Testing the Racial Preference Theory in College Admissions

According to affirmative action proponents, racial preferences in college admissions are desirable because greater diversity leads to a "superior education experience."

California and Michigan and Washington have banned the use of race in college admissions. As a result the admission rate for blacks has dropped.

If affirmative action proponents are correct, then the education experience at those state universities has been denigrated. I would then have expected to see these things happen:

1. Applications to these schools should be declining as parents and students react to the reduced value of education at these schools.
2. Transfers out of UW, UM, UC to more racially diverse schools should have accelerated.
3. Employers would be more reluctant to hire graduates of these school than those of more racially diverse schools.

Has anyone observed or documented any of this? After all, the root of the scientific method (which universities say they hold dear) is that the validity of a hypothesis is whether we observe the results predicted by that hypothesis.