You are correct that the design of a survey can influence the result. A part of an answer to that is to ask the same question in different ways. A check on the possibility of bias is to publish, as part of a research paper, the questions which were asked.

In light of that, your assertion goes beyond doubt into denial. Your assumption that the researchers are biased liars is no better than an unquestioning belief that this report accurately presents the results.

The Social Sciences, done right, can employ powerful mathematical tools. The surveys can be repeated, so as to test the original results. Give a social scientist a thousand orphans, a hedge maze and enough cheese...

We distrust them, and yet so many of us are suckers for politicians.