If we have got a world in which databases have no meaningful copyright protection in the sense that subscribers can take the data, leave the base and use their own algorithms to come out with a competing database, one possibility is no one would create databases in such a world, but I don't believe that one for a minute. We have got a voracious hunger for information in easily accessible form. People will step in to supply the demand. We don't give copyright protection to clothing designs and there's lots of clothing. We don't give copyright protection to recipes and there's plenty of food. We didn't give copyright protection to sound recordings until 1971 and there were lots of records. So I think there will still be databases, but I think that the rational database proprietor will go increasingly to these contractual kinds of protections in order to minimize the risk that any user of the database is going to be a competitor tomorrow. Indeed, the rational way to proceed is to use contract law and to use trade secrecy law to ensure that no use of data in your database occurs without collecting the appropriate fee.
Litman, Jessica D. "Symposium: Copyright Protection for Computer Databases, CD-ROMS and Factual Compilations." University of Dayton Law Journal 17, no. 2 (1992): 617-22. (Symposium Presentation. Work published when author not on Michigan Law faculty.)