PULSCAR ANALYSIS — P0420 IN NHTSA OWNER COMPLAINTS: WHO REPORTS IT, WHEN, AND WHO PAYS PulsCar analysis of NHTSA public data, September 2026. License: CC BY 4.0. Attribution: "PulsCar analysis of NHTSA public data, September 2026." Published alongside https://www.pulscar.io/blog/p0420-code ------------------------------------------------------------------------------ 0. WHAT THIS FILE ADDS, AND WHAT IT DELIBERATELY DOES NOT REPEAT ------------------------------------------------------------------------------ Two PulsCar pages already analyze this code and their numbers are not recomputed here: * The Check Engine Code Index counts the manufacturer bulletins that name P0420 (143 distinct bulletins, 12 manufacturer groups, at least 60% software updates) and publishes its own method and data. https://www.pulscar.io/blog/check-engine-code-index * The oxygen sensor cost guide reads 100 randomly drawn P0420/P0430 complaints end to end and classifies what each story turned out to be about, and reads the summaries of 342 manufacturer documents that mention either code. https://www.pulscar.io/blog/o2-sensor-replacement-cost This file answers a different question: not what the code is caused by, but WHERE ON THE WARRANTY CLOCK the owners who complain about it are standing, WHEN their reports arrived, and WHAT THEY ARE TOLD when they ask a manufacturer to pay. Nothing here contradicts the two pages above; where our arithmetic differs from theirs by a rounding convention, that is stated explicitly in section 3. ------------------------------------------------------------------------------ 1. SOURCE FILES AND DATE RANGE ------------------------------------------------------------------------------ COMPLAINTS NHTSA ODI complaint flat files COMPLAINTS_RECEIVED_2015-2019, _2020-2024, _2025-2026 https://static.nhtsa.gov/odi/ffdd/cmpl/CMPL.txt (field layout) 777,040 distinct vehicle complaints (product type V), de-duplicated by ODI report number, received 1 January 2015 to 8 September 2026. TSBs NHTSA manufacturer communications flat files https://static.nhtsa.gov/odi/ffdd/tsbs/TSBS.txt (field layout) Data pulled 14 September 2026; this analysis run 18 September 2026. No third-party data was used. ------------------------------------------------------------------------------ 2. THE COMPLAINT SET ------------------------------------------------------------------------------ Query: the narrative field (DESCR) matched against the whole-word regular expression \bP0420\b, case-insensitive. SELECT * FROM complaints WHERE descr LIKE '%P0420%'; -- 307 rows then kept only rows matching \bP0420\b -- 302 rows The five rejected rows were read. All five are run-together strings in which the code is not the code being reported: "CODEP0420", "P0420cataylic", "P04200", "P042033" and "P042000". Excluding them costs nothing and keeps the set identical to the one behind the Check Engine Code Index, which uses the same whole-word rule. FINAL SET: 302 complaints. This is the denominator for every percentage below unless another denominator is named. Self-reported, no denominator, not a failure rate. A complaint set is drawn from people with an expensive, unresolved or frightening problem; a P0420 cleared by a $180 sensor rarely produces a federal complaint. ------------------------------------------------------------------------------ 3. ODOMETER, AND THE ROUNDING NOTE ------------------------------------------------------------------------------ Odometer readings kept only where 0 < miles < 400,000: 98 of the 302 reports state one. n = 98 minimum 3,500 lower quartile 81,000 MEDIAN 108,828.5 (the two middle reports are 108,600 and 109,057) upper quartile 132,300 maximum 244,309 ROUNDING NOTE. The Check Engine Code Index reports 109,057 miles for the same 98 reports. That is the upper of the two middle values; this file averages them, as a median of an even-numbered set conventionally is. The underlying reports are identical and the difference is 228 miles. Neither figure is wrong; the article quotes them together so a reader comparing the two pages is not left guessing. ------------------------------------------------------------------------------ 4. VEHICLE AGE AT THE TIME OF THE REPORT ------------------------------------------------------------------------------ Age = (calendar year of the RECEIVED date) minus (model year), kept where the result is between -1 and 40. Available for all 302 reports. It is a coarse measure: a model-year 2016 car sold in late 2015 and one sold in 2017 are given the same age, so treat single years as approximate and the bands as the finding. 