It bothered me the first time you posted this and didn't say anything then, but I didn't imply equal numbers of respondents and I don't know why you think I did. Unless you have some kind of deeper knowledge about partisan alignment in the population, for all we know this is representative. The rankings are very consistent regardless of party identification in the responses across the board, party identification of respondent seems to be a weak factor in the rankings.