Re-arranging bureaucracies has long been a favorite Washington way of pretending to make improvements. It is a handy recourse in the absence of good ideas to make real improvement. Revising a wiring diagram is the sort of change that can be made visible to the outside world. It does not require reaching consensus about significant increases or decreases in the priority given particular programs or their budgets. It offers a basis for convincing ourselves that the bureaucracies involved will perform better, even if the main reasons we don't get everything we would like to get from those bureaucracies are to be found in the inherent, unavoidable challenges of the tasks they are assigned to perform.

The urge to reorganize is not limited to government. Revising wiring diagrams is alluring to senior managers in private sector organizations as well. It is a way of showing initiative and appearing to be dedicated both to improving the organization and keeping pace with changes in the outside world. It is one of the most visible ways for any senior manager to leave a mark and establish a legacy.

Now the Central Intelligence Agency is being hit again with the reorganization bug, with changes that director John Brennan announced last week. The intelligence community has been subjected to this sort of thing at least as much as other parts of the federal bureaucracy. The most notable instance was a reorganization of the community a decade ago as the most visible part of the 9/11 Commission's response to a popular demand to shake things up after a terrible terrorist attack. That change added new bureaucracy on top of continuing old organizations, and in the years since has given us little or no reason to believe that it was a net improvement.

The principal feature of the changes that Brennan announced is to move all of the agency's operational and analytical work, and not just selected parts of it, into integrated “mission centers” covering issue areas defined either geographically or functionally. As with most other reorganizations, both criticism and praise tend to be overstated. Any change in a bureaucracy's performance, for good or for ill, resulting from changing the wiring diagram will not be nearly as pronounced as either critics or promoters usually would lead us to believe.

A criticism of this newest reorganization, for example, is that it would lead to still more focus on current doings at the expense of longer-range analysis. But within each issue area there is no reason to believe that worthwhile long-range analysis cannot be done in the mission centers. Another line of criticism involves a feared compromise of the integrity of analysis because of overly close association of the analysts with operators. This would only be a problem, however, where covert action is involved. Although some unfortunate experiences involving Central America in the 1980s demonstrate the corrupting potential, covert action—despite the public image of what the CIA does—constitutes a small (and usually well-compartmented) portion of the agency's work. There is a substantial hazard of policy preferences influencing analysis stemming from relations with policy-makers, but that is a separate matter from relations between analysts and operators within an intelligence agency.

The justification for the changes is also overstated—or fuzzy and hardly compelling. Mark Mazzetti's article in the New York Times about the announced changes mentions that Brennan relied heavily on “management jargon” to try to explain and justify what he was doing. There were all the unsurprising buzzwords about needing to “wring efficiencies” out of the system and having to modernize and about not wanting to become as obsolete as Kodak, but how this makes one particular wiring diagram better than another one is difficult to see. Brennan talked about the “array of very challenging, complex and serious threats to our national security”—the sort of language that any CIA director, at any time, uses—but what does that say about the supposed advantages of a particular organizational scheme? He said a central aim was to eliminate “seams” in coverage, but aren't there seams in any organizational arrangement, including the seams that will exist between the mission centers?

The particular organizational issues involving the CIA entail, as many such issues in other organizations do, inherent trade-offs, with each possible wiring diagram presenting both advantages and disadvantages when compared to other possible schemes. The main advantage of the announced new arrangement is to make the interface between analysts and collectors working on the same substantive issue as close and smooth as possible. This helps the analysts to understand better the sources of some of the information on which they are relying, and it helps the collectors to understand how the information they are collecting is being used and where are the most important information gaps that still need to be filled.

A significant disadvantage is that bureaucratizing whatever is considered at the moment to be worthy of its own mission center makes for a less flexible and less nimble organization as issues change and especially as new (and sometimes difficult to recognize initially as important) issues emerge. The seeds of future intelligence failures can be found in the seams between the centers. Interface is important not just between collectors and analysts but also between analysts working issues that are different but may turn out to be related in important ways.

Another set of disadvantages stems from breaking up what would otherwise have been critical masses of people working in the same discipline and with the same skill set. Doing so is generally not conducive to enhancing specialized skills, whether those skills involve the craft of espionage, or of analysis, or something else. Particular mission centers, depending on who leads them and what are the relative weights of different types of people assigned to them, may tend to be co-opted by certain disciplines at the expense of the necessary professional care and feeding of those in other disciplines. The further separation of missions and operational control from the management of employees' careers (and the new scheme will retain existing directorates, including those for operations and for analysis, for that latter purpose) will tend to exacerbate issues of personnel management, including loosening the tie between effective contribution to a specific assigned mission and reward in the form of promotions. Retention of the existing directorate structure in addition to more mission centers also makes the whole organizational structure of the agency more complicated.

A principle too rarely recognized is that the advantages of a new organizational structure are uncertain (when compared to the existing structure, which is apt to have to have evolved over time as experience has shown what works and what doesn't), but the costs and disruptions associated with any major reorganization are certain and substantial. The disruption involves everything from having to forge new relationships with bosses, co-workers, and customers, to having to figure out exactly where new lines of responsibility are to be drawn. Rather than impeding accomplishment of the mission with such disruption, it often is better just to let people get on with their jobs—although anyone who makes this observation risks being rebuked as a stuck-in-the-mud resistor of change.

In the face of the inevitable trade-offs, the current organizational arrangement in the CIA, in which there are some integrated centers for selected issues such as terrorism but not for everything, is probably a reasonable compromise. Unmentioned in much of the commentary so far on the announced changes is how much had already been done, outside the centers, to enhance communication and cooperation between collectors and analysts. This includes physical changes made years ago to locate in adjacent office space the analysts and operations officers working on the same geographic areas.

What we most need to be wary of with these latest announced changes in the CIA's organization is not some wave of corrupting influences that will destroy the integrity of analysis. We should instead ask whether this is yet another of the many examples of a senior manager using reorganization to try to make his mark and leave a legacy, especially a legacy that won't be centered on unflattering matters such as strained relations with Congressional oversight committees.