I am using census data which I read is typically clustered by neighborhood. Would it be acceptable to use clustered standard errors even though it cross-sectional data and not panel data? Thank you.