Cluster Estimated Standard Errors (CESE)
vcovCESE(mod, cluster = NULL, type = NULL)
| mod | a model object. It can be the output of the functions |
|---|---|
| cluster | either a string vector with the name of the variables that will be used to cluster the standard errors, or a formula - e.g., ~ rhs, with a summation of the variables that will be used to cluster the standard errors replacing the |
| type | string with either |
The function returns a variance-covariace matrix of the coefficient estimates using the Cluster Estimated Standard Error (CESE) method.
Jackson, John (2019) Corrected Standard Errors with Clustered Data. Political Analysis.
Hayes, A. F., & Cai, L., (2007) Using heteroskedasticity-consistent standard error estimators in ols regression: an introduction and software implementation, Behavior research methods, 39(4), 709–722.
Davidson, R., & MacKinnon, J. G., (2004) Econometric theory and methods: Oxford University Press New York.
mod = lm(enep ~ enpc + fapres + enpcfapres + proximity + eneg + logmag + logmag_eneg , data=dcese) ## -------------------------------------- ## Getting the variance covariance matrix ## -------------------------------------- ## Original variance-covariance matrix (no clustered std. errors) vcov(mod)#> (Intercept) enpc fapres enpcfapres proximity #> (Intercept) 0.34193367 -0.080109404 -6.498717e-02 0.0227605333 -0.0416369373 #> enpc -0.08010940 0.035696550 2.401318e-02 -0.0102825206 0.0059204296 #> fapres -0.06498717 0.024013177 2.734250e-02 -0.0090017922 -0.0004345205 #> enpcfapres 0.02276053 -0.010282521 -9.001792e-03 0.0036429747 -0.0014388263 #> proximity -0.04163694 0.005920430 -4.345205e-04 -0.0014388263 0.0776196264 #> eneg -0.03580496 -0.001476759 -2.517854e-03 0.0007025382 -0.0039084482 #> logmag -0.05448255 -0.006980873 1.742043e-04 0.0021399533 -0.0023836036 #> logmag_eneg 0.02532042 0.001832590 -7.512686e-05 -0.0007721436 -0.0009085951 #> eneg logmag logmag_eneg #> (Intercept) -0.0358049582 -0.0544825503 2.532042e-02 #> enpc -0.0014767587 -0.0069808731 1.832590e-03 #> fapres -0.0025178544 0.0001742043 -7.512686e-05 #> enpcfapres 0.0007025382 0.0021399533 -7.721436e-04 #> proximity -0.0039084482 -0.0023836036 -9.085951e-04 #> eneg 0.0218856408 0.0222887158 -1.190289e-02 #> logmag 0.0222887158 0.0606796325 -2.995518e-02 #> logmag_eneg -0.0119028856 -0.0299551838 1.778317e-02## Variance-covariance matrix using CRSE (sandwish package) ## sandwich::vcovCL(mod, cluster = ~ country) ## sandwich::vcovCL(mod, cluster = ~ country, type="HC3") ## Variance-covariance matrix using CESE ceser::vcovCESE(mod, cluster = ~ country)#> (Intercept) enpc fapres enpcfapres proximity #> (Intercept) 1.51221650 -0.3374255708 -0.308548972 0.087864892 -0.082214926 #> enpc -0.33742557 0.1187065304 0.099198869 -0.033556819 -0.003174936 #> fapres -0.30854897 0.0991988687 0.135525330 -0.036883799 -0.016929595 #> enpcfapres 0.08786489 -0.0335568193 -0.036883799 0.012012890 0.003051838 #> proximity -0.08221493 -0.0031749364 -0.016929595 0.003051838 0.132039888 #> eneg -0.08262488 0.0026806954 -0.006697128 0.001033777 -0.005366609 #> logmag -0.21216040 0.0009545731 0.006332777 0.003599288 0.009265360 #> logmag_eneg 0.07928461 -0.0055220287 -0.010890411 0.000813923 0.004237610 #> eneg logmag logmag_eneg #> (Intercept) -0.082624881 -0.2121603981 0.079284613 #> enpc 0.002680695 0.0009545731 -0.005522029 #> fapres -0.006697128 0.0063327767 -0.010890411 #> enpcfapres 0.001033777 0.0035992881 0.000813923 #> proximity -0.005366609 0.0092653603 0.004237610 #> eneg 0.037272375 0.0455315878 -0.021837494 #> logmag 0.045531588 0.2123302156 -0.099187086 #> logmag_eneg -0.021837494 -0.0991870859 0.057389251#> (Intercept) enpc fapres enpcfapres proximity #> (Intercept) 1.59804215 -0.3565889644 -0.326044522 0.0928614018 -0.086958688 #> enpc -0.35658896 0.1254735451 0.104834222 -0.0354704134 -0.003332760 #> fapres -0.32604452 0.1048342220 0.143205927 -0.0389793642 -0.017878997 #> enpcfapres 0.09286140 -0.0354704134 -0.038979364 0.0126978025 0.003217871 #> proximity -0.08695869 -0.0033327597 -0.017878997 0.0032178709 0.139694748 #> eneg -0.08737194 0.0028258190 -0.007080605 0.0010940285 -0.005680272 #> logmag -0.22422347 0.0009844983 0.006687954 0.0038080340 0.009776123 #> logmag_eneg 0.08380931 -0.0058250477 -0.011500450 0.0008568949 0.004471824 #> eneg logmag logmag_eneg #> (Intercept) -0.087371945 -0.2242234723 0.0838093142 #> enpc 0.002825819 0.0009844983 -0.0058250477 #> fapres -0.007080605 0.0066879544 -0.0115004496 #> enpcfapres 0.001094029 0.0038080340 0.0008568949 #> proximity -0.005680272 0.0097761226 0.0044718243 #> eneg 0.039433115 0.0481561426 -0.0231003489 #> logmag 0.048156143 0.2244236813 -0.1048418029 #> logmag_eneg -0.023100349 -0.1048418029 0.0606625677#> #> Call: #> lm(formula = enep ~ enpc + fapres + enpcfapres + proximity + #> eneg + logmag + logmag_eneg, data = dcese) #> #> Residuals: #> Min 1Q Median 3Q Max #> -3.5592 -0.8192 -0.3615 0.3769 9.0393 #> #> Coefficients: #> Estimate Std. Error t value Pr(>|t|) #> (Intercept) 2.70431 0.58475 4.625 5.65e-06 *** #> enpc 0.30399 0.18894 1.609 0.108711 #> fapres -0.61179 0.16536 -3.700 0.000258 *** #> enpcfapres 0.20779 0.06036 3.443 0.000660 *** #> proximity -0.02243 0.27860 -0.081 0.935890 #> eneg -0.06566 0.14794 -0.444 0.657483 #> logmag -0.18146 0.24633 -0.737 0.461926 #> logmag_eneg 0.36051 0.13335 2.703 0.007266 ** #> --- #> Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 #> #> Residual standard error: 1.653 on 291 degrees of freedom #> Multiple R-squared: 0.3784, Adjusted R-squared: 0.3634 #> F-statistic: 25.31 on 7 and 291 DF, p-value: < 2.2e-16 #>## summary table using CRSE (sandwich package) ## lmtest::coeftest(mod, vcov = sandwich::vcovCL, cluster = ~ country) ## summary using CESE lmtest::coeftest(mod, vcov = ceser::vcovCESE, cluster = ~ country, type='HC3')#> #> t test of coefficients: #> #> Estimate Std. Error t value Pr(>|t|) #> (Intercept) 2.704309 1.264137 2.1393 0.03325 * #> enpc 0.303988 0.354222 0.8582 0.39150 #> fapres -0.611787 0.378426 -1.6167 0.10703 #> enpcfapres 0.207791 0.112685 1.8440 0.06620 . #> proximity -0.022429 0.373758 -0.0600 0.95219 #> eneg -0.065662 0.198578 -0.3307 0.74114 #> logmag -0.181462 0.473734 -0.3830 0.70197 #> logmag_eneg 0.360513 0.246298 1.4637 0.14435 #> --- #> Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 #>