0-3 years 15 ( 5.0%) 4-7 years 145 (48.0%) 8-11 years 111 (36.8%) 12+ years 31 (10.3%) median age 7 years 8 years or older: 142 of 302 = 47.0% Cross-tabulating with the odometer: of the 98 reports that state a mileage, only 18 are inside BOTH halves of the federal emissions warranty for the converter (8 years or younger AND at or under 80,000 miles). 24 of the 98 are at or under 80,000 miles taken on mileage alone. Published as p0420-warranty-position.csv. ------------------------------------------------------------------------------ 5. WHO REPORTS IT, AND WHEN ------------------------------------------------------------------------------ By make and model: published in full as p0420-complaints-by-make-and-model.csv. Raw counts across vastly different sales volumes. NOT a reliability ranking, and they cannot be turned into one, because there is no denominator here — no registration or sales figure is in this data. Model year and year of report: published as p0420-model-year-and-report-year.csv. Model years 2013-2017 account for 170 of the 302. Reports by calendar year of receipt: 5, 5, 4, 9, 30, 30, 25, 22, 37, 33, 45, 57 for 2015 through 2026, the last being a partial year to 8 September. Two things move this line and neither is a change in how often converters fail: a recall or campaign that teaches owners the code's name, and the general growth of the complaint file itself. ------------------------------------------------------------------------------ 6. THE EMISSIONS-TEST SUBSET, AND A KEYWORD FAILURE WORTH PUBLISHING ------------------------------------------------------------------------------ FIRST ATTEMPT (rejected). Matching 'inspect|emissions test|smog' returned 70 of 302. Thirty were drawn at random (random.seed(42)) and read in context. The dominant match was NHTSA's own complaint-form boilerplate — "the vehicle is available for inspection upon request" — plus dealer multi-point inspections, frame inspections and safety inspections at purchase. Only 4 or 5 of the 30 were about a state emissions or smog test. Observed false-positive rate for the intended meaning: roughly 85%. The count was discarded, not published. SECOND ATTEMPT (published). Tightened to: (smog) | (emissions?[ -](test|inspection|check|station)) | ((fail|pass|passing|failing)\w*\s+(the\s+)?(state\s+|nj\s+|ny\s+|ca\s+|annual\s+)? (emissions?|smog|inspection)) | (state inspection) | (inspection sticker) | (vehicle inspection) -> 20 of 302. ALL TWENTY were read in full, not sampled. 15 describe a state emissions, smog or safety inspection genuinely entangled with this vehicle's code — failed it, cannot pass it, cannot register the car, or passed only after repairs; 3 mention an inspection in an earlier year that is incidental to the complaint; 2 are false positives (a battery tested during an oil change; a dealership's own post-repair inspection). So the honest statement is: 20 of 302 P0420 complaints mention an inspection at all, and 15 of those 20 are about one that the code affected. This is a floor, not a rate. Most owners do not tell NHTSA about their state test at all, and not every state runs an emissions program. ------------------------------------------------------------------------------ 7. WHAT OWNERS ARE TOLD WHEN THEY ASK THE MANUFACTURER TO PAY ------------------------------------------------------------------------------ Query: not covered | denied | out of warranty | refus\w* to (cover|repair|fix|replace) | declined (to )?(cover|repair) | won't cover | would not cover -> 31 of 302 (10.3%). ALL THIRTY-ONE were read in full and each was assigned, by one reader, what the refusal was attached to: 19 the converter, exhaust or emissions repair itself 4 an engine claim in which P0420 appears as a symptom 8 something else in the same complaint (a fuel-tank recall, a transmission drum, a rental car, a service contract on an unnamed repair, a paint claim) Reasons given, within the 19: 7 past a mileage or time limit 3 goodwill assistance declined with no rule stated 2 no reason recorded in the complaint 1 the engine variant was outside the campaign's scope 1 the VIN was not on the campaign list 1 the code was not on record at the time of the earlier campaign 1 an extended service contract declined it 1 the owner was told a warranty extension did not apply 1 campaign parts were not available 1 the dealer refused to perform the campaign