more random speculating whilst gesticulating wildly

Marriage
I was trying to assemble a list of reasons why people get married and promptly fail.
1) They are in love (now). Please let me know how that works out in 10-20 years. Or even next year. To use an economic analogy, home prices were going up in 2006. These trends do not always continue. Without some evidentiary support to suggest that it will (such as some active and long-term compatibility, often referred to as a sort of chemistry though it seems to resemble alchemy more than any scientific methods), colour me unimpressed with this reason.
2) They have a child. I suppose this is supposed to be an admirable sacrifice. Though in practice I think people should want to get married first and then want to have children (or simply have children and deal with that scenario, since it is difficult for me to imagine scenarios where people will actually want to have children). The order here seems sort of important if for no other reason than having a relationship solely predicated on raising a biological child together produced by at least one sexual consort together doesn't seem like a relationship at all. Note: I do think that people who sire children should support their offspring. I don't think this requires them to be active partners with each other so much as with the child. To me this reason seems more like laziness in selecting sexual partners more than dedication to children as if we were really "dedicated" to the children, we would seek to have productive relationships surrounding their formative years rather than to simply force ourselves to have relations.
3) Loneliness. I guess I recognize that people want to have companions. But desperation isn't really an inspiring notion to do something positive like start a family or find good companionship.
4) Financial reasons. Tax breaks and some balancing for the fiscal checkbook seems like a great idea. Except that these seem like things that should happen as a matter of course for a positive relationship rather than act as incentives to form them. For whatever reasons, historically men have seemed to grow up when they settle down with a mate. I have found that I grow up more without them during the antebellum period so to speak as I analyze what happened. I could imagine this is a common problem in the modern world (it doesn't appear to be). In practice anyway, the incentive function of financial stability seems to blow up as something to be considered as a valid reason, given that financial instability (and incompatibility) is a key motivator to breaking up marriages.

So basically it looks like the only positive reason left out there is status signaling, ie, I am a serious adult willing to accept certain new responsibilities and (potentially) ready to start breeding, and this partner I have selected also meets those qualifications. There's also some sexual signaling going on, suggesting this partner I have selected will probably not want to sleep with you so move along. I guess those are good reasons.

----
Also trotting through my mind are the various smoking bans in restaurants with now a few years of worth to them. I went out on a limb and said two years ago that I wouldn't vote for these bans despite having a potential positive benefit in reducing annoyance from smoke related dining out. I have still not seen what exact positive benefits they have extracted, at least for the general public, though in addition I might consult the private property owners who run eating or drinking establishments. I will have to do some active research on this one. But my opinion then was something like this:

I don't mind anti-smoking laws personally because the secondhand effects are annoying, but they're offending my sense of liberty by restricting private entities from doing something by choice or voluntary arrangements. If we assume, as I think would be a safe assumption, that most restaurants/bars that allowed smoking would lose business from strong-anti-smoking advocates who might even agitate against the establishment, then they would only remain with a smoking section if they were making up the ground by allowing it. That is, that their smoking consumers are adding enough market value and business to compensate for annoyed consumers who take their business elsewhere. If we also assume that there are easily accessible alternatives, such as crossing a state line or private exemptions to the law, then the law becomes rather meaningless and possibly more dangerous as traveling drunken people are far more hazardous to other people than secondhand smoke. Mostly these sorts of laws (banning something that is marginally hazardous to the individual and very marginally hazardous under very specific circumstances that very rarely apply to anyone else) strike me as majoritarian moralizing rather than advancing some ethical cause. We have established, without real assessments, that smoking is bad for other people and that it imposes unfair costs on them. I agree here, in the same way that someone talking on their cellular phone in a movie theatre imposes a social cost. But we don't need legal bans to enforce every unwritten rule in society. Public campaigns stigmatizing smoking, sufficient complaints by employees and consumers, and loss of business by establishments that violate these unwritten rules and permit smoking seem like plenty of effective methods to control the behavior at a reasonable enough level to prevent any actual social damages to public health and sentiments (beyond annoyance and offense of an occasional patron or employee).

