Thursday, October 8, 2020

Liberty and State Capacity

That Cowen Post
In January of this year, Tyler Cowen wrote a provocative post on the relevance of recent research in institutional economics and economic history to the libertarian movement. In that post he referred to a school of thought that does not really exist with the intentionally clunky name “State Capacity Libertarianism”. I take it that this was to tempt readers to begin to form a more well-defined school of thought that represents an appropriate libertarian response to this trend in research, and a better name along with it. It has already had a number of rises, though mostly from policy wonks and economists. I’ve heard some people say that what he’s referring to has been self-identified as “neoliberalism”. But that term and its referents’ relation to libertarianism is even more fraught. I’m just going focus on what libertarian political philosophy, which is a fairly well-defined set of ideas, might learn from recent research in the state capacity vein.
    Cowen’s suggestion is that the typical libertarian approach of seeking to remove or otherwise consistently undermine and thwart governmental powers that are normatively speaking illegitimate, insofar as they violate the natural liberty of the individual – particularly her economic liberty – is probably not the best way to go. In short, the size and scope of government might be less important than the competency of government, at least when it comes to fostering growth and innovation. Libertarianism therefore needs to make exceptions to its first principles in order to accommodate state institutions that empirically turn out to be necessary for furthering economic growth. But the idea that liberty may sometimes come into conflict with economic growth or welfare generally is not a new or particularly interesting idea, and simply saying maybe that one should accommodate the other more than some subset of advocates say we should, is not very philosophically plausible; it’s ad hoc.
    The challenge that the idea of state capacity raises for libertarians which is most philosophically interesting is not that sometimes state coercion can enhance aggregate welfare. Rather, it is that a certain amount of unjust coercion may be essential to having any liberty that’s worth a damn; especially liberty that we can expect to enjoy for a long period of time.
    I don’t intend to simply buttress the truism of social contract theory. I do not think state capacity shows that whatever coercion is necessary must therefore be just. Rather, coercion that is in fact unjust might be something we need to find a way to ethically work around in ways that are not obvious or straightforward applications of the principles typically articulated in libertarian theories. State capacity does not legitimise the chains, but it does give us liberty-based reasons to hesitate before we break every last one.
    I think that state capacity confronts us with the dilemma of the extent to which we ought to comply with unjust institutions in the name of justice. Perhaps this is a species of a dirty hands/lesser evil/trolley problem, but it is one that I think demands that libertarian political philosophy take a little more seriously in political terms. Empirically, it may not be enough to know what policies involve injustice and which don’t in order to have a road map of a more just society. To beat a dead metaphor: we may have to take detours through valleys of injustice in order to arrive at plateaus of greater liberty. This does not mean rejecting what realists call “moralism”, or the idea that our political ends should ultimately be something that is morally good (what we normally call justice). I don’t think any of what I’ll say implies we should care less about justice. I think it implies that we need to take the invariably unjust realm of politics more seriously for the sake of justice. Jacob Levy has referred to something like this as “Non-Lockean Libertarianism” and later “Political Libertarianism”. I do not think that political libertarianism has to be non-Lockean in terms of its normative and metaphysical commitments, but I do think those commitments are not all that much use when it comes to working out how best to serve them in day-to-day politics, so it likely requires considerable change in the discourse employed by Lockeans.
    I don’t think this warrants a new school of thought, I think it just raises some challenges to libertarianism. In particular, it raises challenges to the theories of government that they think follow pretty straightforwardly from their abstract principles. It is the libertarian theories of what Chris Freiman has referred to as “ideal institutional theory” that I think bears serious reconsideration in light of these ideas. Namely: libertarian minarchism and anarchism. I think it leads us to question key parts of the platforms associated without those institutional ideals. It also raises challenges for natural rights or deontological libertarianism; ones it does in fact have the resources to deal with, but such resources have not yet been transformed into the requisite capital.
