The answer is not “the rule of law.” According to the WSJ, Holder said:
The 9/11 attacks were both an act of war and a violation of our federal criminal law, and they could have been prosecuted in either federal courts or military commissions.
So the U.S. government has the option to try suspected members of Al Qaeda in civilian court or in military court. The “rule of law,” then, does not compel traditional civilian-court protections. However, the question remains unanswered. To say that one has an option is not to say why one exercised that option as one did.
Then what is the answer? It is surely this: the Obama administration has decided to offer a two-tiered system of justice. We might call them the “high-quality” (civilian) tier and “low-quality” (military) tier. The high-quality approach offers greater accuracy; the low-quality approach offers less accuracy. The Obama administration will use the high-quality system against people when it has a strong case, and the low-quality system against people when it has a weak case.
This approach makes sense. Endless detention without trial is no longer a politically viable option. The government will make a judgment as to whether a suspect is dangerous or not. If the case is good, the high-quality system will be used. If the case is bad, the low-quality system will be used. In this way, the government can ensure that people it thinks are dangerous will be locked up.
This system is superior to the two possible one-tier systems. A pure low-quality system (military commissions only) suffers from credibility problems. People will not believe that all the people who are convicted are guilty. A pure high-quality system (civilian courts only) would result in too many acquittals. People who the government believes are dangerous will be back on the streets. The two-tiered system allows for credible convictions when credible convictions are possible, and (non-credible) convictions when credible convictions are not possible. The two-tiered system produces higher overall credibility without sacrificing the incapacitation of dangerous (or supposedly dangerous) people.
The main criticisms of Holder’s approach are that KSM and others will take over proceedings and use them for propaganda purposes, that secrecy will be compromised, and that the approach signals insufficient seriousness about the terrorist threat. The first two concerns are actually irrelevant. The DOJ will decide on a case by case basis, and if those concerns in any particular case are serious, it will opt for military commissions. The last concern is harder to evaluate, but it boils down to the claim that a blunderbuss system that results in outcomes that people distrust is better, on symbolic grounds, than a surgical system that produces the same pattern of convictions but with higher overall credibility. Why would the more intelligent approach signal lack of seriousness about terrorism?