Should we privatise criminal justice? A response to Robin Hanson
A radical case for outsourcing the three Ps: policing, prosecution, and punishment
Amidst the recurring debate on optimal criminal justice policy, Robin Hanson resurrected his novel proposal for “vouching” last week, in the aftermath of the horrendous and barbaric murder of Iryna Zarutska. Under this system, everyone requires insurance for any crimes they may commit. As was the norm for most cultures in most of recorded history, upon conviction, the defendant pays a fine. If they cannot pay, the insurer pays, but punishes the defendant in lieu of his payment. Policing and prosecutions are outsourced to bounty hunters and private attorneys, yet the state maintains a monopoly on the courts and in setting the law. Read Hanson’s post for further details.
In other words: the privatisation of criminal justice. The rationale is the standard argument in favour of markets and competition, albeit in unchartered waters for the economics profession. People constantly decry our criminal justice system: police brutality, prison conditions and overcrowding, even recidivism outcomes - these are a tiny fragment of the issues that are highlighted. Why not let markets solve them?
Whilst we are on this subject, I even have a recent personal anecdote. I am currently blocked from any transactions or from using any cards linked to my bank account, on the basis of moving address. KYC and AML regulations outsource law enforcement to banks, yet banks hardly have a comparative advantage in this industry. A more efficient division of labour, if the state is ineffectual (and the need to outsource law enforcement in the first place suggests such), would surely outsource law enforcement to specialists in that field? Unfortunately, my passport was up for renewal last month, so I have no means to verify my identity. Whilst today my application was approved, the fact they delayed it (it was submitted almost a month ago) just for minor water damage is a perfect encapsulation of how public bureaucracies tend to make our lives harder than necessary. In fact, I cannot think of a single instance where government (beyond the functions it would serve under minarchy) has made me better off, and a plethora of instances where the opposite was true. This is why I am a libertarian, hence why Hanson's proposal resonates with me.
Indeed, the raison d’être of the nascent field of market design is to expand the domain of markets to arenas where they have previously been unapplied. This achieves results [1]. The empirical and theoretical case for markets is as sound and as universal as ever, and we should not expect any different when it comes to criminal justice. Call me a “fundamentalist” all you like: I will wear that label with a badge of pride! There are few beliefs that I endorse unequivocally, yet the supremacy of free-market capitalism is one such view. Of course, if you do not subscribe to this perspective, Hanson also outlines possible vehicles for redistribution and state-driven interventions. Nonetheless, I know of no theory that makes the case for state monopoly on efficiency or redistributive grounds. You do not need a state monopoly on policing or punishment.
There are also two additional justifications for the privatisation of criminal justice:
Even if all prisons were privatised, under the current system the state would still act as a monopsonist. Allowing insurers to choose which prisons to send their convicts to would establish a free market in the penal system. In fact, to generalise this: a free market in punishment. Innovation in punishment. No longer would we face a binary choice between prison or an ineffectual probation service. The scope of possibilities would expand with our imaginations. Recidivism outcomes would likely improve.
Contract theory stipulates the conditions under which vertical integration (or in the sphere of public policy, nationalisation) would occur. Following a Coasian theory of the firm (where firms arise to eliminate transaction costs), the conclusions of Hart and Grossman (1986) imply that vertical integration is optimal when a firm seeks control of noncontractible assets. Yet policing, prosecution, and punishment are contractible. The only noncontractible assets under consideration are the laws themselves, which remain under the purview of the state.
If I could offer one line of criticism, perhaps this proposal is not radical enough. Why not go one step further, and embrace anarcho capitalism? One could imagine communities and organisations requiring insurance for membership. These groups then simply choose which acts they sanction (so require “vouching” to hedge against the risk of commuting those proscribed acts) and which they condone. A privatised court system would finance itself from fines and court fees. People ‘vote’ for policy bundles with their feet by selecting into a community, so all “microstates” must satisfy IR and IC constraints. Think of this as a Hoppean monarchy on steroids (without the monarchy).
