This paper provides a new perspective on the varying efficiency levels among different education systems by comparing data from six Sub-Saharan countries and fourteen Latin-American countries. When comparing the effect of socio-economic status (SES) on education across countries, researchers have always been faced with a trade-off between the accuracy of the SES measure within countries and the comparability of the measure across countries. This has often caused measures of SES to be incorrectly used to compare relative wealth across different countries and contexts. This research paper sets forth a new methodology to adjust the traditional measures of SES and make it more comparable across countries and surveys. Furthermore, the comparable SES measure is applied to compare children in equally impoverished circumstances across countries, sub-samples and datasets to enable the identification of the most disadvantaged children across the world more accurately. More specifically this method will be applied to the SACMEQ (sub-Saharan Africa) and SERCE (Latin America) education datasets and compare the educational outcomes of those students living under the $2 a day poverty line. Most strikingly, the comparison shows that Ugandan and Mozambican children living under the $2 a day poverty line achieve much higher educational outcomes than similarly poor children in middle-income countries such as South Africa and the Dominican Republic.