# Search protocol — WP36 literature review

Date of search: 2026-11-01 (single session). Agent: literature agent for UQO Working Paper 36, "The Hidden Dynamics of Used-Car Markets: Price Discovery, Inventory Liquidity, and Strategic Dealer Behavior".

# Objective

Build a verified bibliography (60–90+ entries, quality over quantity) covering nine topic buckets (A–I, see evidence_matrix.csv), establish what is already known about used-car asking-price dynamics, markdowns, exit and liquidity-type measures (industry indices), and state precisely what WP36 adds.

# Databases and tools

# Queries run (web search)

1. Manheim Used Vehicle Value Index methodology mix mileage seasonally adjusted
2. Canadian Black Book used vehicle retention index Canada methodology
3. Statistics Canada consumer price index used vehicles introduced methodology J.D. Power data 2022
4. BLS CPI used cars and trucks methodology J.D. Power data factsheet
5. used car dealer price reductions time on lot listing data empirical study markdown hazard (extended)
6. Cox Automotive days' supply used vehicle vAuto methodology; CarGurus used car price index; iSeeCars used car price study
7. CarGurus Used Car Price Index methodology; iSeeCars used car price drops study methodology; AutoTrader.ca price index report used vehicle Canada
8. Court 1939 "Hedonic price indexes with automotive examples" The Dynamics of Automobile Demand General Motors pages
9. "used car" "liquidity index" OR "liquidity score" market days to turn dealership inventory academic paper (extended)
10. Kraft Rao odometer thresholds used car time on lot Texas title data paper (no usable result; item discarded)
11. CarGurus "Used Car Price Index" OR "Price Trends" methodology instant market value listings
12. iSeeCars study used cars price drops number of price cuts days to sell methodology
13. Canada used vehicle market depreciation study economics Quebec "used vehicle" prices dealers research paper OR working paper (extended)
14. JSTOR/AER record confirmation queries for Bond 1982, Lazear 1986, Genesove & Mayer 1997.

# Crossref / OpenAlex query batches

• Batch 1 (85 candidate DOIs, buckets A–D), Batch 2 (97 candidate DOIs, buckets E–I plus extras), Batch 3 (57 corrected/new DOIs). Bibliographic searches: 31 + 32 + 11 titles. OpenAlex: 12 + 10 titles. Semantic Scholar: 7 + 6 titles.
• Logs kept in the session scratch directory (/tmp/wp36lit/out*.tsv, s*.tsv, cache.json); the final verification string for each entry is in the verified_via column of evidence_matrix.csv.

# Inclusion criteria

1. Peer-reviewed article, handbook chapter, book, or NBER working paper whose metadata (title, authors, year, venue) were confirmed by Crossref, OpenAlex, Semantic Scholar, IDEAS/RePEc or the publisher page; or an official statistical-agency / industry methodology document fetched directly.
2. Direct relevance to at least one WP36 topic bucket (hedonic pricing of vehicles; durable-goods / used-car markets; dealer inventory and dynamic pricing; search, dispersion and liquidity; survival analysis; spatial competition and regional integration; identification of strategic interaction and event-study methods; price indices and listing-vs-transaction measurement; multiple testing and model evaluation).
3. Preference for the canonical, most-cited statement of a result (e.g. the published AER version over the NBER version where both exist, except Porter & Sattler 1999 which was never published).
4. Items listed in the brief with uncertain identity ("Jerath/Jim??", "Hyndman?", "Hortaçsu?" car-dealer markdown, "Reis & Silva?", "Dunne & Mian?", "Kraft & Rao") were searched; those that could not be identified or verified were discarded rather than guessed.

# Exclusion / safety

• Known retracted items (Kim–Muhn–Nikolaev 2024; Corbet et al. 2019 IRFA) are not cited and did not surface.
• Candidate DOIs that resolved to unrelated records were replaced only when Crossref bibliographic search returned an exact title/author/venue match; otherwise the item was dropped (e.g. Thomadsen 2005 RAND, confirmed by OpenAlex but with no DOI and no resolvable RePEc handle, was dropped).
• Grey literature (Manheim, Cox, CBB, AutoTrader.ca, CarGurus, iSeeCars) is included only as documentation of existing indices, flagged as non-peer-reviewed in notes.

# Output checks

• references.bib parsed with a regex count of @type{key, blocks; duplicate keys checked; all characters ASCII.
• Every key in evidence_matrix.csv checked to exist in references.bib and vice versa (script in session log; result reported in the final memo to the caller).