Appendix F
Data sources
| Subject | Statistic or source |
Where used |
| Family 2 | Japan Payment Service Association, 27th Survey of Issuing Businesses [27] |
Part IV |
| Family 2 | Payment Services Act; FSA register of issuers of prepaid payment instruments |
Part IV |
| Family 5 | MHLW, Tabulation of Employment Placement Business Reports, FY2024 [28] |
Part IV |
| Family 5 | Jinzai Service General Site: office details and refund-scheme filings |
Section 15.3.5.0 |
| Fees | published fee schedules of each platform |
Chapter 17 |
| Industry classification | MOF, Financial Statements Statistics of Corporations by Industry, time series [29] |
Chapter 16 |
| Age at founding | White Paper on Small and Medium Enterprises 2014; Teikoku Databank survey of newly established corporations |
Section 8.6 |
| Comparison with sole proprietorships | SME Agency, Basic Survey of Small and Medium Enterprises, FY2020–2025 final reports |
Section 14.5.5 |
| Regulation of payment dates | Act on Ensuring Proper Transactions Involving Specified Entrusted Business Operators (in force November 2024) |
Section 15.7 |
| Vulnerabilities | published CVE counts, CISA KEV catalogue |
Chapter 9 |
On reproduction. The data obtained and the processing are placed under data/: raw/ holds the inputs as obtained and scripts/ the analysis. derived/ is intermediate data, regenerated by scripts/parse.py.
The calculations of Chapter 16 use ccc.py, e1.py, elast.py and others; the tests of Section 15.3.5.0 use parse_jinzai.py and predY.py. The correspondence is listed in data/README.md.
On handling. All statistics used are public. The Jinzai Service General Site alone, however, publishes the filings of individual offices rather than aggregates, and this work uses them as input to statistical analysis.
What this work presents is the result of the analysis, not the collection of per-firm data that was gathered. That collection is not reproduced, published or redistributed. The repository contains the processing of the analysis and the aggregates obtained from it. Per-firm data retrieved from that site — the search-result listings, the office detail pages and the refund-scheme filings alike — are not included.
What remains is only the list of licence and filing-receipt numbers (data/raw/jinzai/permits.tsv), which contains none of the filed values. Using it as a seed, running data/scripts/fetch_jinzai_detail.py reproduces the realized fee rates, turnover rates and placement counts. Retrieval was carried out at six-second intervals per record and spread across separate days.