--- Drunk driving.
So it's been kicking around for about a month trying to figure out how best to deal with this. This seems like a basic summary of what we do.
1) Arrest and detain people who drink and drive, or at least who refuse to submit to breathalyzers under the assumption that they must be drunk.
2) Create random spot checks, particularly in major cities or major holiday seasons to seize such drivers (while inconveniencing many others, basically a form of security theatre)
3) Seize the license of proven offenders and presumably raise insurance requirements when that license suspension is lifted. Mark their license plates for additional police scrutiny.
4) Suggest alcohol rehabilitation, but don't really follow up on it thoroughly enough to insure genuine alcoholics receive treatment and reduce the relative danger they pose to that of some idiot who decided not to call a cab.
5) Limit the drinking age to 21
This seems like a basic summary of what some people want to do to improve the current system
1) Execute DUI offenders
2) Create automobiles that won't start without a breathalyzer function.
3) Reduce the drinking age to 18, reduce public consumption limits in a scaleable fashion between 18 and 21 (such as by allowing for home purchases but preventing bar or restaurant purchases), or eliminate it entirely
4) Target repeat offenders by seizing property (ie, car being driven) until compliance with alcohol related treatment is demonstrated and persistent.
5) Raise the excise taxes on alcohol.

I will assume that most people aware of me will know I think 1 & 2 are stupid and ineffectual solutions. It's possible you could use some sort of blood alcohol check installed on serial offenders as a first line of defence, but we already have that in the form of a different color on the plates. It seems far more sensible to prevent serial offenders from even getting into a car without considerable attention simply because this is most likely a classic 80-20 Pareto problem. Most drunk drivers do it a lot and are potentially a greater threat anyway for other social reasons (such as public health and disorderly behavior). Most others do so infrequently and violate our often arbitrary limitations at only marginal points and often with decent intention. And while I might agree that we can rather easily execute a few alcoholics without much consternation to the general society, we already are investing far too much energy into the death penalty for real scumbag things like torturing people to death and serial or child rapists/murderers. This energy does not come for free, it costs us the ability to monitor other criminal enterprises and thus to investigate, try, rehabilitate and/or incarcerate people who commit real crimes. A simple DUI event does not constitute a substantive crime in my view. What does? When someone uses a car and harms someone or damages public property. That does mean that it makes sense to warn people or to signal that public drunkenness in the form of operating a motor vehicle should not be tolerated. I think it is perfectly legitimate to treat that as a criminal act if you want to signal its importance. The real danger however is alcoholics. Who didn't and won't get the message. Investing more resources into punishing random people who screwed up and randomly hassling large numbers of civilians in the hopes of stirring up a few who are being idiotic and hazardous is distracting from effectively cleaning up people who seem to need significant help to keep themselves from becoming a (potential) hazard to the rest of us.

So how do you clean up those folks. Well, seizing the property involved is a helluva lot more likely to prevent someone from driving around than taking away a state issued piece of identification. It notifies enablers, in the event of a repeated incidence, that they should seek to help their friend. On some paternalistic libertarian grounds I have some grumbling to do with this idea, but realistically, dealing with a repeated DUI offender is sort of like the restriction on mental patients for firearms. It's a mental condition that requires some limitations on personal liberties to prevent public damages and harms and penalties for abusing those liberties to insure someone seeks help for themselves. Raising the alcohol tax and reducing the drinking age similarly insure that access to alcohol is not substantively affected for people who want to have some wine with their dinner or some beer for their party, but is a noticeable cost on people who need to have a beer when they wake up in the morning to deal with their hangover from having a (few) beers when they went to bed the night before. It also insures that college campuses are a more controlled environment rather than one which seeks to evade legal statutes and invites dangerous behavior. Sure some kids can get drunk legally this way, but it will cost them more to do it frequently and the variation of use can move, as it did when Prohibition was repealed, away from large quantities of harder liquor (bang for the buck is prized in an illicit market environment). This last effect doesn't have much immediate impact on DUI rates, but I would imagine that having some controlled and marginally monitored access to alcohol would cut down on rates of alcoholism in the future and thus lower our potential costs. (I'd have to check and see if this is in fact the case. England has really high rates of alcohol consumption among teenagers too, but I'm not sure that they have our problems with DUIs and alcoholism for adults).