There is a discussion in Economist's view on the long range effectis of kindergarten following a recent study of Raj Chetty and others Do Kindergarten Teachers Matter?. There is an earlier such teport on Perry Preschool program Early Lessons by Emily Hanford and a report by economists here.
On the otherhand, there have been apparently contrasting studies like Benezet experiments here. It would be interesting to see which components worked. It seems to me that emphasis on cognitive development and caring environment rather than rote learning might have played a role in the Perry Preschool experiments.
Some earlier posts on the topic here and here.
P.S.Rajib Khan starts analysing Investing in a nanny state for social returns.