    To repeat: I do not take this to just empirically vindicate social contract theory. These methods of examining institutional development do not presume that peaceful cooperation does not or cannot occur outside of a state; or that only through a state can social life be fair or orderly Indeed, these approaches do not take the state to be – morally speaking – categorically distinct from the rest of society. They evaluate the state’s functionality and performance the same way they do any other social institution. The state is regarded in the same way it is by those at both ends of the radical political spectrum. From Charles Tilly to Murray Rothbard, we see that the state is simply an organised crime syndicate that has violently and financially triumphed over all the others. The state is regarded as the group of persons who at any given time hold a monopoly on violence. This means that they violently extract an income from the population (rather than generating one through work and trade). This monopoly, in part consists in the fact that it protect its population from outside invaders and internal plunderers who would challenge such a monopoly. The state is therefore a nexus of cooperators who extract rents from the rest of society, and divide it up among themselves bureaucratically and hierarchically. The account of liberal democracies one finds in this literature does not say that we have somehow transcended this form of political organisation; it just becomes more complex, sophisticated, and with that, more stably liberal. It remains a bleakly realistic view of what the state is: a predatory organisation controlled by a coalition of elites.
    With that rather convoluted prolegomena in mind, we can begin. In what follows I will try to give a decent characterisation from what I gleamed from the literature, and demonstrate the challenges it raises for libertarian institutional ideals.

What is State Capacity?
State capacity is when states have the capacity to do state-like things efficiently and in the long-term. States are legitimised monopolies on force. What they do is determine which acts of violence will be permitted. Through this function, they generate wealth and status for a class of beneficiaries. In order to provide lots of wealth and lots of status in a sustainable manner, history has shown that states protect what we think of as private rights of it subjects (freedom from assault and theft in the form of the law of torts, contracts, and property). The more they protect their subjects rights, the more space their subjects have to live their lives, which typically involves cooperation and trade which generates a social surplus that the state (the nexus of persons who cooperate to protect rights from insiders and outsiders) can help itself to a portion of.
    Capacity can be understood simply in terms of how much coercion the state can engage it (the sum total of its taxation, the size and strength of its armies or security forces), but this really misses the point of what – if anything – makes state capacity a good thing from a liberal perspective. In order for that coercion to be used in a net-beneficial way, state’s also need administrative capacity; they need dependable information about their subjects. Medieval states, for example, had very little administrative capacity. In order to draw revenues from their subjects they would have to impose a one-off levy on some particular good or activity, and the monarch would have people he knows personally go and make sure the money is raises and delivered to him. For more steady streams of revenue the monarch would create rents through protecting a monopoly in some good or service, and demanding tribute from the monopolist in return for their protection. For example, by creating an elite of landlords as in Norman England, or enforce usury laws against Christians but not Jews, and then tax the Jews on their interest earnings. The King’s ability to raise revenue depended on these fragile equilibria based personal relationship and ad hoc arrangements with intermediary groups such as religious minorities, guilds, the Church, aristocracy, etc.
    The instability of revenue for the state is bad for freedom in two interlocking ways. When the power of the state is used in unpredictable, haphazard ways, one cannot plan one’s life around it. One cannot yield long-term value from one’s freedom if one believes one’s freedom might be invaded without warning at any given moment. Recall Hayek’s remarks at the end of chapter 1 of his Constitution of Liberty in which he says that of course any state has to be coercive, and all coercion undermines freedom. Yet if that coercion is in accordance with general and known rules that one can predict the impact of, the freedom that remains can be made full use of. Whereas where there is no rule of law, even if the state is less coercive in sum, its effects on one’s freedom can be worse.
    Moreover, the instability of revenue itself means that the sovereign’s position as such is pegged to their latest tax, personal alliance, etc. When the King’s position is dependent on constantly shifting personal relationships and ad hoc arrangements to secure revenue, this renders his rule itself more fragile. A poor King is less likely to maintain influence, and therefore perceived legitimacy which are required to keep usurpers at bay. Having to depend upon all these outside contingencies meant the sovereign was not self-sufficient in terms of legitimisation; he depended upon the loyalty of the aristocracy, the Church, various independent military organisations, and his ability to return favours to them. This makes various kinds of coups, revolutions, invasions all the more possible and sometimes predictable, which in turns makes freedom more fragile. Not only do we not know how the King will interfere with us from day to day, but we cannot be sure that his cruel brother, or the French dauphin won’t take over tomorrow and change things all the more.
