Pakistan: Is the War Contractible?
Writing about web page http://www.ft.com/cms/s/0/3574b5ee-ab15-11e0-b4d8-00144feabdc0.html
Relations between the United States and Pakistan have been going downhill since the killing of Osama bin Laden. Ex ante, it seemed likely that the Pakistan Army had offered some degree of shelter to the world's most wanted terrorist. Ex post, Pakistan has halted U.S. access to a drone base on the Afghan border; cut back on visas for U.S. training missions; and detained a doctor said to have helped U.S. intelligence identify bin Laden's family. In retaliation, the United States is now threatening to withhold $800 million in military aid.
The Pakistan response to this threat throws further light on the status of Pakistan as a U.S. ally in the war against Al-Qaeda and the Taliban. On July 10, 2011, the Financial Times reported:
Lieutenant General (retired) Moinuddin Haider, a former Pakistani interior minister, said that the halt on US aid would further strain the two countries’ relationship and called on the US to reconsider. “This move will only add to the anti-Americanism in our country,” he said.
The following day, Reuters reported:
Politically, [the suspension of aid] would be damaging to the relationship, said Pakistan's former ambassador to the United States, retired Major-General Mehmood Durrani said, reflecting a widespread view in Pakistan that it was fighting America's war, for which Washington must reimburse it.
So: Is Pakistan an ally or an enemy? Neither, it seems. There is anti-Americanism, but Pakistan is not an enemy. For Pakistan does make available services and facilities to combat AQ and the Taliban. But Pakistan is not an ally, either.
Nearly 50 years ago, Mancur Olson and Richard Zeckhauser worked out the basic theory of burden sharing in alliances. Consider two countries -- call them A and P -- that face a common enemy. A and P can fight together or separately. By fighting together, they can share the synergies from each other's efforts. For this reason, an alliance is vulnerable to free-riding: the more A contributes, the less P needs to put in. This turned out to explain quite well the pattern of burden sharing in NATO under the doctrine of massive retaliation.
The same theory leads to a straightforward prediction about what will happen if A suspends assistance to P for some reason. On its own, denied the help of A, P will put more resources into the struggle.
But what Pakistanis are saying does not fit this model. What they are saying is that, denied the help of America, Pakistan will fight less, not more. It is "America's war."
If neither an ally nor an enemy, what is Pakistan? Pakistan is a contractor. There is a contract between America and Pakistan, so that America mostly pays and Pakistan mostly fights or helps America fight. Pakistan will help resist AQ and the Taliban if America pays, and not otherwise. It is the same as a contract between me and a builder: I pay, so the builder builds, and not otherwise. There is no intrinsic common interest that we share. Incentives are aligned by agreed payments, not by anything else.
The work that my builder does, however, is contractible. That means we can write a contract that spells out the obligations of both sides with reasonable completeness, and includes most reasonably foreseeable contingencies. He works to my wife's satisfaction, I pay, and we part amicably. Otherwise, we may end up in court.
The war on AQ and the Taliban may not be contractible in the same sense. There are three reasons.
- First, there is asymmetric information. It's hard to tell whether the Pakistan Army is doing a good job, because it is hard for the Americans to monitor progress in harsh terrain among often hostile communities. As a result, America cannot know whether Pakistan is fulfilling its part of the bargain.
- Second, there is moral hazard. The Pakistan Army can explain almost any setback by "We didn't know" (that bin Laden was living under our noses) or "They got away." As a result, Pakistan has a strong incentive not to try very hard.
- Third, the Pakistan Army has residual ownership rights in the literal sense that it owns much of the Pakistani economy and is much more of a sovereign government than the nominal civilian government in Islamabad. As a result, when contingencies arise that are not in the contract, it can control its response in ways that do not formally breach its understanding with the United States over Afghanistan, yet are not ones that the United States would choose. For example, it can secure nominal improvement on the North West frontier by diverting Islamists into Kashmir or on to Mumbai.
A standard solution when a venture is non-contractible is vertical integration. In this case it would require the United States to take on the responsibilities of a colonial power. Ruling this option out on grounds that probably don't need spelling out, we are left with an interesting problem and no obvious solution.