Working Paper: NBER ID: w10862
Authors: Susann Rohwedder; Steven J. Haider; Michael D. Hurd
Abstract: The Asset and Health Dynamics Among the Oldest Old (AHEAD) study shows a large increase in reported total wealth between 1993 and 1995. Such an increase is not found in other US household surveys around that period. This paper examines one source of this difference. We find that in AHEAD 1993 ownership rates of stocks, CDs, bonds, and checking and saving accounts were under-reported, resulting in under-measurement of wealth in 1993, and a substantial increase in wealth from 1993 to 1995. The explanation for the under-reporting is a combination of question sequence and wording in the AHEAD survey instrument.
Keywords: wealth; elderly; survey design; AHEAD; underreporting
JEL Codes: J14; C42; D91
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| survey design flaws (C83) | underreporting of asset ownership (D14) |
| underreporting of asset ownership (D14) | mismeasurement of total wealth (E21) |
| survey design flaws (C83) | inflated perception of wealth growth among the oldest old (E21) |
| phrasing of ownership questions (P26) | underreporting of asset ownership (D14) |
| reporting income from stocks and bonds (G12) | denial of ownership later in the asset section (G32) |
| denial of ownership (H13) | unusually high rates of ownership in subsequent waves (R21) |