    Why didn’t he just impose an income tax or and VAT on all transactions, and sit on his thrown and try to spend the revenue prudently? Because he did not know how much anyone earned, he did not know how much tax they could afford, he did not know what transactions were happening, he did not know how many people lived in his realm. The Lords know their fiefdoms with somewhat more intimacy but would still depend on more lowly ranks of aristocrat for still more local regions of it. Individual money-lenders know their debtors. Bishops knew their priests, and priests knew their parishioners. Guild officials know their artisans. It was through these intermediaries that the King was able to make any use of information regarding his realm at all. This use of intermediaries limited the amount of coercive power the King could claim because they would jealously guard the small amount of localised power they have, moreover, the King’s financial ability to exert any force would be quickly checked if he undermined those intermediaries, since he cannot raise money without them. Even if the King had the sense and ability to just create a bureaucracy to see over a centralised fiscal system, he could not trust their loyalty to him; they might easily go into business for themselves or be captured by a rival for the crown or an intermediary group. It would just be one more organisation that he would have to tend to his personal relationship with. 
    Paradoxically, in order to have a lighter touch, states need to have significant fiscal capacity that can be directed from the centre. As is thematic in James C. Scott’s work, states must make their societies legible before they can plunder them in a sustainable way; that is, by taking a portion of the social surplus whilst undermining the conditions of the production of that surplus as little as possible. In his latest book Scott notes, for example, the imposition of fee-simple land tenure in the place of complex commons arrangements enabled the creation of simple lines of tax liability to individual heads of household. Scott argues that grain-based agriculture was conducive to fee-simple ownership and that is why ancient states only emerged where grain production was viable.
    Douglas North, John Wallis, and Barry Weingast discuss at great length how in pre-modern states – what they call “natural states” or “limited access orders”, there is no little to no freedom to create organisations; whether they be religious, associational, industrial, or political. Since the state depended upon particular organisations and personal relationships with their leaders, the state could not be counted on to automatically provide security to organisations created outside of its purview. No doubt there was social cooperation beyond the scope of the King and his coalition of rent-seekers, but they could not avail themselves of the enforcement of contracts and property rights, nor could they make use of the King’s courts. It is only when states get to a point at which they offer general protection to all organisations regardless of any personal relationship between the King and the organisations leadership that civil society and a market economy begins to proliferate. A social and political order that does not depend on social relationship requires a far more sophisticated and complex legal system – one that gives legal status to organisations or positions as such, rather than to individuals who happen to lead such organisation or occupy such positions. The generalised creation of legal personhood for all organisations, offices, and persons, tends to be a more robust equilibrium because it creates an environment of competition for political rents in accordance with known and enforced rules. An “open access order” begins to emerge where the coalition of elites can create it without threatening their current position of wealth or status; it typically involves the change of rents with high levels of protection from competition, but low levels of long-term security, for lower rents with higher levels of long -term security.
    In modern liberal democracies, rent-seeking takes place almost exclusively via the legislature, and there are set rules laid down for how to go about it – you cannot just make the King an offer. This is far more stable because it gives everyone a stake in a status quo the generally protects freedom by facilitating particular channels of legitimised and limited rent-seeking.
    Mark Koyama and Noel Johnson discuss how pre-modern states lacked the capability of recognising religious liberty. Their coercive capacity was such that they depended upon religion to legitimise themselves, which meant they depended on the Church to legitimise them. This meant they had act as the enforcement arm of the Church so that the Church in turn would tell the faithful that their rulers were Divinely ordained. This meant that states would engage in religious persecution when they were weak. The authors explain why states that gained administrative and coercive capacity as a result of military campaigns were the first to recognise religious liberty – because they could afford to. Without the need to go through religious intermediaries for legitimisation or to raise revenue, they did not need to depend upon group-based legal ordering. A rationalised fiscal and legal system does not need to know who is of what religion because it can hold a place for them simply as an abstract individual. Whereas in lieu of such a system, it is essential to categorise the population into religious groups in order to know who to persecute to get the Church on-side, and whose money lending they can impose large taxes on.
