Despite the neatness of the Hyperbolic Discounting idea in the first answer, I think it extremely unlikely that anyone would find putting a seatbelt on so much of a hassle that the reward gained from avoiding it is worth anything at all, let alone worth taking a known risk.
There are much more nuanced theories as to why people take risks that would cover this situation better.
One that comes to mind is Sykes and Matza's neutralization theory (originally developed to explain moral shifting) applied here http://onlinelibrary.wiley.com/doi/10.1080/0007131032000045888/full to denial of risk from cannabis, but applied in other areas of risk perception also.
Perhaps even more likely here is the concept of Risk Denial as part of a construction of a safe world view as in this study for example http://journals.sagepub.com/doi/abs/10.1177/0022022191221006 . Essentially what happens here is that people fail to mitigate risks (such as with driving in the OP's example) in order to help them construct a safer world view in which the risk does not exist, by acting as if the risk were not there at all.
Here is an article relating the issue to smokers, but the same research has been done on AIDS denial, Adolescent risk-taking, and heart-disease.
The same has been posited as a reason why it has proven so difficult to establish proper hand washing regimes in high risk areas (such a hospitals and day-care). The theory being that if the practitioners there could not easily cope with the reality of working in an environment with such high (and importantly, unobservable) risks, so they have to act as if they weren't there as a coping mechanism. Again, hyperbolic discounting (the hassle) has been cited as the most lilkley reason, but research such as this shows it to have very little statistical influence. Video capture of hand-washing practises in hospitals with sensors have even shown people crawling under the sensors supposed to detect whether they've washed their hands properly or deliberately turning on the water to activate the sensor but not actually washing their hands.
It obviously would show considerable weakness of character (by society's standard, not mine) to admit this inability to cope so it rarely crops up in self-reporting. I personally wouldn't take the agreement with the existence of the risk verbally as anything other than superficial.
It is likely that people who take risks such as not wearing a seatbelt simply can't cope with the risk of driving and so have to deny it exists and reinforce that denial by their actions.