Weighted means are the same as the unweighted means?

When I weight the data on a project, and look at the frequency weighted and unweighted, the weighted data is what I would expect it to be. However when I look at the means for the five cities in the survey, the weighted data is the same for the five cities as unweighted, but the total is different. I would have thought that when the frequencies all change following weighting, the means would not be the same weighted and unweighted. Hope somebody can understand what I mean and throw some light on it. Sample data attached. Thanks.