Chapter 8 — Further Reading

Sources for going deeper on information gathering — the application, the loss run, third-party data, the inspection, and the rules that govern what you may use. Grouped by the book's three citation tiers (see the style bible): Tier 1 verified canonical; Tier 2 attributed but specifics to verify against current sources; Tier 3 illustrative/constructed teaching material.

If you read only one thing: the Federal Trade Commission's public report on credit-based insurance scores in automobile insurance. It is the single best demonstration of this chapter's hardest lesson — that a data source can be both statistically predictive and unevenly distributed across groups, and that predictiveness alone does not settle whether or how to use it. Read it for the discipline of holding two true things at once.

Tier 1 — Verified canonical

  • The Fair Credit Reporting Act (FCRA), 15 U.S.C. §1681 et seq. — the federal statute governing the use of consumer reports, including for insurance underwriting: permissible purpose, the adverse-action notice, and the consumer's dispute and access rights. The core legal framework of §8.6.
  • U.S. Federal Trade Commission — report to Congress on credit-based insurance scores in automobile insurance — the public federal study finding the scores predictive of loss and unevenly distributed across racial and ethnic groups, with effects not fully explained by other variables. The basis of Case Study 1.
  • LexisNexis Risk Solutions — CLUE (Comprehensive Loss Underwriting Exchange) — the consumer-reporting agency operating the industry's shared property and auto claims-history database; public consumer-facing materials explain what is reported and the consumer's FCRA rights to access and dispute.
  • ACORD (Association for Cooperative Operations Research and Development) — the nonprofit that maintains the standardized insurance application and form set used across commercial lines; the source of the "ACORD application" referenced in §8.1.
  • The McCarran-Ferguson Act (1945) — the federal law leaving the regulation of insurance largely to the states, which is why permitted rating factors and credit-score rules form a state-by-state patchwork (Chapter 4 owns this; applied here in §8.6).
  • The Institutes (American Institute for CPCU) — AINS and CPCU underwriting curricula — the professional body's treatment of the underwriting information-gathering process: sources, the submission, loss-run analysis, inspections, and the regulatory constraints on their use.
  • National Association of Insurance Commissioners (NAIC) — model laws and public materials on the use of credit information in insurance and on consumer protections; the venue where much of the state-level policy debate on data use plays out.

Tier 2 — Attributed; verify specifics against current sources

  • Progressive's "Snapshot" and the broader usage-based-insurance (UBI) / telematics literature — the public history of telematics adoption in personal auto (Case Study 2). The program and the movement are real and well documented; specific discounts, participation figures, and program mechanics change over time and by state, so confirm any current detail before citing it.
  • Industry guidance on loss-run analysis and commercial submissions (reputable trade and carrier sources) — the working practice of ordering five years of currently-valued runs across all lines, reading cause and trajectory, and treating open reserves as real but uncertain. The practice is standard; exact valuation-date conventions vary by carrier and market.
  • Consumer-advocacy and civil-rights analyses of credit-based insurance scoring — the disparate-impact critique summarized in Case Study 1. The fairness arguments are well represented in the public record; quantified claims should be checked against their underlying studies, which differ in scope and method.
  • State insurance department bulletins on permitted and prohibited rating factors — the operational source for "know your state." These are authoritative per state and change; always check the current bulletin for the jurisdiction and line you are writing.

Tier 3 — Illustrative / constructed

  • The Harbor Steel & Fabrication underwriting file (Harbor Steel, Port Hadley, Meridian Risk Partners, the two fires, the loss runs, the information order, and all attached figures) — a constructed teaching example, realistic but not drawn from any real account. The dollar figures (\$180K and \$1.2M fires, \$20M building, \$45M revenue, \$11M payroll) are illustrative.
  • The worked loss-run reading exercises (Figure 8.1; the year-by-year property history in the exercises) — illustrative claim histories chosen to make frequency-versus-severity and trajectory legible.
  • The constructed distributor and plastics-manufacturer submissions (in the exercises) — teaching scenarios built from realistic patterns, with all numbers constructed.