Generally, I think the idea that the majority of people can vote for something and restrict freedoms of other people who opposed the idea is something completely crazy and it's so sad that a lot of people will, I'm sure, continue thinking that is the way to go.

Update: also, let me ask you, cos I'm not sure, why do you think it's a bad idea to argue with a libertarian?

There will always be some demand for violence (meaning the application or threat of force to cause other autonomous agents to act a certain way), and there will always be suppliers of violence. The question then becomes of market structure: do we want decentralized and competing suppliers of violence in a given geographic area, or a monopoly? Both can be Pareto-efficient, so that provides no guidance.

Having a monopoly supplier, however, gives that supplier an incentive to maximize the net economic value of its customers: this allows it to extract more of that value. Competing suppliers, however, have no such incentive: if one cultivates a client, there's no reason for the others not to swoop in and extract the surplus. It's a race to the economic bottom. If economic well-being and surplus are good, then having a competitive market in violence is bad.

Okay, so we have a monopoly supplier of violence. What then, though? How do we structure it internally? Well, the way it arises will be the way it's structured, regardless of what we might like in an imaginary world. And that will be some group of agents who have to negotiate contracts with each other and with the customers who are exchanging labor in exchange for violence.

What remains is squeezing out the information and liquidity gaps that make every market inefficient. People should have a clear set of rules that they can assume will be okay with the monopoly violence supplier: then they will not miscalculate when they do something thinking it will not cause violent reprisals. That's what we call rule of law.

That says nothing about what the contract is, though: it may be "everyone pays a flat fee," it may be commission based, it may be commission based on a sliding scale. Who should decide that? Well, you ultimately don't want internal factions to splinter off to compete with the original monopoly (after all, we like the monopoly! At least at the margins.) So you've got to set up an corporate structure where no large enough group of customers feel that they could have a better deal with another supplier in the market. Everyone needs equal input.

We've tended to solve that problem with majority vote.

On the other hand, when you don't have a monopoly, even if the other side is weaker than you are, you would still have to factor in your potential losses and see if you can strike a deal that would allow you to avoid violence and, thus, losses completely.

But you're always fighting over stuff and, moreover, people. In all the cases I can think of, violence-suppliers would look at all the people it can extract value from, look at each other, and then combine forces to extract the most value from people. The only time they'd fight each other is if someone miscalculated who had more capability to deliver violence. But either way, violence suppliers would either combine forces or get destroyed, ultimately leading to a monopoly and making violence a profit center.

On a side note, libertarians call violence suppliers "private protection agencies". Reflects their nature better, since we agreed that the equilibrium status quo would be no violence.

That's why "private protection agencies" really doesn't get at their nature: being plural is contradictory with having a single cartel supplier of violence. And it's hard to say "private from whom" when there's only one party you negotiate with. And when you're negotiating for protection from violence with the only likely source of that violence... the word racket comes to mind better than protection agency.

It's not some god in the sky that does. We do: not just a generic we, but most people reading this, you and me included. We choose to participate in the markets it establishes; it tells us what property rights are and says they're property rights because it says so and has the guns to back them up. The corporations we work for rely on it to break down people's doors when someone gets ahold of a next gen device that they misplaced, and we feel safe that the money we put into our banks is safe for us because the State will yell at the bank if it takes our money and pay for it if the bank loses it.

There's no reason a protection agency wouldn't do the same to establish its own legitimacy. It seems like a good business model, in fact: convince everyone that questioning it is morally bad, create some ideal of "citizenship" that amounts to brand loyalty to violence, and then expropriate value from a populace that's not only cowed but actively consenting.

Don't get me wrong: it gets my goat as much as yours that that's the situation. But I wasn't describing before how a hypothetical, ahistoric government-like entity might develop: I was offering a general theory of how it actually did. If it's right, the State isn't some aberration but an emergent, self organizing maelstrom that has gradually grown to envelope virtually all people in all places.

