PulsCar analysis of NHTSA public data, September 2026 Topic: manual-transmission clutch failures - owner complaints, recall campaigns and manufacturer communications License: CC BY 4.0 (attribution: "PulsCar analysis of NHTSA public data, September 2026") SOURCES - NHTSA Office of Defects Investigation flat file COMPLAINTS (FLAT_CMPL), field dictionary https://static.nhtsa.gov/odi/ffdd/cmpl/CMPL.txt 777,040 owner complaints received 1 January 2015 - 8 September 2026. - NHTSA ODI flat file RECALLS (FLAT_RCL), field dictionary https://static.nhtsa.gov/odi/ffdd/rcl/RCL.txt - NHTSA ODI flat file MANUFACTURER COMMUNICATIONS (FLAT_TSBS), field dictionary https://static.nhtsa.gov/odi/ffdd/tsbs/TSBS.txt - Downloaded and indexed 17 September 2026; queries run 18 September 2026. QUERY - "a complaint about the clutch in a car with a clutch pedal" This is the hard part of the topic. The word "clutch" appears in 6,655 complaints, and most of them are not about a foot clutch: Ford's PowerShift dual-dry-clutch automatic alone accounts for roughly half the file, and the phrase also covers A/C compressor clutches, fan clutches, hybrid disconnect clutches and motorcycles. The published set was built in five steps. Step 1. Full-text search of the complaint narrative for "clutch". -> 6,655. Step 2. Motorcycle and powersports makes removed by make name. -> 6,344. Step 3. Foot-clutch test. The narrative must contain at least one phrase that only makes sense on a car with a clutch pedal: "clutch pedal", "manual transmission", "manual trans", "stick shift", "N-speed manual", "slave cylinder", "throw out bearing" / "throwout bearing" / "throw-out bearing", "release bearing", "clutch master cylinder", "pressure plate", "flywheel", "depress/press/push the clutch", "clutch fork", "clutch kit", "pilot bearing", "clutch disc". -> 1,257. Step 4. Automatic exclusion. The narrative is dropped if it names an automated or torque-converter transmission anywhere: "dual clutch", "dual-clutch", "PowerShift", "DCT", "DSG", "CVT", "automatic transmission", "torque converter", "clutch pack", "Getrag", "Tiptronic", "paddle shift", "S tronic", "PDK", "automatic gearbox". -> 1,183. Step 5. Subject test. The clutch has to be what the complaint is ABOUT, not a passing mention. The narrative must contain a clutch-failure phrase - slipping, burned, failed, went out, worn, replaced, "clutch job", "clutch kit", friction disc, pressure plate, flywheel, release/throw-out/pilot bearing, slave or master cylinder, clutch fluid or hydraulics, clutch fork, clutch pedal went / stuck / to the floor / broke, pedal assembly or bracket, "clutch will not / does not disengage", clutch cable, clutch chatter / shudder / noise. -> 654. Step 6. Two carve-outs. (a) Complaints whose only clutch content is the clutch-pedal SWITCH - cruise control that will not cancel, hill-hold assist - are dropped unless a real clutch part is also named (3 dropped). (b) Ford Focus and Fiesta narratives are kept only when they say "manual transmission", "stick shift" or "N-speed manual", because PowerShift dominates those two nameplates and many narratives simply say "the clutch" (65 dropped). -> 586 complaints. This is the published set, received 6 January 2015 to 7 September 2026. SAMPLE READING (required before publishing any keyword count) Two rounds of 30 random matches were read in full. At the Step-3 stage (1,176 kept, before the subject test) roughly a quarter of the sample was off-topic: engine-stall complaints that happened to mention pushing the clutch in, cruise-control complaints about the pedal switch, hill-hold complaints, and PowerShift Focus narratives that never use the word "automatic". Steps 5 and 6 were written against those specific misses. In the final two samples of 30, three narratives in each were about something else - a seized gearbox, a rev-matching / camera fault, a truck that lost forward gears - so the observed false-positive rate is about 10% in samples of 30. Counts in the article are described as "complaints whose narrative is about a clutch on a car with a clutch pedal", never as a count of confirmed clutch failures, and never as a failure rate. 586 is a small set by the standards of this database, and the article says so on the page. FIGURES QUOTED IN THE ARTICLE (published set, n = 586) Towed 4.95% (29) against a 5.11% baseline over all 777,040 complaints. Crash 1.54% (9) against 4.37%. Fire 0.85% (5) against 1.82%. 7 injuries, 0 deaths. Mileage: median 48,000; quartiles 26,000 and 78,586 (n = 387 with a usable odometer reading; miles <= 0 or >= 400,000 dropped). Baseline median over all complaints with a usable reading (n = 392,935): 70,000. 147 of the 387 (38.0%) are at or under 36,000 miles; 46 are at or under 12,000. What the narrative names (a complaint can name more than one part, so the shares do not add to 100): hydraulics - slave or master cylinder, clutch line, clutch fluid - 211 (36.0%); flywheel 118 (20.1%); pressure plate 109 (18.6%); release or throw-out bearing 72 (12.3%); friction disc, clutch plate or clutch kit 63 (10.8%); clutch pedal assembly or bracket 50 (8.5%); clutch fork 7; pilot bearing 1. Symptom words: pedal to the floor / no pressure / will not return 130 (22.2%); will not go into gear, grinding, drag 86 (14.7%); slipping 59 (10.1%); noise - rattle, chirp, squeal, whine 36 (6.1%); burning smell or smoke 26 (4.4%); judder, shudder or chatter 18 (3.1%). 45 complaints (7.7%) contain warranty-refusal language - "wear item", "wear and tear", "normal wear", "not covered", "refused to cover", "would not cover". Of the 30 of those with a usable odometer reading, the median is 26,657 miles and 19 are at or under 36,000. 