Please note-there are some issues with this piece: Social polling, during the third quarter, and much of the last quarter of the 20th Century, was still a rather new thing, and, for various reasons was often prone to accuracy issues, especially on rather complex and nuanced subjects such as interracial marriage; there is also the fact that interracial marriage, while somewhat controversial for many years, was not as polarizing of an issue, as, say, same-sex marriage. Furthermore, the sudden apparent jump between 48% approval in 1994-95 to 64% in 1997, very strongly suggests that a major revision in Gallup's survey methods took place right around this time. It's an interesting historical piece for sure, but any researchers out there should definitely be aware of the shortcomings of early social polling.