Tag Archives: autosafety

Which is the Safest Subaru: Outback, Legacy, or Forester? IIHS Driver Death Rates (2017)

Per the IIHS' 2017 driver death rate math, a Subaru Outback isn't any safer than a Legacy or Forester--they're all just really safe vehicles.
Per the IIHS’ 2017 driver death rate math, a Subaru Outback isn’t any safer than a Legacy or Forester–they’re all just really safe vehicles.

Subarus have a reputation for safety among middle-to-upper class educated car buyers, and it’s not hard to see why: their flagship wagon, sedan, and SUV, the Outback, Legacy, and Forester, constantly land on the IIHS’ “top safety pick” lists year after year.

However, the ultimate question of vehicle safety doesn’t have to do with how vehicles test in controlled crashes, but with how they actually protect drivers and passengers on the daily bloodbath that is the US road network. If you’re up for crunching NHTSA FARS data, you can pull the figures yourself, but if you’re short on time, the IIHS periodically runs their own calculations and publishes lists of driver death rates for recently sold new vehicles. I enjoy reviewing the data, even though it involves a.) large margins of error and b.) overlooks two thirds of what’s most important in auto safety–road design and driver behavior.

This is part of a series on the IIHS’ summer 2017 Status Report (volume 52, number 3); previous articles include what’s the safest electric / hybrid: the Volt, the Leaf, or the Prius?, Civic vs Accord safety showdown, and my Cruze is as safe as your Suburban. Today we’ll look at whether or not there’s a safety difference between the sales-leading Subaru Outback, its sedan equivalent Subaru Legacy, and the SUV / CUV Subaru Forester.

2011-2014 Subaru Outback – 12 driver deaths (5-20)

Per the IIHS, the Subaru Outback’s estimated driver death rate during the 2011-2014 model years was 12, with 8 of those deaths predicted to occur during multiple vehicle crashes, 3 from single vehicle crashes, and 1 of those from a rollover. The exposure came from 1,116, 891 registered vehicle years. As the IIHS notes, a registered vehicle year refers to one vehicle registered for a full year. This basically means that the IIHS took NHTSA FARS data about fatal car crashes and crossed it with IHS vehicle registration data to make a model of driver death estimates. Their calculations essentially state that if 1 million drivers drove 1 million ’11-’14 Outbacks throughout the US for a year, we’d expect 12 of them to die.

2011-2014 Subaru Legacy – 20 driver deaths (4-36)

The Legacy’s estimated driver death rate during the 2011-2014 model years was 20, with 14 estimated to occur from multi-vehicle crashes, 5 from single vehicle crashes, and 4 of those from rollovers. The exposure resulted from 428,322 registered vehicle years, a figure roughly 1/3rd the size of the Outback’s, which resulted in a much larger confidence bound (spanning 33 instead of 16 with the Outback). As with the Outback, it’s important to remember that the figure above doesn’t mean 20 drivers died while driving Legacy’s; it means that in a sample of 1 million drivers driving 1 million ’11-’14 Legacys for a year, we’d expect 20 of them to die over the year.

2014 Subaru Forester – 28 driver deaths (3-53)

Finally we come to the Forester. It had the highest estimated rate of driver deaths at 28 for the 2014 model year, with 17 predicted multiple vehicle fatalities and 11 deaths resulting from single vehicle crashes. The exposure was based on 134,402 registered vehicle years, the smallest of the three vehicles, which is why the confidence bound spanned 51 instead of 33 as with the Legacy or 16 as with the Outback. For reference, the smallest exposure figure the IIHS uses for their calculations is 100,000 registered vehicle years. Below that, they likely find the figure either introduces too much error into their models or results in a larger than desirable confidence bound. At any rate, the driver death figure means that if we let 1 million drivers loose with 1 million 2014 Foresters, we’d expect 28 of them to die in the course of a year’s worth of driving throughout the country–in New England, in the MidWest, in the South, and in the West.

Does this mean the Outback is the safest, followed by the Legacy, then the Forester?

Not quite. Although the Outback had the lowest driver death rate, followed by the Legacy’s at 20 and the Forester’s at 18, none of the driver death rates were statistically different from the others. This is where the confidence bounds become important. The 95% confidence bounds give us an estimate of where the true driver death rate would fall 95% of the time when we sampled the car and driver pool (e.g., when we allowed 1 million drivers to drive 1 million of the given vehicles for a year, or 500,000 drivers to drive 500,000 of the vehicles for 2 years).

