I'm having trouble finding it, but there was a recent study on people who were just above vs just below the hard SAT cutoff for their state school (presuming that these tiny differences - less than 20 SAT points - are fairly meaningless) and found that it does appear to be causal.