The move away from limited access orders and toward open access orders in North, Wallis, and Weingast, and that from identity rules to general rules in Johnson and Koyama are moves from situations in which political institutions are weak, but arbitrary and unpredictable, to one in which they are strong, but more liberal. Moreover, the latter are relatively more stable equilibria because there are positive feedback mechanisms between access to organisational protection and the conditions that maintain that, and high state capacity and general rules.
    In Acemoglu and Robinson’s latest book, The Narrow Corridor, the title describes the regression between strong civil societies and strong states. Robust protection of civil organisations, understood broadly to include all organisations and institutions not internal to the elite coalition that controls the state at any given time, stands in a virtuous circle with strong states. The combination of the two create robust liberalism. If civil society is too weak it is incorporated into the state by the elite coalition and you get a totalitarian government that subsumed all of society. If the state is too weak you don’t get the development of general legal forms that are required for the long-term planning required for socioeconomic development. In the middle you have the “Shackled Leviathan”. 
    The inconvenient truth for libertarians is that we must pick one point within this corridor, and there is no point within it where coercion is only ever used to defend against prior coercion. Nonetheless, the maintenance of a less than perfectly just system of coercion – often through unjust acts of coercion – is necessary to there being any freedom that is of any value at all.
    The first problem is that the path taken toward the open access order involves serious injustice. Whether it be the war-making that selects for states that can raise revenues to fund them; the imposition of foreign legal forms on peripheral communities that are necessary for legal rationalisation, as well as any other coercion used by the central state to over-correct for local tyrannies. The second problem is that the institutional mix that is necessary for the preservation of a system in which individuals have secure private rights involves institutions that many libertarians see as at odds with private rights. Of course, different open-access orders have different kinds of mass democracy, different kinds of welfare states, and different degrees of regulatory rent-seeking, etc. But if these authors are right, you cannot have a self-sustaining market economy without any mass democracy, welfare spending (which rises as a proportion of GDP), and systematic rent-seeking.
    Firstly, it is by funnelling rent-seeking activities through legal channels that maintains the security of the market economy (by stopping random plunder and unpredictable intervention or cartelisation), and the state itself. If such channels were not available through the state elites would create them extra-legally, which would threaten the status of the state as legitimate monopoliser. Mass elections as well as freedom to create political parties and organisations generates legitimised and more peaceful avenues by which to challenge the elite’s rule at any given time. This forces the ruling coalition to act as if they were more long-termist than they actually are. They cannot create new rents because if they do not get re-elected these gains will be transferred to their rivals.
    North, Wallis, and Weingast also argue that a welfare state that expands as a proportion of GDP is also an interlocking component of the open-access order. At the primary level public works are necessary to integrate peripheral communities into the national economy. (Such economic integration is necessary so that everyone is interdependent and hence limiting the incentive to create rents that would be net wealth destroyers over time) This means building roads, sewage systems, public transport, electric grids, single currencies and all the other “distance destroying” goods of state-building James C. Scott talks about (or what neo-Marxists call the “annihilation of space by time”). These public goods can probably often be provided on a market, but when the state undertakes it jump-starts the integration of a large society into an interdependent market by taking on and socialising transactions costs that might otherwise be hurdles.
Insurance against bad luck and bad decisions in the market (including the failure to insure oneself), North, Wallis, and Weingast also believe is an essential thing for the state to do. Firstly to avoid the creation of a class of people excluded from the gains experienced in the rest of society by the market order. Everyone needs to be given a stake in the social surplus. The second is that socioeconomic inclusion is necessary to give visual reality to rather abstract and invisible notion of legal equality. Ideationally speaking, legal equality may not be stable without being able to see it through some parity of material advantage.