The Outback’s true driver death rate would, per the model, almost always fall between 5 and 20. The Legacy’s? Between 4 and 36. The Forester’s? Between 3 and 53. In other words, there’s a chance (5-20, or 16, out of 3-53, or 51) of 16/51, or 31%, that all three vehicles share exactly the same driver death rate. There’s a chance that the Outback had the lowest true death rate for drivers, a chance that the title would go to the Legacy, and a chance that it actually belonged to the Forester. The broad point is that there’s no way to statistically prove that any of the three vehicles was safer than the other two based on the IIHS’ calculations. Or to put it even more bluntly, we can’t statistically prove that any differences in the driver death rate weren’t simply due to chance.

Does this mean that my children, husband, wife, or family are equally safe in a Forester, Legacy, and Outback?

Per the calculations of the IIHS, yes. Despite the differences in size and price of the vehicles, their actual driving safety as measured by records of driver fatalities in each vehicle did not show a significant difference in odds of survivability from one vehicle to the next. Keep in mind that the IIHS only surveyed driver data, rather than passenger survival data (including that for children). However, since all three vehicles include the same core safety features (i.e., good frontal and side crash scores, side airbags with head protection, and ESC), each could be expected to provide the same amount of passenger protection as individual driver protection. An Outback is a good choice for a family, but so is a Legacy, and so is a Forester. If you’re lucky enough to drive any of these vehicles, whether you survive or not will depend far more on how and where you drive than on which one you’re driving–which will make, statistically speaking, no difference whatsoever.

What’s the single most important thing I can do to keep my family safe in an Outback / Forester / Legacy (or any other vehicle)?

If you’re already in one of these vehicles, or even if you’re not, you’ll get the lion’s share of protection from simply observing the three following rules: Choose safe speeds, follow best practices with car seats,  and choose safe roads. Do these three things, and you’ll increase your family’s odds of both avoiding and surviving crashes by more than you’d manage through any vehicle choice.

If you find the information on car safety, recommended car seats, and car seat reviews on this car seat blog helpful, you can bookmark and shop through this Amazon link. Canadians can bookmark and shop through this link. You can also buy my books on auto safety here.

Small Car Safety: A Chevy Cruze is as Safe as a Chevy Suburban, Per IIHS Driver Death Rates (2017)

Per the IIHS' 2017 driver death rate math, a Chevy Cruze is as safe of a car to drive as a Chevy Suburban, despite being significantly smaller.
Per the IIHS’ 2017 driver death rate math, a Chevrolet Cruze is as safe of a car to drive as a Suburban, despite being significantly smaller.

Every few years, the IIHS lets loose calculations on which new vehicles were recently safest and not-so-safe in terms of driver death rates. I enjoy taking a look at the data and fishing out trends. It’s important to remember that a.) there are huge margins of error and b.) vehicle selection is only one third of what’s important in auto safety–the other two being road design and driver behavior. This is part of a series on the IIHS’ summer 2017 Status Report (volume 52, number 3); previous articles include the safest hybrid / electric cars are currently the Volt and Leaf and a Civic is as safe as an Accord. It’s also part of an ongoing series on small car safety; previous articles include how small cars can be safer than the largest SUVs and why we should all drive small cars. Today we’ll look at how the Chevrolet Cruze fared in driver safety compared to the behemoth model-mate Chevrolet Suburban.

2011-2014 Chevrolet Cruze – 42 driver deaths (32-52)

According to the IIHS, the Chevrolet Cruze’s estimated driver death rate during the 2011-2014 model years was 42, with 29 of those estimated deaths predicted to occur through multiple vehicle collisions, 13 from single vehicle crashes, and 4 of those 13 from rollovers. The exposure was due to 2,220,302 registered vehicle years. What this means is that the IIHS looked at NHTSA data of auto deaths and their data of registered vehicles and made up a model of how many deaths we could expect to see if everyone drove an ’11-’14 Cruze. Their math suggests that if something like 1 million drivers drove 1 million ’11-’14 Chevrolet Cruzes around the US for a full year, we’d expect 42 of them to die. This was almost identical to the figures for both the Civic and the Accord, by the way.

2011-2014 Chevrolet Suburban 1500 4WD – 39 driver deaths (11-67)

In a statistical dead heat with the Cruze, the 2011-2014 model years of the 4WD Chevy Suburban logged a driver death rate of 39. Of these 39 estimated deaths, 23 were predicted to occur from multiple-vehicle collisions, 16 from single vehicle crashes, and 10 of those 16 from rollovers. The confidence bound spanned 11-67 with an exposure of 293,380 registered vehicle years, a figure close to 1/10th the size of the Cruze’s, which resulted in a much larger confidence bound (spanning 56 instead of 20 with the Cruze). As with the Cruze, it’s important to remember that the DDR doesn’t mean that 39 drivers died driving these Suburbans. It means that if 1 million drivers drove 1 million ’11-’14 Suburbans for a full year, 39 would be predicted to die in the course of the year. It’s also important to note that the 2WD version of the Suburban had a much lower driver death rate of 7 (0-38) with 147,811 registered vehicle years of exposure.

