Survey statistics

A researcher cannot observe outcomes for nonrespondents but compares people who responded after the first invitation with people who responded only after several follow-ups. Why might this comparison be useful?