PublicOrgTheory

Though over a week old, this story nests somewhat well with yesterday’s thoughts on the A.Q. Khan network’s sustainability and legacy:

LONDON, England (CNN) — A second set of confidential papers on British terrorism policies were found on a train, The Independent reported on Sunday.

A member of the public found a set of papers on British terrorism policies at the Waterloo station in London.

A member of the public found the documents at the Waterloo station in central London and turned them over to the London-based newspaper, Simon Evans, a journalist at The Independent, told CNN.

The paper gave them to the Treasury Cabinet on Friday evening, Evans said.

The documents outline how trade and banking systems “can be manipulated to finance illicit weapons of mass destruction in Iran,” The Independent reported.

The article goes on to note that each page of the documents bears the legend “Australian/Canadian/UK/US Eyes Only”, a nugget that would seem likely to arouse interest if found on, say, a train. After all the complex procedures put in place to keep these documents under wraps, it would seem this internal control has the unintended consequence of being far more intriguing to the uninitiated than an innocuous-seeming interoffice memo.

However, about those procedures: these incidents are somewhat reminiscent of Scott Sagan’s eloquent discussion of normal accidents and high reliability organizations, The Limits of Safety. The quick summary for the uninitiated (and this really is a book worth picking up) is that Sagan was intrigued by the question of why we haven’t had a serious nuclear accident. Is it because the organizations running these programs have massive reliability built in, or is it that the accident one might expect just hasn’t happened yet? Recent incidents involving accidental transfers of nuclear weapons within the US and export of weapons parts outside it certainly make a case for the latter.

Unfortunately, there is no lack of accidents in the intelligence field, and there never has been. The salient question, then, is what form of system can help compensate for human fallibility? The high reliability organization tends to offer some degree of false comfort through tight coupling and redundancy, but undermines itself by fomenting complexity beyond the capacity of organizational actors to comprehend. Workarounds follow, and mayhem ensues.

I don’t imagine complex systems and the craft of intelligence are all that compatible. Sure, there have to be some safeguards in place to ensure that hard-won information is both secret and useful. Still, doesn’t it make sense that

1. Rigid, cumbersome systems would spawn workarounds;
2. Detailed regimes of rules will be impossible for all people to remember at all times;
3. The importance of intelligence will diminish with volume; and
4. Those who aren’t supposed to see this stuff might find it more intriguing if labeled, “Look! Spy Papers”?

More to the point, how likely do we think it is that the system undermines itself? I would say pretty likely. Again, there must be safeguards in place, and classified information must be handled differently. Still, it’s much easier to say that some wanker left some documents on a train than to try to understand why that would happen–twice. It is the rules themselves, not the actors, that provide the most levers for change. An investigation to assign blame might feel good, but the cure is almost certainly in the hard work of looking at the organizational system.

Study question:  what advantages and disadvantages does a network structure have over a bureaucratic structure for trading in secrets?