In the case of the car accident loss being left where it fell, the government is clearly failing to protect the rights of the victim. The goal of the government (under the aforementioned philosophy) is to secure individual's rights to life, liberty, and property. It would be ludicrous to suppose that this could be accomplished without torts (or likewise without the use of police power to try to prevent wrongful conduct)Even accepting that a tort system is a necessity, this still leaves a huge range of determinations to be made with respect to process and defaults. Default legal burdens of proof and persuasion, burden-shifting rules, rules of evidence and courtroom procedures, etc. And to suggest that only the victim's rights are implicated neglects the fact that the tortfeasor's property rights are equally implicated. A faulty or unfair process could easily result in an unjust invasion of his rights. Protecting everyone's rights is a complicated business, packed with value-laden judgments.
In fact, one of my (many) objection's to Nozick's theory is that he imagines there will be numerous regimes of justice developed by individuals and associations in the state of nature, and he is indifferent to which one prevails (which would occur through economic competition of the associations). Any rights-protecting regime is apparently as good as any other in his view. But process can be slanted heavily to favor particular groups and actions, and it seems to me fairly important that there be a conscious and rational effort to create the "best" system, rather than the one most appealing to those with the greatest economic power.
I sincerely believe that the principle is merely a backdoor approach to regulation, and that his use of the word libertarianism is a usurpation of the term (this is the slippery-slope argument that Sunstein says is silly). Many of his examples involve at least some components which are involuntary and do not allow for choice. My second point along this line is a much more interesting one. Can such policies really be implemented (meaningfully) without imposing upon the liberty of the parties involved?Sunstein's argument is fairly limited in this regard. I don't think he wants to get into the argument about whether or not programs like the 401k program should exist, but rather given that it already does exist wouldn't it be smarter to set the defaults in a manner that will produce the socially optimal result? We can get into the topic of what optimality means (and I will below), but let's assume Congress created the 401k program for some purpose and that purpose can be furthered by adjusting the default as Sunstein suggests. It would be a paternalistic change to an existing regime that creates no additional burden with respect to interfering in anyone's rights and that produces a better outcome. That seems fairly unobjectionable.
So it makes sense that we would be better off providing people with a rationally determined set of defaults to prevent their credulity mechanism from accepting a bunch of random crap. But this is harder than it sounds. I mean the 'determining a rational set of defaults' part. It can be done, but it requires a great deal of thoughtfulness, trepidation, and investigation. From my point of view, doing so through the auspices of modern democratic government virtually guarantees that we will botch the process.Modern American government warrants a healthy dose of cynicism, but this is more like nihilism. Our institutional system is a bit of mess, and in my view needs some significant reform. But that stands entirely aside from the discussion of libertarianism as far as I can tell. Even the classic night watchman government requires a structure of governance. And if that system of governance sucks, you kiss your night watchman government goodbye in short order.
This, in fact, has always been my biggest objection to libertarian theory. I simply cannot see how a libertarian government could ever remain so for any appreciable length of time. The initial structure of the government will not bind it in libertarian form, even if you had a constitution without amendment procedures. It was not a constitutional amendment that allowed the development of the administrative state in the US, but judicial interpretation. And it wasn't just uppity judges responsible for that (although I'm not sure it would be significant to my argument even if it were), but intense political pressure placed on them by political branches motivated by quite serious concerns about social stability and unrest (and rising socialist sentiment). Whether it's a Great Depression and populist outrage or rent-seeking businesses and other economically powerful interests, a state will always be subject to pressures (of varying intensities) to do non-libertarian things. The state will always have the power to do those things; it is inherent in being a state. That power cannot be ignored, and there is no structural way to make it go away. It just sits there begging to be used.
I have a strong suspicion that some of the policies necessary to keep those pressures at bay (to the greatest degree possible) are themselves non-libertarian (e.g. some level of redistributive tax policy and regulation political speech financing, etc.). In the end, I tend to believe that our rights (which may or may not be limited only to life, liberty, and property) are likely best protected by a state that overtly recognizes the threat of instability and subversion of state power and is structured to best address that threat, even if that structure is not a strictly libertarian one. The state may need to be more powerful and far reaching than a night watchman state in order to be a stable, free-standing system. What good is the night watchman if it ends up being overthrown by communists or turning into a plutocracy?