Finder makes money from featured partners, but editorial opinions are our own.

What we learned from watching a car crash test

Hard lessons learned the easy way.

The sound of the impact was jarringly loud in the enclosed booth above the scene of the crash. At the moment of impact, the car’s hood rippled and then began to crumple like an accordion. The first ripple saw the seatbelt pretensioners fire, yanking the dummies in the front seat backwards into their seats and bracing them for impact. The airbags came along a moment later.

As the crushing continued, the car eventually ran out of crumple zone. The crunch arrived at the main body of the car, eating up the windscreen and doing some real damage.

The driver’s seat dummy was safely strapped in, but it still got a lot of use out of the head-level side airbags as the seat shook loose and jumped around. On the whole, the dummies in the front seat seemed to have come out relatively unscathed.

The dummy kids in the back seat weren’t so lucky. Someone deliberately forgot to face the child seats backwards, so the force of the impact whiplashed the dummy kids’ heads forward in an unmistakably disabling motion.

It had taken days to carefully set up the crash-test dummies, but the crash itself was all over in a split-second glimpse of flying debris and a bang followed by the steady hiss of deflating airbags.

The once fully functional brand-new SUV would now be carted away and written off.

Just another day at the crash lab

In 1985, Australia’s car accident fatality rate was 18.6 per 100,000. That was much too high, so they invented road safety.

By 1999, that new invention halved the fatality rate to 9.4. Then the turn of the millennium brought a new generation of technology and another big jump down in the fatality rate. The fatality rate dropped to 7.7 in 2007, and by 2014, it was only 5.0.

Then 2015 happened. For the first time since the invention of road safety, Australia’s fatal accident rate actually went up. By 2016, it was back up to 5.4.

Fortunately, the Australasian New Car Assessment Program (ANCAP) is on the case. They’re responsible for a lot of those lives saved over the years, and they have a pretty good idea why road death tolls have started increasing again.

What makes a car safe?

“Experienced drivers, safe cars and good roads,” explains James Goodwin, the CEO of ANCAP. “You need at least two of three.”

These are the three pillars in the road safety trifecta, and all have been improving since the 1980s, with corresponding drops in accident rates.

The problem is that tests change over time as standards improve. A car that got five stars in 2002 might be closer to zero today, but still showing as five stars.

So starting on 1 January 2018, you’ll be seeing the year of testing next to the star rating, so you can compare like-for-like vehicle types. Autonomous safety features will also be required for a vehicle to be awarded a 5-star rating.

Ideally, Goodwin says, you’ll want to be looking for vehicles showing a test date from within the last three years.

That won’t always mean buying a newer car though. Manufacturers will be able to apply for updated results, in which their older vehicles are re-tested in line with current standards.

This will help highlight the old and cheap, yet safer, cars that are still for sale. You’ll be able to do the following:

  • Compare car safety based on age. Look at two cars with the same testing year, and it means they’ve been through the same tests. A newer testing year will generally mean a safer car.
  • Pick out extra-safe older cars. If an old car punches above its weight, a manufacturer might apply to update the test. An old car with a high score on a recent test year is probably safer than an old car with a high score on an old test year.

If safety standards are improving, why is the road toll going up?

From next year, it will be more difficult than ever to get a 5-star rating – but that doesn’t mean older cars are off the road.

“No one expected that,” said Goodwin. “That car was not unsafe in its day.”

He was standing in front of scenes from the video you see below. It shows a 1998 and a 2015 Corolla in a head-on collision. The dummy in the 98 Corolla would have died on the scene, while the one in the 2015 Corolla would probably have walked away.

That head-on Corolla test was very different in two ways from the SUV crash we witnessed.

First, it involved two real cars going head to head, rather than just one slamming itself into a wall. Destroying two cars for every test would be needlessly expensive and the results inconsistent.

Second, the now-trashed 98 Corolla was an old car bought second hand from a public dealership for $5,000 – for “probably a bit too much” in Goodwin’s words. For the most part, Crash Labs destroys brand new market-ready vehicles instead.

The numbers are clear.

  • The average age of vehicles on Australian roads is 9.8 years, but the average age of cars involved in fatal accidents is 12.9 years.
  • Old cars built before the year 2000 are involved in 33% of road fatalities, but only account for 20% of the fleet.
  • New vehicles built in 2011-2016 account for 31% of the fleet but are only involved in 13% of fatalities.
  • In 2015, the fatal accident rate among the oldest cars (built prior to 2000) was four times higher than the newest cars (built 2011 or later).

Incentivised to drive dangerously

Somewhat ironically given its function, car insurance clearly incentivises younger drivers to buy old and cheap. Inexperienced drivers might pay around twice as much for cover than better drivers.

