Chapter 31 — Further Reading

Sources for going deeper on data-driven underwriting, grouped by the book's three citation tiers. Tier 1 is verified canonical material we stand behind by name; Tier 2 is real, attributable practice and commentary whose exact citation you should confirm before quoting a specific figure; Tier 3 is the constructed teaching material in this chapter. Where a source's precise statistics are not pinned down, treat the claim qualitatively — the chapter's accuracy rule applies to your reading as much as to its writing.

Tier 1 — Verified canonical

  • The Fair Credit Reporting Act (FCRA), 15 U.S.C. §1681 et seq. — the federal statute governing information from consumer reporting agencies used in eligibility and pricing decisions, including notice, accuracy, and dispute rights. The legal backbone for how much of the third-party data in this chapter may be used. (Owned in Chapter 8; foundational here.)
  • The NAIC (National Association of Insurance Commissioners) — model laws and the ongoing regulatory work on insurers' use of data, third-party data sources, and artificial intelligence. The NAIC's published bulletins and principles on the use of AI/Big Data are the central regulatory reference for data-driven underwriting; consult the current versions directly.
  • State insurance departments — the actual regulators (under McCarran-Ferguson, Chapter 4) of what data and factors a carrier may use in a given state and line. The binding rules on alternative data are state law; there is no single national standard.
  • The Institutes (CPCU / AINS / AU curricula) — the professional bodies of knowledge covering information gathering, automated and algorithmic underwriting, and the underwriting workflow; the canonical certification-aligned treatment of this chapter's material.

Tier 2 — Attributed, specifics unverified

  • Industry and trade press on aerial/satellite-imagery underwriting — reputable insurance trade publications and the property-intelligence vendors themselves have documented, over the past decade, the adoption of imagery-derived roof and property data in underwriting. Read these for the pattern (Case Study 1); confirm any specific adoption figure before quoting it.
  • Consulting and rating-agency commentary on InsurTech and data-driven underwriting — major consultancies and rating agencies have published extensively on pre-fill, real-time scoring, straight-through processing, and the economics of automation in underwriting. Useful for the strategic picture; treat specific percentages as illustrative unless you can source them.
  • Telematics / usage-based insurance program disclosures — public materials on real UBI programs document what telematics data is collected and how it is used (the §31.2 IoT discussion). Confirm specifics against the program's own current disclosures. (UBI is owned in Chapter 14.)
  • Data-governance and data-quality literature (general) — the "garbage in, garbage out" discipline of §31.7 is a standard topic in data-management practice; the dimensions of data quality (accuracy, currency, completeness, consistency, provenance, relevance) are widely treated and worth reading on outside the insurance context.

Tier 3 — Illustrative / constructed (this chapter)

  • The Harbor Steel & Fabrication Underwriting File — the running constructed commercial account; this chapter adds the pre-fill and the satellite-roof corroboration (Figure 31.1). All facts and figures are illustrative and frozen in the continuity ledger.
  • Tindall Stores — the constructed post-breach cyber submission, enriched in parallel here (introduced in Chapter 24).
  • The model-override case — the constructed anchor in which a model scores Harbor Steel a 7 and the underwriter writes it at a 6; logged here, resolved in Chapter 32.
  • Case Study 1 (the roof seen from space) and Case Study 2 (the pre-fill failure, an explicitly labeled composite) — the chapter's two cases, the first describing a real industry shift with an illustrative carrier, the second a teaching composite of real failure patterns.

If you read only one thing: read your own carrier's (or a public carrier's) data-governance and third-party-data-use policy alongside the current NAIC bulletin on the use of AI/Big Data in insurance — together they show you, more concretely than any textbook, the exact line between what data-driven underwriting can do and what it is permitted and trusted to do. That line is where this chapter lives.