Here's another common misconception ready to be exposed: old cars aren't safer. By older cars it is typically meant late 60's and prior, and the thought goes that because they are heavier, they are safer. This is absolutely false, weight doesn't matter in vehicle safety; what really matters is the ability of the frame to absorb impact while maintaining structural integrity of the passenger space. The safety of a vehicle accident is very simple physics: by spreading an impact over time and distance the force of impact is also minimized. This matches our intuitive understanding, just imagine dropping an egg from 10 feet. If the egg hits concrete, it breaks--it goes from speed to stopped instantaneously. If the egg hits 5 feet of padding, it will be fine--it will slowly go from speed to stopped over time and distance, as the padding absorbs energy from the moving egg. The "padding" in a car accident is mainly of one form, the crumpling of steel. Of course this is all moot if the passenger cabin is compromised, as our soft, fragile bodies are no match for hard things moving at high speeds, and that's the rub; no matter how much energy is absorbed, if the engine block ends up in the driver's seat or the vehicle explodes, there is little hope of walking away from the accident. Excluding air bags, seat belts, and other obvious safety features, modern cars still have the advantage because they are designed to crumple up to the cabin, which is in turn designed to be as rigid as possible. As far as I know, older cars weren't designed with any energy absorption in consideration, and thus a double edged sword: if the car doesn't crumple at all, no energy is absorbed and it is like an egg hitting concrete; if the car does crumple, it will most likely continue to crumple well past the engine bay and into the cabin, rendering all energy absorption for naught. And now, for the demonstration:
The cars collided each moving at around 40 mph. As is usual for the Internet, a number of people (I think it's safe to assume they are classic car enthusiasts) have stepped forward challenging the veracity of the video, suggesting that the chosen car was not representative. As in any scientific pursuit, contentions are often valid and desired, and the responsible scientist will acknowledge, explore, and respond to any valid concerns. I found this on Consumerist, where amongst the comments arose criticism (from user Nighthawke) as follows: the appearance of reddish dust that may indicate the presence of structure-compromising rust; the lack of seat belts used in the test vehicle (which were available as a dealer option); The expense of an unsafe frame for the aesthetics of the curved front pillar specific to Bell-Airs; Finally, the frontal offset test is unfair because the skinny engine didn't have the opportunity to absorb energy.
The responses are easy, as only the first point is really valid. The IIHS, which conducted the test, assured that the "rust" was just accumulated dirt and the car appeared structurally sound. For the final 3: the optional seat belts were lap belts only, and almost certainly wouldn't have made a modicum of difference; perhaps the Bell-Air has uncommonly poor structural integrity (I'm not sure, but I know of other classic cars with the same pillar shape), but the whole point is to show that collision safety design has improved tremendously and no other modern American car has performed anywhere near as bad as this Bell-Air; Last but not least, life isn't fair and the frontal offset test is one of a scant few standard tests that all cars undergo. Likewise, it's an important test for how common this type of accident is; James Dean died in a frontal offset collision. Frontal offsets have a particular propensity to cause extensive damage--the energy of the collision is focused on a smaller portion of the vehicle, thus causing more damage. In fact, in terms of energy absorption, a direct, in-line/"head to head" collision is safer! Clearly our intuition begins to fail us at this point, our instinct even more so; two people destined for a head on collision will swerve, unfortunately magnifying the danger of the impact be reducing the surface area of the collision. Nonetheless, this idea of applying a force over increased surface area is one that is often understood (or at least utilized) by people using snowshoes. This same principle is what allows people to lay on a bed of nails.
At least we have some kind of standards! Check out other poorly fairing vehicles here and here. In closing, I want to point out that heavy modern cars aren't safer either--in fact many large vehicles (trucks, SUVs) fare worse in passenger protection than smaller vehicles for a few fairly obvious reasons. Also, it's a matter of perspective: presume large vehicles are safer for the occupants, what about the people in any smaller car that may be hit? You'll probably walk away from your Suburban with a few scratches, but how will you feel about having possibly killed several or all of the people in that Yaris? The truth is, large vehicles aren't safer, they're more dangerous for everyone. The only reason huge cars can be viewed as safe is because there are other huge cars out there, and that's just an unsustainable and foolish perspective--keep it going and before long we're all driving monster trucks. Unfortunately even that won't help just as our huge SUVs haven't helped because more and more people will be getting injured in single-vehicle rollovers.
Certainly cars have gotten a lot safer, but as long as they are being driven by people, they will never be safe enough.