[mikeash]

I think a lot of the regulation makes more sense once you understand that the FAA prioritizes different lives differently.

Roughly, they consider these groups, from lowest to highest priority in terms of keeping them safe:

1. Pilots. These are the lowest priority because they're the most in control and best understand the risks. If a pilot wants to turn himself into a red splat, that's his own concern.

2. Knowledgeable passengers. These are people who may not be pilots but understand aviation to some extent and have an idea of what they're getting into. They can't necessarily evaluate all the risks completely, but they can do a pretty good job of it. The FAA can't directly determine this, of course, but they use "for hire" as a proxy: if you're just taking people for fun or as a favor, it's assumed they have some idea of what they're getting into. If they're outright hiring you to fly them around, it's assumed they don't.

3. Passengers who are random members of the general public. Most of these people know little about airplanes and about the risks. They can't be counted on to evaluate things for themselves. People paying random pilots/companies to fly them around are assumed to be in this category. (And people buying tickets on a regularly-scheduled airline flight are assumed to be even more random than people hiring a charter, for example.)

4. The general public on the ground. These people aren't even involved in the process and have no choice in the amount of risk they're exposed to.

This is why the requirements get steeper as you move from recreational flying to commercial flying. If you want to fly solo over empty land, they don't care too much if you get yourself killed. If you're going to carry a hundred vacation travelers who just want to get to the beach, things are more strict.

[microcolonel]

Sure, but no aviation company actively dabbles in endangering the life of its customers or members of the public through negligence anyway. In the case of an incident, settling with the victims is going to be enormously costly with or without aviation-specific regulations. Regulations that insure the NTSB has resources to follow up on incidents make a lot of sense, so do ones on accounting (in manufacturing and maintenance); but it's hard to imagine what positive effect the FAA could have outside of that.

[mikeash]

Plenty of companies have dabbled in it through lax maintenance, insufficient training, or crew overwork. It may be hard to find ones doing it now, but much of that is because of the FAA's push for safety.

(And I don't know that it is actually hard to find ones doing it now. Certainly the airlines are almost ludicrously safe, but the context here is GA, where things can be much more lax.)

[microcolonel]

Safety is overrated. In CA it makes sense to strive for it, and boy, it's safe enough that people are worried about radiation! But in GA I feel like a bit more risk is incurred for productivity and efficiency reasons. Some stupid risks are taken, but I figure that's because the market tolerates it. Smaller planes also tend to generate less risk on the ground. Who am I to say that they should be safer; people are happy to drive cars of all things, no matter how crazy they must be to undertake that risk.

[pdelbarba]

Statistically speaking light aircraft are closer to motorcycles in terms of fatalities. On the bright side though, you're much less likely to me maimed in an aircraft accident :)

When people get upset about the FAA's safety stance in GA, it has less to do with the part 61/91 regulations (airman certification/operating procedures) and more to do with the equipment certification standards. A common example is the reliance on vacuum driven gyroscopic instruments when MEMS technology provides better performance with a much lower probability of failure. Up until recently the FAA made it almost impossible to retrofit old aircraft with generic glass panel systems even though these systems provide vastly greater safety margins when used correctly. I personally have experienced a vacuum system failure, but luckily not while in IFR conditions.

Beyond instruments, the bulk of the fleet of 30-50 years old and beginning to show it's age, but it's prohibitively expensive to certify new designs that incorporate more safety features like CAPS (parachute systems), composite energy absorbing seats/fuselages or digital engine management systems. We're also still entirely reliant on leaded fuel (100LL) due to cost hurdles in certifying engines that can run on JetA or anything else for that matter.