The study compared children enrolled versus forced to delay entrance according to an arbitrary date.[1] Socioeconomic status has nothing to do with it. Further

The study compared children enrolled versus forced to delay entrance according to an arbitrary date.[1] Socioeconomic status has nothing to do with it. Further, the study explicitly claims causal relationships, rather than merely correlative. I'm guessing as a result of the kind of fancy-pants regression analysis they used. [2] [1] "children who were born a certain number of days before the Dec. 31 enrollment cut-off and those born after the cut-off" http://www.nber.org/papers/w21610 http://www.nber.org/papers/w21610 [2] "We estimate the causal effects of delayed school enrollment using a "fuzzy" regression-discontinuity design" http://www.nber.org/papers/w21610 http://www.nber.org/papers/w21610