There are a few complications with the implementation of anarcho capitalism in practice:
Transaction costs may become a salient feature of daily life. A proliferation of “microstates” may adopt wildly divergent laws, which may then hamper the flow of trade, capital, and people. Of course, these microstates could coordinate to form a confederation, but how would this be different to a nation state in practice? Theory of the firm seems to select against anarchy, when applied to the realm of governance. [2]
Anarcho-capitalists assume that the alternatives to government are benign. Empirical results reject this hypothesis. When inner-cities are subject to under policing, gangs emerge. As Levitt documented in Freakonomics, some gangs act as “mini” governments. They levy taxes (extortion, but functionally the same) on businesses in a territory, in return for protection from rival gangs. In addition, rentier states that profit from natural resources may also emerge. Look at how life was for the denizens of the East India Company! Of course, hypothetically, some jurisdictions may reintroduce slavery. People cannot vote with their feet when their liberties are curtailed, so the premise that all communities satisfy IR and IC constraints breaks down.
However, on point 2, the East India Company was a state-supported monopoly. We have gangs and rogue states such as North Korea today: it is far from obvious that these problems will worsen on the extensive margin under anarcho-capitalism. Indeed, under the correct incentives, we should expect improvements.
A realistic, incentive-compatible model of anarcho capitalism would likely revolve around our current startup cities such as Próspera in Honduras. For-profit governments’ incentives are obviously to maximise economic growth, hence most startup communities will conclude that it is in their interest to safeguard individual autonomy. Yes, rogue states will still exist, yet they will be a minority as they struggle to profit relative to other for-profit microstates.
This begs the question: how will a for-profit government generate its revenue? Off the top of my head, I can think of a few ideas. Of course, these proposals are not mutually exclusive, and any combination of these may be profit-maximising:
A vertically-integrated vouching firm that owns land, and requires insurance to live in its territory. This firm may also generate profit from land sales, so this hypothetical is appealing to Georgists.
Lump-sum fees as a condition of residence. As voting with one's feet induces successful for-profit states to satisfy IR and IC constraints, this will approximate an economist’s dream of Lindahl taxation. As multiple outcomes in governance are possible under for-profit experimentation, no longer will the standard bugbears of mechanism design in terms of preference aggregation (such as Arrow's impossibility theorem) be an issue. A Walrasian equilibrium then becomes possible for institutions too.
A microstate generates revenue via the provision of public goods such as roads. Yes, the difficulty of enforcing comprehensive tolling of an entire road network is obvious, so this non-excludability means that a microstate would likely not rely on this as its principal revenue stream (i.e. roads would be cross-subsidised under an integrated firm). Yet it resolves the issue of private provision of public goods.
Moreover, this model of anarcho-capitalism is in fact likely the most realistic vision of implementing libertarian ideals. In democracies, libertarians must either form coalitions with non-libertarians, or if they stand under their own parties, then they are reliant on people voting to strip away their handouts. With our current generation of startup communities, gradually (over a few centuries perhaps), I expect them to replace established nation states. These charter cities endure rapid economic growth, people migrate to reap the rewards, then nation states become depopulated whilst microstates thrive. Crucially, this mechanism is not dependent upon appealing to a tyranny of the majority in a democracy, and sustaining perpetual incumbency (as if libertarians get voted out, then the state would expand beyond minarchy).
Hence, I salute Hanson's revolutionary proposal, and my only gripe is that we could possibly push the frontiers of markets even further. Yet, if you accept the legitimacy of a minimalist government, then there is no reason why criminal justice cannot, nor should not, be privatised.
Note that this example also demonstrates why libertarianism, and not just the bleeding-heart variety, is fully consistent with compassion for the poor. The welfare state crowds out philanthropy. The former is a stagnant monopoly; the latter innovates, is subject to creative destruction, and is open to market mechanisms.
One could also stretch this framework further to advocate for a world government. If such a government was a nightwatchman, I would support the idea, but that would be unlikely. In any case, a world government is a nonstarter (although arguably the whole point of this post is to engage in political fantasy).