It’s based on a paper by a pair of Chicago economists, Jesse Shapiro and Matthew Gentzkow, who noticed that American cities were not all hooked up to television in the same year. Specifically, television started up after WWII, then stopped expanding between 1948 and 1952 for regulatory reasons. Comparing the later test scores of those who had access to TV during their early years to those who did not, they found the former group did noticeably better on these standardized measures.
In case you’re thinking that those who had TV first were richer, and thus their kids would do somewhat better – my first reaction – the statistics controls for that, so no dice there. Read the full article for more possible explanations. HT: MR.