Figure 2 - uploaded by Ivan Fernandez-Val
Content may be subject to copyright.
Source publication
The Counterfactual package implements the estimation and inference methods of Chernozhukov, Fern\'andez-Val and Melly (2013) for counterfactual analysis. The counterfactual distributions considered are the result of changing either the marginal distribution of covariates related to the outcome variable of interest, or the conditional distribution o...
Similar publications
The gender pay gap measures the differences in the wages earned by female workers when compared to male workers. Accurately measuring the gender pay gap is an important exercise in order to assess how far we are from the ideal of equal pay for work of equal value. Determining the different sources of the gender pay gap is also very important in ord...
Citations
... We next present the results of the matched sample in the right panel of Table 3. Full details on the matching procedure can be found in Appendix A. 10 As expected, matching amounted to a reduction in sample size in each country. In Australia and Portugal, the coefficient on private is no longer statistically significant for maths, with the (Chen et al., 2016). (4) The y-axis denotes the learning profile of private students under the hypothetical case of having the same distribution of systemic characteristics of public school students. ...
We use measures of competitive pressure, administrative autonomy and staffing practices to explain the private-public performance difference in Australia, Portugal and Spain using the TALIS-PISA dataset. We employ OLS regression and counterfactual decomposition analysis on matched sub-samples. These school factors do not explain the overall private-public performance gap in the three countries except at the higher-end of the distribution. In other words, these factors appear to benefit only the high-performers in private schools in Australia and Spain. The results point to the potential limits of adopting private school practices for improving learning across the performance distribution especially for low-performing students.