    It is the rent-seeking that maintains the elite’s sake in the market economy, and it is the market economy that funds the state, and it is the state’s legal apparatus that provides the rails of our socioeconomic liberty (such as it is), and secures the rights of anyone to participate in electoral democracy, it is that electoral competition that makes the political elite maintain a system that strongly limits rent creation, and it is the welfare state that ensures the masses have a stake in the market economy, and ensures that – ideationally speaking – our legal equality is buttressed by a degree of socioeconomic inclusion. The “double-balance” of the civil society and political elites via the democratic process maintains a liberal order.
    Given the level of abstractness and generality involved, states with the rule of law tend to be relatively liberal, and if we want to maintain it we need to look at what keeps it healthy. The fact that rent-seeking (but not rent-creation), electoral democracy, a growing welfare state, appear to be part of that, as well as the dependency that a path be taken via warfare, is quite disturbing. Not disturbing because it tells us that the ideal society as we see it may not be feasible (we probably always knew that), but that taking steps toward our ideal society may often undermine the conditions for liberty’s best chance: actually-existing liberal democracy.
    The point is not – at least not always – that we need high state capacity and open access in order to have private rights; but rather, if we don’t have the former, we won’t keep the latter for very long. Whether or not state capacity is the cause of socioeconomic development or merely a correlative filter, the point is that if we do not maintain this kind of government, we may get something far worse.

The Challenge For the Minimal State
Minarchist libertarians argue that we should have a Weberian government, but it should be limited to the narrow functions of national defence, and enforcement of the law of property, contracts, and torts. That is to say only ever uses coercion to protect innocent individuals from coercion. One often hears people describe such a regime as one of “small” or “limited” government. And the temptation for minarchists would be to say “We never said the state should be weak, of course it must be strong, but nonetheless minimal it its functions.” The problem is that that kind of state just does not appear to be a stable equilibrium. The provision of a good system of property, contracts, and torts, combined with efficient taxation, is probably not incentive compatible with the state not also being a vehicle for rent-extraction for private parties. Nor is it likely that a state of this form could emerge without engaging in any of the activities states do when they are trying to create a nationally integrated market economy, like provide public goods that connect the periphery to the core. How did the ruling coalition come to control the state without any promises to special interests that they will share in some of the benefits of the state’s fiscal power; how is the state legitimised when it appears to sustain an economy the benefits some wildly more that it benefits others?
    Many minarchists will accept all this and say they never expected their ideal to be a reality, for precisely these kinds of empirical facts about human motivation. Nonetheless, they would maintain, this is the ideal of government we should pursue. We should fight every tax increase, support every tax cut, fight to abolish the welfare state and privatise all public goods. Abstain from all partisan involvement because politicians are nothing but criminals, whether they’re on the blue team or the red team. So on and so forth. The problem is that is that this kind of approach will not only fail to achieve the ideal result, but it will probably achieve worse results even relative to that ideal.
    Government will never stop doing all the things minarchists want them to stop doing. But if minarchists keep fighting the way they do, it just guarantees that all these things are done badly. 
    As Jacob Levy has argued at length, drawing on Benjamin Constant, this political abstention also opens up the way for political leadership of populist strongmen who promise to “cut through all the bullshit politics”. Meanwhile in Trump’s America you still have a welfare state, you still have rent-seeking, you still have public debt, and the public education system is still awful (and is here to stay), the welfare system is still awful (and is here to stay) there are still people who cannot afford basic healthcare, and the country is still at war. It is just that now all the avenues through which those things could be fixed are compromised by Trump’s defiance of constitutional and democratic norms.
    Viewing the government as an illegitimate criminal enterprise is the (partially) correct normative position, but that should not make us think that we ought to deal with it on its own terms. Given that the ideal limited state is impossible, we should remember that far worse governments than the ones we have are imminently possible, and we should work to make our own governments better, which means, in part supporting the continued existence as the kinds of things they actually are. Rent-seeking and the welfare state are endogenous to the kind of society in which we want to live, so we should work out how to make them better and more conducive to liberty rather than act as if we can make them go away when we know we can’t.