How can the Cruze be as safe as the Suburban if it had a higher driver death rate?

Remember the margins of error I discussed at the start of the article? This is where they become relevant. Looking at the numbers, it would seem that the 4WD Suburban was slightly safer to drive than the Cruze in the aforementioned model years due to a DDR of 39 vs a DDR of 42. However, the single number isn’t what makes the difference; the confidence bounds provide the context, and context is important.

The 95% confidence bounds give an idea of where the true driver death rate would occur in 95% of samplings (e.g., situations where 1 million drivers drove 1 million of the given vehicles for a year). The Cruze’s bounds (32-52) and the 4WD Suburban’s (11-67) overlapped a lot. Specifically, the Suburban’s confidence bound encompassed the entire range for the Cruze. There’s no way, statistically-speaking, to say that either vehicle was actually safer than the other based on the IIHS’ calculations. To put it another way, we can’t statistically state that any DDR difference wasn’t due to chance.

What about the 2WD Suburban?  Wouldn’t that be safer than the Cruze?

It’s possible, but it’s still impossible to say given the confidence bounds. Because the confidence bounds of the 2WD Suburban (0-38) still partially overlap with those of the Cruze (32-52), we can’t state that there was a statistically significant difference between the driver death rates of both vehicles. Once again, it’s possible that the difference in DDRs could have occurred purely due to chance. The overlap was much smaller (32-38, or 7 out of 52), suggesting only a 13% chance that they shared the same true DDR. But as long as that chance exists–as long as there’s any overlap–the two vehicles were not statistically different in driver safety. To this end, it’s worth keeping in mind that in the last DDR Status Report (Volume 50, No. 1 in January 29, 2015), the 2011 Cruze had a lower DDR at 42 than the 2WD 2008-2011 Suburban at 60. This is a quick example of how wildly the statistics can swing from one survey to the next with essentially the same vehicles, and why it’s so important not to take the single DDR number as gospel.

How about the 2WD Suburban vs the 4WD Suburban?

As with the 2WD Suburban vs the Cruze, it’s possible the difference in DDRs between the 2WD and 4WD Suburban could be completely due to chance. This also applies to the differences between the 2WD Suburban (0-38) and the 4WD Suburban (11-67). It doesn’t mean the 2WD was actually a safer vehicle than the 4WD. In fact, it’s rather common to see 4WD and 2WD vehicles of the same make and model trade places from one DDR Status Report to the next. In fact, in the last DDR Status Report (Volume 50, No. 1 in January 29, 2015), the 2008-2011 4WD Suburban had a DDR of 17 (0-34) while the 2WD had a DDR of 60 (13-107). The results were completely reversed, even though one of the model years (2011) was the same from one study to the next and the Suburban didn’t see any major design changes in the full span of those years (2008-2014). Most importantly, the confidence bounds overlapped handily and there was still no statistically significant difference between the two vehicles.

Does this mean my son, daughter, husband, wife, children, and family are as safe in a Cruze as in a Suburban?

Statistically speaking, yes. Although the Suburban (2WD or 4WD) is a much larger and expensive vehicle than the Cruze, the actual driving safety as measured by the records of who lived and died driving them didn’t reveal a significant difference in survivability probabilities between the two vehicles. It’s important to remember that the IIHS only looked at driver data, and not survival data for passengers (including children). However, since both vehicles are equipped with the same essential safety features (i.e., good frontal and side crash scores, side airbags with head protection, and ESC), they would likely protect their passengers the same way they did their drivers. A Cruze is a safe family car. A Suburban is a safe, if difficult to drive and fuel-inefficient, family SUV. Vehicles in the US are safe enough; the deaths primarily occur due to how and where we’re driving them.

What can I do to keep my loved ones safe in either (or any) vehicle?

Regardless of what you’re driving, it’s important to remember that vehicle safety only makes up one of the three components tied to auto safety. What make a far greater difference are how you drive and where you drive. Fortunately, driving safely and choosing safe infrastructure doesn’t take money; they just take good judgment. Choose safe speeds, follow best practices with car seats,  and choose safe roads, and your family’s odds of being in a crash (or not surviving one) will be much lower than those you’d find from choosing any vehicle, no matter how much it cost.

If you find the information on car safety, recommended car seats, and car seat reviews on this car seat blog helpful, you can bookmark and shop through this Amazon link. Canadians can bookmark and shop through this link. You can also buy my books on auto safety here.