The Older Cars Budget Buyers Are Being Pushed Into: We Counted NHTSA Complaints on Every 7-Year-Old Model

Key facts

• Source data: NHTSA ODI Complaints flat file (FLAT_CMPL.txt), retrieved 2026-09-03, file dated 2026-09-02. 2,241,804 total records processed.
• MY2019 cohort: 58,132 component-rows collapse to 40,297 distinct complaints (one complaint = one ODINO) across 724 make/model strings; 90 models carry ≥100 complaints — the tested universe for any "worst" claim.
• Why 2019: iSeeCars' late-August 2026 study found only 11.4% of 3-year-old used cars now sell under $20,000 (a 76.9% drop from 2019), pushing budget shoppers to 7-year-old cars — model year 2019.
• Raw-count leaders: Jeep Cherokee 1,620, Ram 1500 1,464, Ford F-150 1,312, Ford F-250 SD 1,161, Honda CR-V 1,085.
• Normalized universe: the per-100k ranking covers the 37 nameplates with a reliable free 2019 US sales figure (floor 36,285 units). It is a high-volume ranking, not a ranking of all 724 model strings.
• Per-100k leaders (high-volume set): Honda Odyssey 946, Jeep Cherokee 846, Honda Passport 755\*, Ford EcoSport 754, Acura RDX 690.
• Disclosed reversal: widen the denominator set and the Odyssey loses the top spot — Ford Flex is 240 complaints on 24,485 units (980 per 100k) and Nissan Leaf 253 on 12,365 (2,046 per 100k). We therefore never claim "worst 2019 car overall".
• The one constant: Only the Jeep Cherokee ranks top-2 on both measures (raw #1, rate #2) — genuinely worst-in-class however you count.
• Safety flags in the cohort: 1,529 complaints allege a crash, 446 a fire, 963 report injuries, and 19 report a death (20 deaths).
• Complaints are unverified owner reports, not adjudicated defects — a signal to investigate, not a verdict.

Why are budget buyers now shopping 2019 cars?

Affordability has shifted the whole used market down by three model years. iSeeCars, analyzing 11.4 million-plus used sales, reports that in 2026 only 11.4% of 3-year-old cars sell under $20,000, versus roughly half in 2019 — a 76.9% collapse. The median-affordability age moved from 4-year-old cars in 2019 to 7-year-old cars in 2026, where 53.0% still list under $20K. A 7-year-old car in 2026 is a model-year-2019 vehicle. That is the cohort we examined, and every published take covered the prices — none counted the defect record of the exact cars buyers are funneled into.

The affordability figures above are iSeeCars' and are used here as context only. The load-bearing numbers in this article are our own count over the federal complaint file.

Which 2019 models have the worst owner-complaint records?

It depends entirely on whether you count raw complaints or complaints per owner — and the two disagree sharply. Raw counts reward volume: the biggest sellers naturally accumulate the most reports. Normalizing by 2019 units sold (a vehicles-in-operation proxy) is the fairer read, and it reverses much of the list. The table below shows both, sorted by the per-100k rate, across the 37 high-volume nameplates we could give a reliable denominator.

\* Passport launched in early 2019, so a partial year of units backs a full year of complaints — its rate is inflated; footnote it, don't headline it. ‡ Pickup denominators use the full line (including heavy-duty), so the true 1500-only rate is higher than shown — which only strengthens the finding that these trucks fall on normalization.

Does a high complaint count mean the car is unreliable?

No — and the robustness check is the whole point of this piece. Raw complaint volume correlates with fleet size, media attention and recall activity as much as with how a car behaves. When we re-rank by complaints per 100,000 units, the "worst" list flips: the Silverado 1500 drops from raw #4 to rate #31, the RAV4 from #8 to #29, the Ram 1500 from #2 to #25, and the CR-V from #3 to #19. Meanwhile the Honda Odyssey jumps from #6 to #1, the Ford EcoSport from #20 to #4, and the Acura RDX from #22 to #5. This is the same volume-versus-rate distinction we found when checking whether owner complaints predict a car recall — raw counts flatter big sellers.

The single exception is the Jeep Cherokee: #1 raw, #2 per-owner. It is the only 2019 model that lands in the top two no matter how you count, driven by 784 power-train complaints centered on its 9-speed ZF automatic and 454 engine complaints on the 2.4L Tigershark, against 11 NHTSA recalls including two automatic-transmission campaigns. If you want the underlying pattern, that gearbox shows up across our count of used cars with the most transmission problems.