   The view that the minarchist implicitly holds (though would probably deny holding when pushed) is that the state’s function is to set up the conditions for a socially beneficial spontaneous order, but is itself in a way insulated from that order. But the state is co-constitutive of the market and civil society. To have a theory of how the ideal government ought to look that is based on all the arrows of causation going away from government is obviously flawed. As yet, minarchism does not include a theory of how a free market order can generate the political conditions for its own maintenance.

The Challenge for Anarchism
Anarchist libertarians believe either that the market-place will select for decent rights enforcement without the need to coercively impose one form upon everyone (or at least, they it could so select, perhaps with some concerted help from conscientious actors), or that norms that end up protecting rights will emerge from consensual cooperation. Obviously, not all emergent institutions are good for freedom. And exogenously imposed ones might be better than the ones that have the best chance of emerging.
   Asserting that we have to wait for people to consent before we impose any coercive constraints on them certainly ensures that those constraints respects their rights, but it does not guarantee that their rights are respected. Non-consent based social contract theory argues that it is a guarantee that their rights will not be respected. Though they often put this too strongly (namely, when they say that right action is conceptually impossible outside of a state), I think empirically it is probably mostly true. 
    Monopolies on the use of violence can be incredibly dangerous, but sometimes they develop in ways which are far less dangerous than others. In order to make their channels of rent extraction and status conferment sustainable over time they administer violence rather than merely exerting it. Franz Opppenheimer, the anarchist sociologist who would be an enormous influence on Murray Rothbard, referred to a particular stage of state consolidation at which point that state’s power transitions from “violence” to “might.” Others have called it the “rule by laws not by men”, Kant made the distinction between coercion being “unilateral” and “omnilateral”. The distinction is basically between when force is exerted in order for certain people to benefit and get their way, and they get to exert it pretty much as they please, in a way that is highly unpredictable in the long term. You might know that Mary I is going to kill protestants, but you don’t know what will happen after she dies, or if she fails to produce a male heir, or if her marriage to Phillip of Spain doesn’t work out, and so forth. You don’t know what will count as heresy from one day to the next. Moreover, you don’t know which protestants will be deemed the greatest personal threat and therefore be persecuted. On the other hand, you have codified, formalised use of coercion. Though there is the constant and very real threat of coercion carried out by individuals (and we should not pretend political society ever transcends that), it is predictable, we know how to avoid it, and we have relative trust that others won’t use force on us except where permitted or required by law. It is coercion in accordance with general rules applying to everyone, rather than coercion in accordance with some set of elites will. Of course, there may be an extent to which general laws represent the will of the ruling elite, but it is important that their will is now refracted through general laws that will pretty much be applied independent of their particular changing will over how they should operate. (If the research done in the state capacity literature had been done by legal historians or jurists, rather than economists, it would no doubt have been tacked onto the literature on the conditions for the rule of law).
    What this does is it opens up pathways of free action in which we have trust where we won’t be subjected to coercion, so we will always have some recourse if someone breaks the law and interferes with that sphere of freedom, such as it is. It is not just that we are able to evade coercion, it is that we know where the coercion is, and plan our lives around it, and so derive a lot of value from it.
    Now, as we all know, the rule of law is more or less a formal characteristic of a state’s legal system, rather than a substantive one. Rule of law is, in principle, compatible with laws being horrendously illiberal. So, monopolies on the use of force, even if they do generate a situation in which the law is known to everyone, demands high compliance, which in turn generates trust, it will always involve at least some legitimised injustice. This is bad. In some ways, worse than injustice that is not legitimised since. The legitimisation of injustice adds insult to injury by constituting an expressive injustice on top.
