Try some statistics

Before you can conclude "soaring rates", you need to know what the margin of error is. The sample size is 73600 people, and the rape rate is one per thousand in 2007, compared to 0.8 per thousand in 2005. 
 
This means that the standard error is about 1/8500 or 0.116 per thousand. Given random sampling, you can expect 95% confidence that the true value is within two standard errors, which means the true rate is somewhere between about 0.7 per thousand and 1.2 per thousand. 
 
Given that the 2005 rape rate falls within the margin of error of the 2007 rate, I don't think it's valid to conclude "soaring rates". Rape is so rare that a survey of 73,600 people is just not enough to detect a change of this magnitude. 
 
(Using a maximum margin of error, the 90% confidence margin of error for a survey of this size is 3 per thousand, and 95% is 3.6 per thousand.) 
 
Even the survey itself reports that the rape rate change is not statistically significant, whereas the declines in aggravated assault and property crimes are significant. These offenses are more common than rape, and so the rate changes can be detected by a study with this sample size.

Guest