In computer science, particularly in theoretical computer science we can really judge the quality of a student's research in a more objective way than a paper in say sociology. It does not matter where you got your Ph.D., if you didn't have strong results as a graduate student you won't find employment at a top university. If one does have solid results from a lesser known school one can find a top academic job though admittedly this happens less often.
I won't deny a strong correlation between where you got your Ph.D. and where you get employed: The best undergraduates choose the best graduate schools. At the strong schools you can usually find stronger faculty as potential advisors, stronger fellow students to work with and push you and you will have letter writers with stronger credentials when you do enter the job market. There is a social network among stronger schools that we cannot ignore. And some places, particularly those without experts in an area, will put too much emphasis on a student's department.
But we can better measure the quality and importance of a student's research and impact and that always plays the most critical role in whether one offers that person a position in a particular field.