This translates into more numerous and serious accidents, which lead to higher apparent risk, which results in even higher premiums. This means even more incentive to buy a cheaper car. It’s a vicious cycle.

The good news is that you probably won’t die in an accident no matter how old the car is. The bad news is that you’re about 100 times more likely to be left seriously injured or permanently disabled instead.

More than 100 Australians are killed in car crashes every month, but about the same number are seriously injured in a crash every single day.

This may also compound the economic cost of road accidents. When a younger driver is seriously injured, the lifetime economic costs, in terms of care needed and lost productivity over the years, are much higher than for older drivers.

ANCAP, crash tests and safety

Most of what happened in the crash test we saw occurred too fast to follow with the naked eye. Even at the sedate-by-freeway-standards speed of 64km/h, it was over almost instantly.

Cars have been operating beyond the physical limits of human ability for a long time as have features such as seatbelt pretensioners and airbags.

But there’s only so much they can do. If two cars have a head-on collision at highway speeds, someone’s almost certainly going to die, no matter how safe their car is. And so the next frontier is accident prevention technology to stop accidents before they happen.

But not everyone is keen on letting their car take the wheel. ANCAP expects increasingly autonomous safety features like lane correction and automatic emergency braking to be a divisive issue.

Against all statistical evidence, there will always be those who doubt that autonomous systems could be better drivers than they are.

Are they right? Probably not. As Goodwin points out, in a mathematically impossible turn, most people reckon they’re better-than-average drivers.

And what do we know anyway? We’re still incentivising unsafe drivers to choose the most dangerous cars available and then marvelling at the rising fatality rate. For road safety at least, the robots can’t take over soon enough.

ANCAP expenditure

During 2016 and 2017, ANCAP’s crash testing generated the following costs:

ANCAP’s scrap yard

To ascertain 2016-17 crash test ratings, ANCAP and NCAP completely destroyed 142 vehicles. Across the six test labs used, 378 dummies met their demise.

ANCAP worldwide sister organisations

ANCAP is part of the broader New Car Assessment Program. It oversees ANCAP’s sister organisations, including the Russian ARCAP, India’s BNVSAP, C-NCAP in China, Euro NCAP, IIHS (United States), JNCAP in Japan, KNCAP in Korea and the Latin NCAP of South America. By working as part of this larger consortium, ANCAP is able to publish twice as many ratings per year.

Compare different vehicles

Name Product Boot displacement ANCAP rating Price (from) Finder Score
Volkswagen Golf
5 stars
Mazda 3
5 stars
Mazda 2
Mazda 2
5 stars
Hyundai i30
Green Company
Hyundai i30
5 stars
Ford Focus
5 stars
Holden Astra
Holden Astra
5 stars
Toyota Corolla
5 stars
Volkswagen Polo
5 stars
Honda Civic
Honda Civic
5 stars
Subaru Impreza
5 stars
Suzuki Swift
Suzuki Swift
5 stars
Nissan Leaf
5 stars
Name Product Number of seats Boot displacement ANCAP rating Price (from) Finder Score
Mazda CX-5
Mazda CX-5
5 seats
5 stars
Toyota RAV4
5 seats
5 stars
Hyundai Tucson
Green Company
Hyundai Tucson
5 seats
5 stars
Kia Sportage
Kia Sportage
5 seats
5 stars
Nissan X-Trail
Nissan X-Trail
7 seats
5 stars
Jeep Cherokee
5 seats
5 stars
Honda CR-V
5 seats
5 stars
Mitsubishi Outlander
Mitsubishi Outlander
7 seats
5 stars
Renault Koleos
5 seats
5 stars
Mitsubishi ASX
5 seats
5 stars
Name Product Payload ANCAP rating Price (from) Finder Score
Ram 1500
845 kg
5 stars
Ford Ranger
945 kg
5 stars
Volkswagen Amarok
1,020 kg
5 stars
Isuzu D-Max
935 kg
5 stars
Mazda BT-50
1,039 kg
5 stars
Mitsubishi Triton
855 kg
5 stars
Holden Colorado
1,029 kg
5 stars
Nissan Navara
1,013 kg
5 stars
Toyota HiLux
995 kg
5 stars
1,050 kg
5 stars
Mercedes-Benz X-Class
1,188 kg
5 stars

More guides on Finder

Ask a Question

You are about to post a question on

  • Do not enter personal information (eg. surname, phone number, bank details) as your question will be made public
  • is a financial comparison and information service, not a bank or product provider
  • We cannot provide you with personal advice or recommendations
  • Your answer might already be waiting – check previous questions below to see if yours has already been asked

Finder only provides general advice and factual information, so consider your own circumstances, or seek advice before you decide to act on our content. By submitting a question, you're accepting our 1. Terms Of Service and 6. Finder Group Privacy & Cookies Policy.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Go to site