Using Stata 14 Serial Number
LINK === https://shurll.com/2t8mea
^ Note: Serial numbers without a filing date were excluded from the 2018 update. To obtain data on the excluded serial numbers use a prior version of the dataset and filter based on a missing filing date.
Data from the year 1995 to 2007 were used, for a panel of 14 EU member states: Austria, Belgium, Denmark, Finland, France, Germany, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain, Sweden and the United Kingdom. Greece was excluded for lack of data. Due to the inexistence of data prior to 1995 and subsequent to 2007 for some of the variables, the maximum time span was ascertained (1995-2007). Furthermore, because the remaining countries of EU27 only offer data from 2000 for some variables, we had to limit the study to EU15, except Greece. Otherwise, the actual period of thirteen years (1995-2007) would be only eight years (2000- 2007). Although the number of observations is not exactly the same for all countries, missing values are few, isolated, and purely random. Therefore, we can apply the estimators in our unbalanced panel without causing inconsistency in these estimators.
the number of countries is larger than the number of periods. This estimator turns out to be adequate both in thepresence of panel-level heteroskedasticity and contemporaneous correlation, and in finite cases performs better than the asymptotically efficient FGLS [33]. Moreover, in order to check the robustness of the results achieved with the PCSE estimator, we follow two options. The first one is to apply the common panel data estimators, RE and FE. Results accomplished with the PCSE estimator are robust if the other models, such as RE and FE estimators, return different results. In that case, it seems that there is inefficiency in coefficient estimation and biased standard errors, by using the common panel data estimators. The second one is to test various assumptions about the variance across countries and serial correlations. If the results remain in essence unchanged, then the option of using the PCSE estimator is strengthened.
Income and the price of gasoline contribute towards mitigating road transport CO2 emissions. On the other hand, population density and the average power of new diesel PCs registered have the opposite impact, i.e., they contribute to exacerbating those CO2 emissions. As far as dieselization is concerned, our findings are crucial to fully understanding this trend by showing that saving emissions from using diesel tends to be surpassed by the increased kms driven. Indeed, we show that a large share of new diesel PCs contributes to more road transport CO2 emissions. This result with a positive signal does not depend on the debatable assumption that diesel and gasoline PCs travel the same number of kms per year. 2b1af7f3a8