I imagine these surveys just tend to omit the required consequences of each choice. E.g. asked "Do you want to pay less in tax or more in tax?", most people wou

I imagine these surveys just tend to omit the required consequences of each choice. E.g. asked "Do you want to pay less in tax or more in tax?", most people would likely answer "less", but if you outlined all the things they'd lose with lower taxes, a lot of those votes would flip. In this case, I imagine people said that they'd prefer more relevant ads (I know I do), but then T-Mobile used that as justification to do all sorts of other things that weren't asked about.