I've always been a bit sceptical about these types of statistics, seems like it's a classification problem as well; the big established companies that traditionally take on lots of grads have many recruiting pathways, and don't always list many roles.

Smaller places just need "a guy" and don't always specify experience.