2025-02-16-美联储-揭示调查准确性——SHED与人口普查局调查结果的比较（英）_19页_273kb
报告摘要
Summary of SHED and Census Bureau Survey Comparison
Report Title: Shedding Light on Survey Accuracy—A Comparison between SHED and Census Bureau Survey Results
Authors: Kabir Dasgupta, Fatimah Shaalan, Mike Zabek
Date: February 2025
Institution: Federal Reserve Board
Research Goal:
To evaluate the accuracy and consistency of the Survey of Household Economics and Decisionmaking (SHED) by comparing its aggregate results with the Current Population Survey (CPS) and American Community Survey (ACS) from the U.S. Census Bureau.
Key Findings
- Overall Comparability
- SHED aggregate statistics generally align well with nationally representative surveys like CPS and ACS, especially for identical survey topics.
- Question and methodology differences can lead to minor variations.
- Detailed Comparisons:
| Domain | Key Differences |
|---|---|
| Demographics | SHED demographic distributions match CPS and ACS after weighting; gaps of up to 4 percentage points in educational attainment and homeownership rates. |
| Employment | Employment rates and self-employment match CPS with slight variations (up to 2%). Differences in recall period (e.g., month vs week) affect part-time employment estimates. |
| Homeownership Status | SHED reports higher shares of "neither owning nor renting" due to narrower definitions centered on the survey participant only. |
| Health Insurance Coverage | Insurance coverage rates are highly consistent across surveys, except for slight discrepancies in employer-provided insurance (SHED: 56%; CPS: 53%). |
| Food Sufficiency | SHED reports higher monthly food security (93%) compared with the weekly measure from the Census Household Pulse Survey (HPS: 88%), reflecting recall period differences. |
| Bank Account Ownership | SHED's narrower definition of participants (only the respondent and partner) results in a lower bank ownership rate (94%) vs. the FDIC (96%); disparities vary by age and race. |
Methodological Notes
- SHED employs an online sampling method via Ipsos's KnowledgePanel, while CPS and ACS incorporate mixed-mode sampling (phone, mail, internet) with larger sample sizes.
- Variability in survey methodology (timing, mode of interview, sample weighting) can amplify differences even for similar questions.
Concluding Remarks
The SHED remains a credible source for economic and financial data but is subject to discrepancies arising from subtle differences in survey design and definition. The report recommends further research to explore the underlying causes of these variations and the temporal dynamics of survey responses.
试读结束，高清完整版pdf/doc/ppt,请点下载
