A survey uses weights to align its respondent age distribution with a reliable population age distribution. Respondents systematically underreport their weekly expenses. What can be concluded about weighting's ability to fix this problem?