What are the recurring problems on these 2019 models?

These are complaint patterns from the dominant COMPDESC categories and well-documented model histories — not adjudicated root causes.

• Honda Odyssey (rate #1): electrical (217), forward-collision-avoidance false triggers (215) and lane-departure (152), plus sliding-door/latch reports. It also carries the most recalls in the top group — 20 — including a back-over camera, a door-latch and the Denso fuel-pump campaign. High complaint volume plus heavy recall activity is exactly why it tops the per-owner list; for a longer read on how we count a single nameplate, see our Honda Odyssey NHTSA complaints analysis for the prior generation.
• Honda CR-V: 296 forward-collision/phantom-AEB complaints and 252 fuel/propulsion reports (the 1.5T oil-and-fuel-dilution pattern); 8 recalls including the Denso fuel pump.
• Acura RDX: wiper/visibility (96) and engine (89); 5 recalls, three of them Denso fuel-pump campaigns.
• Ford EcoSport: engine-cooling and overheating dominate (215 + 191); 2 recalls.
• Subaru Outback: Starlink infotainment reboots/blackouts (362) and windshield cracking (326).
• Ford F-150: 722 power-train complaints on the 10R80 10-speed (harsh/erratic shifting) and 316 engine; 9 recalls. Ford reports only F-Series combined (896,526 units), so the family rate is 328 per 100k and individual F-150-versus-Super-Duty rates aren't derivable from free data.

The Denso low-pressure fuel-pump recall of 2020 swept many 2018–2019 Honda, Acura and Toyota vehicles, which is why it appears in the CR-V, Odyssey and RDX recall lists.

By the numbers

• 40,297 distinct MY2019 complaints; 724 models; 90 with ≥100 complaints.
• 946 per 100k — Honda Odyssey, the highest normalized rate in the high-volume set (Ford Flex and Nissan Leaf are higher once low-volume models are admitted).
• 95 per 100k — Nissan Rogue, the lowest in the analyzed set.
• −27 ranks — the Silverado 1500's fall from raw to per-owner, the largest drop.
• +34 ranks — the Honda Passport's jump, flagged for its partial launch year.
• 8,413 / 7,571 / 7,117 — the cohort's top component categories: Engine, Electrical, Power Train.
• 20 deaths alleged across 19 complaints; 446 allege a fire.

Based on 2,241,804 NHTSA complaint records (MY2019 subset = 40,297 distinct complaints), FLAT_CMPL flat file retrieved 2026-09-03; rates use 2019 US calendar-year sales as a vehicles-in-operation proxy, across the 37 high-volume nameplates with a reliable denominator. Rates are complaints per 100,000 units sold — directional, not exact. Frozen ranking: my2019-nhtsa-complaints-ranking-2026-09-03.csv.

Method

Source file. NHTSA ODI Complaints flat file FLAT_CMPL.txt (from FLAT_CMPL.zip), downloaded from static.nhtsa.gov/odi/ffdd/cmpl/ on 2026-09-03; the file is internally dated 2026-09-02. It contains 2,241,804 records. The file is tab-delimited and 0-indexed fields used here are ODINO (1), MAKETXT (3), MODELTXT (4), YEARTXT (5) and COMPDESC (11).

Universe. Rows where YEARTXT is exactly 2019. That is 58,132 component-rows, which collapse to N = 40,297 distinct complaints across 724 make/model strings. NHTSA files one row per cited component, so counting rows would double-count a complaint that names three parts; we count distinct ODINO throughout. Counts are complaints, not vehicles and not owners.

Model keying. Models are keyed on the exact trimmed MAKETXT|MODELTXT pair with no substring matching. This matters: JEEP|CHEROKEE (1,620) and JEEP|GRAND CHEROKEE (362) are different vehicles and a substring match would silently merge them and inflate the Cherokee by 22%.

Denominators. 2019 US calendar-year sales are used as a vehicles-in-operation proxy, taken from GoodCarBadCar and best-selling-cars.com. A denominator was only accepted where a per-nameplate figure is publicly available and matches the complaint nameplate, which yields 37 models (floor: Honda Passport, 36,285 units). Rates are complaints per 100,000 units.

Aggregation. For each of the 37, complaints per 100k = distinct ODINO / 2019 units x 100,000, then ranked against the raw-count rank. Both ranks and every caveat ship in the frozen CSV.

