This data is from the Delta Cost Project, a longitudinal study of college and university finance and administration, showing hundreds of institutional variables over time. This particular set shows the years 2000-2010, but there is no admissions data reported until 2003, which is itself interesting.
The race for more applications has been successful: Apps are up about 50% over that time, even though the high school population has not grown that much. Enrollment is up, too, but only by about 16%, which proves the old adage of mine that, "If they weren't born 18 years ago, they aren't going to enroll this fall."
You can look at this by region or by Carnegie Type: I clustered the myriad of Carnegie classifications into four groups. Although there are some variations, for the most part, the trend applies to all types, and all regions. Of interest is the drop in the Draw Rate, which is yield/admit rate, a proxy for market position. You can shift prestige around, but you can't seem to create or increase it.
The money spent to generate more applications, which generate paranoia among students, all in the pursuit of prestige, seems to have worked. The actual cost of doing so has yet to be determined.