Working Paper: NBER ID: w20972
Authors: John Ameriks; Andrew Caplin; Minjoon Lee; Matthew D. Shapiro; Christopher Tonetti
Abstract: Wealth, though crucial for modeling economic behavior and understanding well-being, is difficult to measure in surveys. This paper introduces a new, comprehensive account-by-account approach for eliciting asset holding. This approach is implemented in the Vanguard Research Initiative, a panel of wealthholders designed to yield high-quality measurements for a large sample of older Americans with significant financial assets. Because survey responses are linked to administrative account balances, this paper can show that the approach yields precise, unbiased estimates. Having accurate and dense data on the wealth of wealthholders provides sharper inferences on wealth management behavior as well as on relationships between wealth and economic behavior than is possible in leading datasets.
Keywords: Wealth Measurement; Economic Behavior; Retirement Planning
JEL Codes: D91; E21; H31; J14
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| comprehensive account-by-account approach (H60) | unbiased estimates of wealth (D31) |
| correction mechanism (D51) | accuracy of reported wealth (G51) |
| ample observations of wealthholders (D30) | understanding of behavioral relationships (D91) |
| improved measures of wealth (E01) | insights into behaviors like saving and portfolio choice (D14) |