6 complaints (1.0%) say the shop found nothing or called the condition normal. Owner-quoted dollar figures: every "$N" between $100 and $20,000 in a narrative was collected - 105 figures, median $1,800, quartiles $1,012 and $3,316, 75.2% above $1,000 and 36.2% above $2,000 (file clutch-owner-reported-amounts.csv). These are NOT prices. They are amounts owners chose to type into a federal safety complaint, which selects for the expensive end; a person who paid $700 and drove away rarely files. Signature-failure table (clutch-signature-failure-by-model.csv): for every nameplate with 8 or more complaints, the count of narratives naming each part group. Nissan 370Z 47 of 49 name the hydraulics; Fiat 500 31 of 51 name the pressure plate; Jeep Wrangler 24 of 41 name the pressure plate; Toyota Corolla 36 of 41 name the hydraulics; Subaru WRX 15 of 34 name the release bearing and 13 the pedal assembly; Mazda RX-8 11 of 15 name the pedal assembly or bracket. QUERY - recall campaigns All rows in FLAT_RCL whose COMPONENT names a clutch assembly or a manual transmission AND whose DEFECT text names a clutch, flywheel or pressure plate. Excluded: motorcycle and powersports makes, heavy-truck and bus makes, and defect text naming a dual-clutch or automatic transmission, a torque converter, an A/C compressor clutch, a fan clutch, a hybrid disconnect clutch or a drive belt. Campaigns counted DISTINCT on the campaign number. -> 18 campaigns, 222,414 potentially affected vehicles, report dates 4 June 2013 to 9 June 2026 (file clutch-recall-campaigns.csv). Note on double counting: Jeep campaigns 20V-124, 21V-028 and 23V-116 cover overlapping populations of the same Wrangler and Gladiator vehicles, and 23V-116 says in its own text that it "expands and replaces previous recall 21V-028". The sum of the three (145,325) is therefore not 145,325 distinct vehicles, and the article says so. QUERY - manufacturer communications All rows in FLAT_TSBS whose SUMMARY contains "clutch" plus at least one manual-clutch term (manual transmission, clutch pedal, flywheel, slave cylinder, release or throw-out bearing, pressure plate, clutch master, clutch disc, clutch kit), minus any summary naming a dual clutch, PowerShift or a torque converter. Volkswagen's monthly "VIN4APIN" documents are index files that list dozens of unrelated customer complaints in one summary, and were removed - without that step Volkswagen leads the table on documents that are not clutch bulletins. Counted DISTINCT on NHTSA_ID. -> 247 documents (file clutch-bulletins-by-theme.csv). 68 (27.5%) name a noise; 40 (16.2%) name the hydraulics; 36 (14.6%) put software, reprogramming or calibration in the remedy; 30 (12.1%) name slipping; 17 (6.9%) name the flywheel; 11 (4.5%) tell the technician the condition is normal or not to repair it. LIMITATIONS - Complaints are self-reported, unverified, and not a failure rate. There is no denominator: nobody publishes how many cars with a clutch pedal are on US roads by model, and manual take rates differ by more than an order of magnitude between nameplates. A model can lead this table because it sold well to enthusiasts who file complaints, not because its clutch is worse. Nothing on the page ranks reliability. - 586 is small. Sub-counts below about 20 are reported as what they are - a handful of reports - and no percentage is quoted off a base under 10. - Only 387 of 586 complaints carry a usable odometer reading, so every mileage figure describes that subset, not the whole set. - Zero of the 247 manual-clutch bulletins use the phrase "dual mass flywheel". That is a fact about the one-paragraph summaries in the flat file, not evidence that no such bulletin exists - the full PDFs are not in the flat file and were not read in bulk. - Manufacturer-communication summaries are written by the manufacturer, and the flat file carries one paragraph per document. Theme counts are counts of what the summary says, not of what the full bulletin instructs. FILES clutch-complaints-by-make.csv complaints and median failure mileage by make clutch-top-models.csv nameplates with 3+ complaints, with model-year span clutch-signature-failure-by-model.csv which part each nameplate's owners actually name clutch-parts-owners-name.csv part groups named across the whole set clutch-symptom-and-context.csv symptom words, towed/crash/fire against baseline clutch-owner-reported-amounts.csv owner-typed dollar figures, bucketed clutch-recall-campaigns.csv 18 light-vehicle clutch recall campaigns clutch-bulletins-by-theme.csv 247 manual-clutch bulletins by theme and by make ------------------------------------------------------------------------------ CASE SENSITIVITY — SECTION ADDED 19 SEPTEMBER 2026 ------------------------------------------------------------------------------ NHTSA stores complaint narratives in ALL CAPS for records filed through roughly 2020 and in mixed case for records filed from 2021 on. A case-SENSITIVE search of the narrative field loses about 92% of the complaints filed after 2020, so this is the single largest way an analysis of this database can go wrong while still looking finished. Every query behind this file is case-insensitive. SQLite's LIKE and its FTS5 full-text index both ignore case by default, and no query in this work uses GLOB or any case-sensitive comparison. This was not taken on trust. On 19 September 2026 every published folder was re-checked against the same database, comparing the figures here with what a case-sensitive query would have returned. The result: the complaint set published here (6,655) is LARGER than the absolute ceiling a case-sensitive search of the same terms could have returned (4,287 narratives). A case-sensitive query cannot produce a bigger set than itself, so the bug is ruled out for this folder by arithmetic, not by assurance.