My issue with libertarianism as it exists in popular culture is that its idea of liberation seems to be getting the State to deign to rewrite property rights to privilege one part of society instead of the current one (that's all of electoral politics, actually). Once that's achieved, one gathers, we can all become happy and free consumers because a certain property right regime has become the law of the land, imposed from above by our now-purified and almighty State.

Events, power struggles as people rise and fall in strength are all likely. Economics is a dynamic process, modelling as a stable or convergent systems is an extremely naive approach.

Unfortunately, it looks like nobody explained this to criminal gangs. They obstinately use violence against each other, especially when no one gang has a near monopoly...

Could you explain this claim? I wouldn't agree with it if you we're talking about fulfilling the demand of other sectors (like music or food), so I'm curious why it would apply to violence.

Represent an economic transaction with {M, X, V}, where M is currency, X is some good or service, and V is violence.

For a music cd, there's three parties involved, though the third is sometimes not explicitly accounted for.

1:{-M, +X, 0}

2:{+(1-p)M, -X, 0}

3:{+pM, 0, V}

(aside: money, you see, plays a nice role of making it a 3-tuple instead of an infinite-tuple)

The first two elements of a tuple have the constraint that they must sum to nullity across all the other parties involved. You can't make labor or goods appear out of thin air.

Violence, however, doesn't have that constraint. You can always throw in more violence, and it's always a positive amount.

(Why doesn't 3 just set p = 1? It certainly could, but that isn't the optimal proportion to maximize sum(M). The incentive structure breaks down: Laffer bites with a vengeance. So there's some other optimal p 0<p<1 that maximizes sum(M) over all clients--what that value is is simply an empirical question.)

So let's look at a market of violence, with multiple suppliers:

1:{-(q+r)M', 0, 0}

2:{+qM', 0, V'}

3:{+rM', 0, V''}

where q + r = p' and M' is the total value that 1 provides at p' where p' maximizes M'.

This satisfies the required constraints, but it's not a stable equilibrium. q+r is that same optimal value p from before, but with a twist: it's not set by only one agent trying to maximize sum(M) but by two trying to maximize their own. There's no reason for 2 not to go for (q+e) instead of q, and the same is true of 3. Taking only q and r as varying qualities, and you rapidly end up at q + r = 1, which means no one gets anything.

In practice, what happens is that 2 and 3 have their own separate transaction since they can foresee that upcoming equilibrium:

2:{0, 0, V'}

3:{0, 0, V''}

That itself is unstable: assuming that V' > V'', the new equilibrium is

2:{0, 0, V'}

Leading to

1:{-p'M', 0, 0}

2:{+p'M', 0, V'}

(...let's be clear that I'm not calling libertarians pigs. The point is that you can't change a dedicated person's mind via argument, and neither I nor Shaw found the effort enjoyable or productive.)

I understand libertarianism and I very much appreciate that it is built on a foundation of reason. However, being an imperfect being, a wise man tempers his reason with empiricism.

But libertarians are the ones who want to take direct moral responsibility for their circumstances; it's statists who seek to outsource that responsibility to distant formal institutions or systems of pre-defined abstract rules.

This includes choosing what trade-offs to make in the interest of maintaining or creating specific relationships with others - this is what society is. But what you're referring to as "society" is merely a logical construct, abstracted so far from the actual circumstances of people's lives that it does not represent anything that really exists.

Actual societies only emerge from the substantive interactions of particular human beings. Suppressing those individuals' ability to freely negotiate the terms of their relationships by forcing them to adhere to generalized rules, promulgated in the abstract, actively damages society.

Every time I argue with a libertarian, it's an endless series of "No True Scotsman" arguments.

See how that works? "Your argument doesn't accurately represent my viewpoint" is what both No True Scotsman and straw men are about. So rather than opting out of an argument by calling "logical fallacy" on someone, why not clear up the definitions and be sure of what each other are talking about?