Known limits. (1) Complaints are unverified owner reports, not adjudicated defects. (2) Sales in one year are a proxy for the fleet, not the fleet — they ignore scrappage, imports and multi-year production. (3) Complaint volume is driven by recall publicity and media attention as well as by the vehicle; a heavily recalled model attracts reports. (4) Two pickup denominators cover the whole line including heavy duty (Ram 1500, Chevrolet Silverado 1500, GMC Sierra 1500), so their true 1500-only rates are higher than shown — which only deepens their fall on normalization. (5) The Honda Passport launched in early 2019, so a partial year of sales backs a full year of complaints and its rate is inflated; it is footnoted, never headlined. (6) The per-100k ranking is over high-volume models only — see the reversal below. (7) Complaints accrue over a vehicle's life, so a 2019 cohort measured in 2026 reflects seven years of exposure, not a first-year defect rate.

Robustness checks

We ran the checks most likely to reverse the finding. One did, and it is disclosed here and in the CSV.

Label consolidation. NHTSA's MODELTXT is free text and one nameplate can occupy several strings, which has reversed a ranking of ours before. We swept all 724 MY2019 strings for same-make prefix pairs and found 87. Every one is a genuinely distinct variant sharing the same model year — RAV4 / RAV4 HYBRID, ACCORD / ACCORD HYBRID, ROGUE / ROGUE SPORT, FUSION / FUSION ENERGI — rather than one vehicle renamed mid-range, so they are correctly left split. None of them involves the Odyssey or the Cherokee, and merging any of them does not lift a model into the top two.

Models we could not normalize. Four models outrank the Odyssey on raw count but carry no clean free denominator: Ford F-150 (1,312), F-250 SD (1,161), F-350 SD (466) and Ford Ranger (264). Ford reports F-Series only as a combined 896,526 units, giving a family rate of 328 per 100k — well below the leaders — so none of them would displace the top two.

The one reversal. The per-100k ranking covers only nameplates selling above ~36,000 units in 2019. Extend it downward and the Odyssey is not #1: Ford Flex is 240 complaints on 24,485 units (980 per 100k) and the Nissan Leaf 253 on 12,365 (2,046 per 100k). Both denominators are small enough that a few dozen complaints swing the rate hundreds of points, which is exactly why the analyzed set has a volume floor — but the honest consequence is that we do not claim the Odyssey is the worst 2019 car, only that it leads the high-volume set. The full 724-model raw count is published so anyone can re-run the normalization with their own denominators.

What a VIN check can and can't tell you here

Our ranking describes how a model behaves across its whole population. It does not describe the individual 2019 car in a listing — that is a separate, per-VIN question. Use the model ranking to decide which cars deserve extra scrutiny, then verify the specific vehicle. Because older cars in a tight market get flipped repeatedly, an individual car's history matters as much as its model reputation — the same reason we track at what mileage cars actually start to fail when sizing up a high-mileage listing.

To screen a 2019 car for open, unrepaired recalls, start with NHTSA's free VIN lookup at nhtsa.gov/recalls — it is the authority on remedy status. As a practical next step, a Zilocar VIN check confirms recall presence and adds the history NHTSA doesn't: accident and airbag-deployment records, salvage/junk-auction and odometer/rollback checks, theft (NICB), ownership, and sales-listing history. It does not confirm a recall was remedied, track investigations, or show firmware status — take those to the dealer with the VIN. For the same open-recall step framed around one campaign, see what NHTSA complaints show before a recall, and how to check a used VIN.

Related checks

• Used cars with the most transmission problems
• Used Nissan Rogue problems to check before buying
• Do owner complaints predict a car recall?

Sources

• NHTSA ODI Complaints flat file (FLAT_CMPL.zip) — primary federal dataset our ranking is computed from; retrieved 2026-09-03, N = 2,241,804 records.
• NHTSA datasets & APIs hub — complaints, recalls (recallsByVehicle campaign counts) and product-information references.
• NHTSA CMPL flat-file field layout — 51-field spec.
• iSeeCars — "What Happened to Affordable Cars" study — affordability hook and 7-year-old cohort definition (context only).
• Motor1 — coverage of the iSeeCars sub-$20K study — independent corroboration of the hook.
• best-selling-cars.com — 2019 full-year US sales and GoodCarBadCar — per-model 2019 US sales — denominators for the per-100k normalization.
• Frozen data: my2019-nhtsa-complaints-ranking-2026-09-03.csv (normalized table, both ranks, the analyzed-set flag and caveats) and my2019-nhtsa-complaints-raw-all-models-2026-09-03.csv (all 724 model strings, raw distinct-complaint counts).

Last verified: 2026-09-03.