DWAS Productive Ownership Benchmark Methodology v1.0: 2022 Survey of Consumer Finances Reference Implementation and Replication Archive
收藏资源简介:
DWAS Productive Ownership Benchmark Methodology v1.0 is the first frozen methodological vintage for constructing a classified U.S. household productive-ownership benchmark from the Federal Reserve’s Survey of Consumer Finances (SCF). The archive contains the governing methodology, supporting 2022 SCF variable and property/business crosswalks, the reference Python implementation, five-implicate statistics, 999 bootstrap-replicate statistics, combined survey-uncertainty results, the original compressed computational archive, and cryptographic manifests supporting reproducibility and provenance. Applied to the 2022 SCF, Version 1.0 estimates an ownership-exclusion parameter of g = 0.37124, indicating that approximately 37.1 percent of SCF families had no positive net productive ownership under the classified benchmark. The estimated concentration parameter is α = 1.30368. Positive productive ownership is estimated at approximately $73.17 trillion. The top 10 percent holds approximately 87.90 percent, the top 1 percent 49.19 percent, and the top 0.1 percent 22.47 percent of measured positive productive ownership; the middle 40 percent holds approximately 12.03 percent and the bottom 50 percent approximately 0.0778 percent. The archive is designed to permit independent replication. Federal Reserve source files are not redistributed. The README identifies the exact 2022 SCF public-use inputs, byte counts, SHA-256 hashes, authoritative Federal Reserve sources, computational procedure, and expected results. Version 1.0 is a controlled methodological vintage rather than a claim of methodological finality. Material changes to classifications, estimators, fitting conventions, or uncertainty procedures should be released as explicitly versioned successors. External independent replication had not been completed at the time Version 1.0 was frozen. Documentation and DWAS-produced research outputs are licensed under CC BY 4.0. The Python implementation is licensed under the MIT License.



