Working Paper: NBER ID: w26063
Authors: Anmol Bhandari; Serdar Birinci; Ellen McGrattan; Kurt See
Abstract: This paper examines the reliability of survey data on business incomes, valuations, and rates of return, which are key inputs for studies of wealth inequality and entrepreneurial choice. We compare survey responses of business owners with available data from administrative tax records, brokered private business sales, and publicly traded company filings and document problems due to nonrepresentative samples and measurement errors across several surveys, subsamples, and years. We find that the discrepancies are economically relevant for the statistics of interest. We investigate reasons for these discrepancies and propose corrections for future survey designs.
Keywords: Survey Data; Business Incomes; Valuations; Rates of Return; Wealth Inequality
JEL Codes: C83; E22; H25
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| survey data (C83) | overstate aggregate business incomes (E25) |
| survey data (C83) | overstate business incomes per tax return (H32) |
| discrepancies in survey data (C83) | systematic overstatements of business incomes (M41) |
| underrepresentation of low-income business owners (N86) | overstatement of business incomes (M41) |
| misreporting of losses (G33) | overstatement of business incomes (M41) |
| survey designs (C83) | fail to capture true economic activity (E26) |
| survey methodologies (C83) | upward bias in estimates of business rates of return (G31) |