It's cheap. A physicist I work with says it was a cheap way to have clout in the 80's in a physics department. Fundamentally, paying for a couple of people to do math on paper all day is a lot cheaper than doing what they do at cern.