The paper even admits that it probably exaggerates the effects of grants and loans since their model doesn't incorporate private and public schools with different funding characteristics or competition[1].

[1]: http://www.nber.org/chapters/c13711.pdf