    Whilst it is in principle possible for there to be reliably minimal violence without any monopolisation of violence, no one has yet been able to explain the institutional conditions that would maintain it and reproduce it. The essential feature of anarchist legal order is that if anyone attempts to monopolise violence, there is always legal space for challenge to it. This is what defines a lack of a legal monopoly – the omnipresent possibility of legal appeal. However, just because the legal order itself renders it permissible to resist monopolisation this does not mean monopolisation is rendered impossible or unlikely. Obviously, since states just are the result of monopolisation where there was none before. However, the very essence of anarchic legal order is openness to challenge. No monopoly means no final pronouncement on law. Therefore, that the order remain open to challenge is itself open to challenge. If one is concerned with a just anarchy and not merely anarchy per se, it is surely important than when justice is done, there be no further challenge that is legally possible.
    As my colleague Carmen Pavel has recently been arguing, whilst law per say can be grounded in consent, rule of law cannot. For the simple reason that it provides no legal rules with regard to how do deal with unjust non-consenters. They will either have to be coercively subjected to just law (which ought not be subject to legal challenge) or they will have to be dealt with extra-legally or not at all, meaning no rule of law. There must be a residual system of coercion against which consent enables you to make departure from, and that just is a legal monopoly on force. I don’t doubt that there can be anarchist law, but I doubt that there can be rule of law such that each individual is guaranteed equal standing before it, under anarchy.
    If I’m right, what does that mean? It means that a stable, liberal anarchy is out of reach, and is therefore not something we should reach for. This means lowering our time preference for social change. Making states better often means making them, not only bigger, but more permanent. For example, American anarchists believe that protecting private gun ownership is essential for (inter alia) strategic reasons – it maintains an entry point for disrupting the state’s monopolisation of violence. So, if anarchy is the goal, private gun ownership must remain legal. Releasing ourselves from the idea that the institutional ideal must be one that is stateless allows to consider the possibility that prohibition of private gun ownership might be something that increases state capacity in a way that makes the state more permanent, but also society less violent. (I’m not staking out any position on the empirics of gun control in the US here, I’m only speaking conditionally)   
    The political and social conditions of trust make the state’s existence more permanent and robust, and can also make governance much more liberal. Actively undermining the perceived legitimacy of the state would obviously erode this without yielding anything like a robustly liberal stateless world. Attempts to achieve anarchy on the pretence that only a world without a state can be just will typically not only not only fail to generate anarchy (least of all, just anarchy) but will also make government worse by liberal standards.
    The state monopolises violence, and permits only non-violent challenges to its rule. Enabling the development of state capacity does not mean leaving its power unchallenged, it means constantly challenging it to be more liberal, but not in ways that undermine its ability to maintain a background of rough liberality.

The Challenge for Natural Rights Theory
Natural rights libertarianism lacks a complete or adequate theory of government. I don’t think this is a flaw, it’s just not that kind of theory. Natural rights theories are theories of right action, the necessary and sufficient conditions for right action, and the conceptual and normative structure of their order. A theory of government would be a theory of the institutional conditions for the facilitation or enforcement of right action. Often natural rights theories are held up as if, since they tell us what right action is, they tell us by extension how government must be constituted since ultimately government is constituted by human actions that should be regulated by right. However, if certain amounts of injustice are necessary to have any meaningful justice at all, I think theories of natural rights will need to account for this. If the institutional conditions under which right action can robustly proceed are ones in which certain wrongful actions must also proceed, this muddies the picture.
    Moreover, even for those natural rights theorists that are comfortable deriving a certain form of government in perfect accordance with right action, like minarchists, they will still need to account for the actions that take place within the realm of the legitimate state. How much taxation, what should be taxed, what is the role of partisanship, what is the legislature for since we already know which actions should be illegal, should judiciaries be independent? Even if one can derive a state with clean hands, all of these questions surely require answers that are informed by justice, even though they clearly cannot be mere applications of justice.
