Once again, I don't understand your point. If by majority, you mean a majority of the electorate then no, that would not have been the sample size.
However, assuming that a survey is well-designed, a relatively small sample will allow conclusions be made about the larger population.