Book Discussion “Outsourcing War and Peace”: The Efficient Regulation of Private Security Contractors

[This is the third day in our discussion of Professor Dickinson’s book Outsourcing War and Peace: Preserving Public Values in a World of Privatized Foreign Affairs. Links to the related posts can be found below.]

The dean-emeritus of US government law of war lawyers, Hays Parks, spoke in passing a year ago about private security contractors and the law of war at a conference in honor of Michael Walzer.  Parks noted that for better or worse there was no going back to a world in which even the use of force, let alone other matters, was confined to formal government military forces – that bridge had been crossed with such things as State Department security and much else.  But he expressed concern with the way in which it had come about, with insufficient attention to the issues of accountability.  And also regret, from the standpoint of the laws of war, that certain of these functions had ever been allowed to be privatized.

Because, as Parks says, the bridge has been crossed, I’m fundamentally in agreement with the concerns that Laura Dickinson’s splendid book raises about accountability.  Point being that one can be a government laws of war lawyer and share these concerns, as Parks does; and equally share concerns for establishing mechanisms that lie within the realistic possibilities of legal policy for addressing the accountability concerns.  Contractual mechanisms are a very important category of that, and the questions raised below are less about principle than cautions about the limits of contractual mechanisms.  Like Chris and several other OJers, my teaching day-job is in business, finance, contracts, and law and economics, and I also draw here on experience as the board chair and general counsel for several NGOs engaged in much cross border development finance and related contracting – microcredit and all that.  So let me run through a standard set of law and economics issues that might arise in these attempts to regulate via contract:

1. Might contract mechanisms under-deter or over-deter the use of private contractors for particular functions (or under some circumstances, might they do both)?  One of the things I most like about Dickinson’s nuanced approach to the possibilities and limits of contracting mechanisms is the awareness that contract enforcement in our civil court system is not well suited to the kinds of situations that would most concern us in the situation of private contractors, for example, using force in some foreign situation of high insecurity if not straight up conflict.  This is so, I think, for contract claims as well as tort claims.  For example, the possibility of wrongful death of a foreign national in some not-exactly-quiet-not-quite-war-zone by a security contractor is not very well suited to the social setting in which the US civil litigation system of wrongful death has meaning – which is a settled domestic legal system in which the participants all share certain expectations about what is okay and not okay in the circumstances.  Moreover, much of that settled, legitimate, domestic system has to do with situations of risk in which we as a society accept that there are crucial social benefits that lead us to accept risks to life and limb, even ones that implicate wrongfulness.

I am skeptical that those shared social assumptions, which deeply inform and legitimate our domestic civil legal system, can be applied well in all these other situations.  I doubt that the adjudicators think they can do so, either, even if tasked to do so.  The tendency is either to hold contractors to a stricter standard than one might think is efficient to the ends – ends, however, which in quasi-war zones are themselves contested – for which some modicum of force is justified.  In that case, contractors will be over-deterred once they understand the costs they must internalize.  Or else the adjudicatory system, concerned that it is being asked to apply purely domestic standards to highly contingent and fundamentally different conditions of conflict, will excuse too much, whether overtly or indirectly, for fear of penalizing unfairly something where command and control, and ultimately accountability, ought to have rested with the government agency, not the contractor.

These are not mutually exclusive possibilities, and the system might both over-deter and under-deter, in different ways and circumstances.  Firms might opt out of contracts that they understood to force them to internalize costs that they could not properly police, but which represented tasks that ought, all things considered, to be done even with those risks to innocent persons.  And other firms might take on those tasks, screw up, but then not be penalized because the domestic adjudicatory system thought it would be unfair to hit them with responsibility.

Dickinson is admirably clear about these possibilities of the limits of contracting as a regulatory mechanism, and I think she is right to see contract as being far more suited to insist on terms that deal less with liability as such – negligence or strict liability – and instead with ways in which governments can require training, vetting of personnel, and so on.  These are requirements that are far more likely to mean something as than the effort to write final conduct standards into contract covenants.  But the limits of what can reasonably be expected out of contract terms leads to a different set of concerns.

2. Agent and principal.  The broader concern that private security contractors are not government actors under the direct command and control of government authority is that contract – privately contracting security personnel – is not finally an effective mechanism of control.  (This often seems odd to contemporary law and economics scholarship, but that’s because the field has difficulty conceiving of any relationship as other than contractual, in substance if not law, from family law to national security; every institution is just a nexus of contracts.  But that is a weakness of law and economics as a field – but it has special resonance here, because war is so often not a matter of rational actors pursuing their economic self-interest.)