work Limitations that matter more than usual here. These are ONE SIDE of a conversation: the owner's account of what a dealer or manufacturer said, written down after the fact, with no reply recorded. We have not verified any individual refusal, we cannot tell a correct application of a coverage rule from a wrong one, and 31 is a small number. What the set does show is the SHAPE of the argument — which is overwhelmingly about limits, lists and scope rather than about whether the part failed. Per-complaint categories and a short quoted snippet are published in p0420-coverage-refusals.csv so the classification can be checked. No VIN, name, city or dealer address is published; snippets are capped at 110 characters. ------------------------------------------------------------------------------ 8. PRIMARY DOCUMENTS OPENED FOR THIS PASS (18 September 2026) ------------------------------------------------------------------------------ Ford emission recall 24E12, dealer letter dated 22 October 2024, NHTSA document 11009827. 2018-2020 F-150 with the 2.7L EcoBoost; US population of affected vehicles 505,633; build dates 10 January 2017 to 22 October 2020 across Dearborn and Kansas City. Quoted: the PCM calibration "could, over time, reduce the effectiveness of the emissions control system and the monitoring accuracy of the catalytic converter's efficiency. As a result, the Malfunction Indicator Light (MIL) may illuminate for low catalyst efficiency even though a catalytic converter repair may not be necessary." Remedy: PCM reprogramming at no charge. The letter also records that California and Massachusetts require emission recall repairs before registration renewal, and that dealers issue a Vehicle Emission Recall Proof of Correction certificate. https://static.nhtsa.gov/odi/tsbs/2024/MC-11009827-0001.pdf GM Special Coverage 15810-03, released July 2018, NHTSA document 10144226. 2010-2011 Chevrolet Equinox and GMC Terrain with the 2.4L LAF engine: the MIL "may illuminate due to thermal damage or face erosion of the catalytic converter". Coverage: 10 years or 120,000 miles from the in-service date, "regardless of ownership". Correction: replace the catalytic converter at no charge. Labor 2.6 hours AWD, 1.3 hours FWD, plus 0.1-0.3 diagnosis; a separate labor code exists for "Engine Exhaust - Customer Concern Not Duplicated". Customer reimbursement provisions are set out with a $20 administrative allowance. https://static.nhtsa.gov/odi/tsbs/2018/MC-10144226-9999.pdf Subaru service bulletin 09-84-22R, dated 20 January 2022, revised 1 November 2022, NHTSA document 10226829. 2017-2018 Forester 2.5L: a rattle or fluttering noise from the undercarriage during acceleration "may be caused by cracking in the front exhaust pipe"; the pipe material was changed from ferritic to austenitic stainless steel, and the part replaced is listed as "CONVERTER COMPLETE-FRONT". The bulletin extends the warranty on the front exhaust pipe to "15 Year (180 months) or 150,000 miles". https://static.nhtsa.gov/odi/tsbs/2022/MC-10226829-0001.pdf Each document was downloaded from static.nhtsa.gov and read as text on 18 September 2026; every population, date, limit and quoted phrase above comes from the document itself. ------------------------------------------------------------------------------ 9. FILES IN THIS FOLDER ------------------------------------------------------------------------------ p0420-complaints-by-make-and-model.csv make, model, complaints (302 total) p0420-model-year-and-report-year.csv model year and year of report p0420-warranty-position.csv age bands, odometer bands, the 8/80 test p0420-coverage-refusals.csv 31 refusals, categorized, with snippets method.txt this file ------------------------------------------------------------------------------ 10. LIMITATIONS IN ONE PLACE ------------------------------------------------------------------------------ * Complaints are self-reported and unaudited. They are not failure rates. * There is no denominator: no sales, registration or fleet figure is in the data, so no model can be called better or worse than another from these counts. * Odometer and age are missing or coarse for most reports. * Categorization in sections 6 and 7 is one reader's judgement; a second reader would not assign every borderline case identically. The raw rows are published so anyone can disagree with a specific call. * Keyword sets under-count: an owner who does not use the word is not counted. Every count here is a floor.