Although test scores are not perfect measures of the success of education systems (or of the future success of student), internationally comparable data attract the attention of the media, policy makers , academics, and the public. Superficial analysis has shown that there is no direct relationship between per pupil spending and test scores. For example, countries like Finland and South Korea, which spend much less on education than the US and France, received higher scores on these tests. However, as we have learned by now, correlation does not always mean causation. We need a more careful analysis addressing problems like omitted variable bias and reverse causality. Numerous researchers have studied the factors influencing test scores by addressing some of those problems. Hanushek and Woessmann [2010] carefully surveyed these studies and concluded that there is not a significant cross country positive correlation between per student spending and test scores. This result holds even when other factors influencing educational success are included in the analysis . Including factors at the individual level, such as family background , at the school level, such as class size, and at the country level, such as accountability of educators, does not Change the conclusion that spending more does not improve test scores.