The proper relationship here is agent and principal, and a very special one at that – government agent. We are trying, in effect, to recreate what are properly agent-principal relations through contract mechanisms.  In the security area, this is especially difficult for two reasons.

One is that contract is a very problematic mechanism for controlling the behavior of an agent in which there is substantial or even radical uncertainty as to the situations that are likely to be faced. The problem is not the traditional one of agent-principal asymmetry of information, but rather symmetry of uncertainty: neither the private security contractor nor the government actor truly knows what kinds of situations will arise or the best response.  Since this often describes armed conflict, post conflict, or insecure situations in which private security contractors often operate, the ordinary response in the law to such symmetric uncertainty is not to rely on contract at all – or, at most, to use the form of a contract to memorialize what is essentially the discretionary, standards-based, judgement-based relationship of an agent acting for a principal.

Contract in its purest form seeks to fully specify all relevant contingencies and assign a covenant prescribing behavior and a breach term for each.  But that is not going to be effective or even possible in the difficult and uncertain conditions of conflict and insecurity.  Contract here is really just a form for acknowledging reliance upon concepts of faithful agency.  I was once asked to draft a contract for a private corporate telecommunications company, with supposedly great experience in difficult environments, to set up a television network in Bosnia following the Dayton Accords in preparation for elections. Since neither party had any real idea what to expect on the ground – post conflict situation in the Bosnian hinterlands – the contract was mostly littered with terms like “reasonable,” “best efforts,” “good faith,” and so son – and in the event proved to be not just useless but worse: we had papered over our mutual uncertainties with terms like “reasonable” that might mean something in commercial circumstances in London or New York, but could provide no concrete measure of behavior for the parties, or for a judge afterwards, and allowed everyone to kick the can of entirely different expectations down the road.

A retired judge in London with whom I consulted on possible litigation shook his head and said, this is not in substance a contract, it is in bare legal form a contract committing everything over to the undefined reasonable judgment of an agent in the field, in matters having nothing to do with the ordinary commerce that defines “reasonableness.” The telecom team in the field thought it was patently unreasonable to expect them to work where there might possibly be mines or some amount of gunfire; I thought it was patently obvious that these had always been understood as the conditions under which the company had offered itself.  Reasonable?   I raise this example because it seems to me a small example of what contract cannot do in these situations – and the mismatch of expectations that it invites.

That said, however, Dickinson’s detailed research points to the fact that a lot has been learned about best practices for the conduct of private security forces in various circumstances.  It’s not purely a matter of reliance on the judgment and discretion of the agents; there’s a lot that can be described more exactly.  Not necessarily as a matter of trying to draft specific covenants about what agents must do in circumstances x or y, but instead as a matter of specifying the upfront matters of training, certification, best practices, vetting, etc.  That can go a long ways.  It is also true that a lot of this knowledge is local and context specific – what works in Iraq today might have little to do with teams seeking out Kony or confronting pirates in Somalia.  But there are important things that can figure in contract terms.

The other agent-principal issue is the more traditional one in the law and economics literature – the misalignment of agent and principal interests, and that in the context of the agent having greater information and control of the situation on the ground.  I don’t think this is the main problem in the private security contractor context – in general, I read Dickinson’s work to suggest that, whether they do well or badly, in general the private security agents prefer to remain as closely tied to, and by implication legally protected by, the government umbrella as they can.  This is likely not true in all cases, but I think the much greater problem of is the first one I started with – what is the best legal regime for addressing radical uncertainty on the part of both parties, with a recognition that contract is not by its nature well suited.

3.  Governments wanting it both ways.  Overall, my skepticism about private military contractors in such fraught matters as using force is that governments really want it both ways.  They want the cheapness that outsourcing through the contractual form can bring – essentially, lowering costs on the front end.  They tell themselves something that is, indeed, oftentimes true in the domestic regulation of private actors in a domestic economic setting – that it is better to outsource in part because the regulator is separated from the regulated.  True. But military and quasi military security situations are different, and traditionally the mechanisms of contract have not been seen as sufficiently robust to police the behavior of agents.  That which is saved on the front end comes back to bite, with interest, on the back end.

Governments recognize that – and yet wind up drafting contracts that risk the worst of both worlds – the appearance of contractual certainty, but in substance reliance upon terms that convey an agent-principal relationship, but in a murky and ill-defined way.  Dickinson clearly understands this, and one salutary effect of her book is to point to ways in which it is possible for contractual form to embrace much greater substance – without turning these very peculiar agency contracts into a mechanism of disguised tort or even criminal liability.