Consumer Reports Car Brand Reliability Rankings 2015

cr_carcoverConsumer Reports recently released the results from their 2015 Annual Auto Reliability Survey of over 740,000 vehicles owned by its subscribers. Below are the complete rankings from Consumer Reports, including the change from last year. Taken from this CNBC article. Lexus and Toyota remain on top. In terms of big movers, Honda went down 4 spots, while Kia moved up 4 spots.

One of the trends they note is that fancy infotainment systems and complex transmissions (including CVT, 8+ speeds, and dual clutch) are a growing source of complaints. Many brands, including Acura, were significantly hurt in their rankings due to issues in these areas.

cr_reliability_2016

Here’s another view that takes into account the range of scores taken from individual models (the brands are ranked by averages). Taken from the public version of the Consumer Reports page.

cr_reliability_2016_2

Jaguar, Land Rover, Mitsubishi, Scion, Smart, and Tesla were excluded due to a lack of data on two or more of their models. However, Tesla’s Model S was individually given a reliability rating of “below average”.

Comments

  1. Yet another consumer reports automotive ranking that bears no reality to my experience in car ownership.

    The new Subaru and Toyotas we bought were defect prone. No more for us.

  2. I’d like to understand the baseline on these. The scores and rankings are relative to the average. But what does average reliability really mean? Does it mean that 1 in 1000 new cars are basically lemons or does it mean that 1 in 10 new cars has a “complaint” such as, “I can’t figure out the touchscreen thingy in my new car.”? I think the details matter and what they’re scoring and grading matters. I mean +/-50% on stuff like .001% failure rates or complaints about “ease of use” of an infotainment system are really not too useful. But if this is more hard concrete stuff like major drive train failures in the first year or other even minor concrete flaws then thats more useful.

    • I don’t know the specific answers but average is, well, the average. But being 50% to 100% better than average should be statistically significant in my mind.

      I want to say a “serious problem” requires a fix by the mechanic. That’s how I recall these surveys working.

      http://www.consumerreports.org/cro/2012/04/understanding-reliability-ratings/index.htm

      “Consumer Reports provides the most available to consumers (available to subscribers). It’s based on CR’s annual surveys of our over 8 million magazine and website subscribers. These surveys ask about any serious problems they’ve had with their vehicles in the preceding 12 months. They generate hundreds of thousands of responses—the 2015 survey, for instance, provided information on more than 740,000 vehicles—which give us a solid foundation for our reliability Ratings.”

      • Is it significant though?

        I do think it really does matter what average is.

        If the average failure rate is 1 in 1000 then +/- 50-100% of that is pretty small. Its almost insignificant.
        Would you pay a high premium for something that is 99.9% perfect versus only 99.8% perfect?
        Thats a huge difference between +/-100% of an average 1 in 4 failures.
        I’m straining the example here to illustrate the point.

        And I think it really does matter a lot if these are serious or minor problems. They do talk about infotainment systems and transmissions as examples. We’ll I’d take 10 infotainment failures over 1 transmission failure any day.

        I did find this FAQ with some info on methodology:
        http://www.consumerreports.org/cro/2011/10/consumer-reports-car-reliability-faq/index.htm

        It talks about the classes of problems and it does at least say they weigh the serious problems like drivetrain failure more heavily. But theres stuff like “bluetooth” pairing for the audio system or “loose trim or moldings”. So does 3-4 of those trivial things equal a lemon engine?

        These ratings have value and I’d take a lexus over a fiat for sure. But the devil is in the details.

  3. Jeff, no car company is going to put out 100% perfect cars 100% of the time. With any car purchase, new or used, you’re playing the probability game. How likely is it that the car that you buy will have problems? For more reliable brands, it’s less likely (but not unheard of) and for less reliable brands, it’s more likely (but not a certainty).

    The question here is: do you trust your own anecdotal experience over Consumer Report’s semi-scientific amalgamation of the experiences of many car owners? Unless you have some reason to believe that Consumer Reports is biased, you’d be better off trusting them and continuing to buy Subarus and Toyotas.

    Incidentally, what I’d find most interesting is some reliability ranking that also took into account the cost of repairs. Porsches may be marginally more reliable than Lincolns, but if they are astronomically more expensive to fix, the fact that you may be 3% or 5% more likely to need repairs in a Lincoln ends up financially being pretty irrelevant in evaluating the cost of owning these two cars.

    Finally, does anyone else find the spreads really interesting? My understanding of reliability is that it’s based largely on consistency of process and parts. (Read about the GM-Toyota Nummi car plant for more on that.) I’d assume that the spreads in reliability for these companies is a reflection of inconsistency in their factories: some Toyota factories are better than others, which is why some models are more reliable than others. Mazda, here, is the interesting counter-example. Basically all Mazda cars are equivalently reliable. But none of them are as reliable as any Toyotas. I wonder how that happens…

    • To answer your question, no, not unilaterally. However, when I compare CR data to that of JD Power, True Delta or Edmund user reviews I see differences that reflect my personal experience.

      To the question of bias, yes I believe there is bias until proven otherwise. As others have said CR should publish what they consider a “problem” (they are not all equal, so what is minimum threshold?) and the sample size for each make they are drawing from. Entirely possible CR readership and participation in surveys is skewed to a particular demographic. Bottom line is their method would never pass the scientific community as documented, so why should we simply trust it?

  4. I agree with the comments.. would be nice to get more data behind the analysis…

    “how does this thing work” vs “my transmission broke in year 3”

  5. Whats interesting is that Kia is more reliable than hyundai! They are the same company and Kias are cheaper.

  6. Anecdotally, I am very happy with my 2015 Mazda 3 hatchback. Just hit 10,000 miles, no issues so far. A lot of technology and safety features built in standard that you would have seen on a high end luxury car less than five years ago. It’s interesting how tight Mazda’s range is too. Maybe they have very strict quality control?

  7. Funny on Audi. Given my anecdotal experience and the recent brouhaha over their lying to customers… would sooner walk than go near a VW/Audi product. We’ve had 4 and all 4 were a mess.

    True on toyota/mazda/lexus. Just bulletproof.

Speak Your Mind

*