary Response Panel Data Models with Serial Correlation

Contact Info

Product

Resources

About