Cost of a false positive here would have been a nurse who was inconvenienced. Cost of a false negative is at a minimum the time and expense of contact-tracking 132 people who flew with her, the cost to the airline of taking the plane out of service (which they've already done), and a severe drop of public confidence in the CDC. The pane made 5 more flights before it was taken out of service: http://www.latimes.com/business/la-fi-frontier-airline-ebola...

Now, the chances that anyone else on the same flight got infected is very very low, close to zero. That's even more true for the people who took other flights on the same plane, even the people who were sitting in the same seat that she sat in - though it's non-zero, as I doubt the airline disinfects the armrests and fold-down trays after every single flight. But just doing the contact tracking and being prepared to field telephone and email inquiries from all the other people who traveled on that plane with and after the nurse did is going to consume a bunch of CDC resources that could be better spent elsewhere.

Meanwhile, it seems that staff at the hospital were not wearing adequate protective gear for the first 2 days after Duncan was admitted until they got a positive blood test for Ebola: http://thescoopblog.dallasnews.com/2014/10/presbyterian-work...

False positives are bad but IMHO they're a lot cheaper than false negatives. When dealing with the limited subset of people who had contact with known Ebola carrier at the Texas hospital (a total of 76 medical personnel AFAIK), the slight economic cost of limiting travel and monitoring all of them for a few weeks seems low compared to the new situation.