The empirical evidence that types affect productivity and correctness

This was just posted a few days ago -- here's the discussion for anyone interested:

https://news.ycombinator.com/item?id=8594769