    One could plug in certain outputs that a state must have, say “justice”, and say whatever is most efficient at bringing about those outputs counts as legitimate. However, it’s not obvious that whatever individual actions by citizens, politicians, etc. result in just outputs from politics are compatible with the one’s that will reproduce the system that keeps the worst injustice at bay over very long periods of time for very large populations. For example, ending mass incarceration in the United States is no doubt a demand of justice, but there are definitely ways that one could achieve that which are not advisable if we have a commitment to the institutional conditions for robust and long-term liberalism, such as it exists. Violating all the informal norms of democracy could be one way of ending the war on drugs, but if it chipped away the US state capacity too badly, it would not be the right way to go. Given the institutional complexities, politics might not be the kind of machine where you can ensure justice outputs by just inputs, and vice-a-versa. We may need principles for running the machine itself that are informed by justice but not mere applications of it to myopic human actions.
    I think Kant may have been on the right track in the Doctrine of Right when attempting to derive principles of government from individual rights, without those principles simply being applications of rights to particular cases. Infamously, the relationship between public right and private right is not very clear, and when it’s made clear it’s not very plausible. But the need for principles of government being grounded in the insecurity of rights without settled rule of law is definitely the right starting point, I think. Good principles of government are not going to make it impossible for governments to commit injustice, but that doesn’t mean they can’t be morally informed by justice. (When I say “principles” I mean that in the broadest sense; just a framework for understanding what makes government good in the light of justice. I don’t mean to say it has to be all a priori and we shouldn’t use empirics – anything but!)
Prospect
Will Wilkinson and Tyler Cowen have said that the core of libertarian political philosophy collapses in the face of the empirics about the political conditions of freedom. The fact that freedom from constraints imposed by other people’s actions is best secured by certain kinds of institutions leads them to think that what we thought was negative liberty is in fact positive liberty. If we want to be left alone, we need certain institutions in place. Therefore, we cannot be said to have a right to negative liberty without it being that case that we have a right to whatever positive conduct is required by others to create and sustain these institutions. But conceptually speaking, it can be true that we have a right to non-interference, but fairly minimal (if any) rights to positive assistance, whilst it still being true that the presence of institutions is the best way to constrain people’s actions such that they respect one another’s negative liberty. There is a distinction between the content of a right and the conditions required for securing, enforcing, or enjoying the right. Negative libertarians have always believed rights need to be enforced.
    I don’t think the experience of state capacity indicts the conceptual content of libertarian political philosophy. It only means we need to alter the way we understand how to apply it. Political philosophers (as well as everyone else) are very quick to think that they are merely stating an abstract normative principle when really they are describing the conditions they think concretely realise that principles. Everyone – and political philosophers are not immune – make implicit judgements of social science when they articulate their political world view, conception of justice, etc.
    Libertarians are very good at identifying this tendency in their opponents. The clearest example is Rawls claim that regimes that do not intentionally aim at x necessarily are incompatible with x as an outcome, and therefore if justice demands x, that regime is unjust. This of course assumes that justice can never be an unintended consequence. Jason Lee Byas and I have described the tendency to regard principles of justice as necessarily demanding their realisation through the intentional enactment of explicit public policies in their name; we call this “the policy framework” and contrast it with “methodological anarchism.” However, libertarians – even anarchist ones – often have their own version of this. It is thought that justice must be the conscious and direct output of the ideal political regime, whether it be a nexus of overlapping rights enforcement agencies and the like or a minimal state. These are the political conditions we take to be necessary to freedom. But we neglect the conditions that are necessary to those political conditions. Libertarians have a lot to say about what the rules should be, but very little about how to maintain institutions and organisations that in turn sustain those rules.
    As far as I understand it, this is an important part of what the German ordoliberal tradition has always been. It takes a classical form of liberalism as a normative standard, but takes the construction of government, and its reciprocal relationship with the market and civil society to be the political-economic problem to be solved, and not one that can be solved by simply re-asserting the normative standard itself. Unlike the settler/frontier liberalism of Locke and Jefferson, it takes historically constructed and constantly evolving institutions as essential to freedom. Perhaps the future of libertarianism lies back in the “Old World.”