Science and private "big data"

When research is conducted using data from private sources, should the underlying data be made public?  See the New York Times article at

http://www.nytimes.com/2012/05/22/science/big-data-troves-stay-forbidden-to-social-scientists.html?smid=pl-share