After 70 years, it's time for some major restructuring at the CIA — at least according to its director, John Brennan, who announced earlier this month his plan to overhaul the agency by creating "mission centers" that would concentrate resources on specific challenges or geographic areas. Brennan also announced the formation of a new "Directorate of Digital Innovation" to lead efforts to track and implement new intelligence-gathering cyber tools.

While "cyber" and "innovation" are certainly newsier buzzwords than "restructuring," Brennan's planned changes to the CIA's internal architecture are generating controversy and conversation, and rightly so. Brennan's reform would extend the model of the Counterterrorism Center (CTC), a division already inside the CIA to the entire agency. This means analysts and operations officers will no longer work separately, but side by side in regional and thematic units. This is a major shift which overcomes a decades-old division of labor. The CTC model is not without its advantages. Analysts with deep knowledge of their enemies would be able to target operations more efficiently, and they could more easily assess the reliability of incoming information. Intelligence would circulate more freely among those working on the same topic and officers would be able to apply their talents and expertise to a wider range of tasks.

But breaking down the wall between analysts and operations officers has some even bigger drawbacks. With the CIA conducting more and more paramilitary operations, its budget has come to depend on the perceived effectiveness of these operations. With analysts and operations officers working side by side, it will become harder to find analysts willing to say that some operations are not working. Brennan dismissed this concern saying that the history of the counterterrorism center demonstrates that its analysis has remained objective.

Unfortunately, that history says otherwise. The CIA described two of the CTC's biggest programs, drone strikes and interrogations, in terms that were simply too good to be true, misinforming policy-makers and the public.

In 2011 then senior counterterrorism adviser John Brennan stated that drone strikes had not killed civilians in over a year. The data very likely came from the CTC, the unit of the CIA that conducts drone strikes and analyzes their consequences. Brennan later had to retract his statements to the press, saying that the CIA had not found credible evidence of civilian casualties, considering every male of fighting age killed in the strikes as a combatant. Had there been a cadre of analysts working separately from the operations officers, it is far more likely that Brennan's claim would not have made it out the door.

The interrogation program offers even stronger evidence of analytic biases. For instance, between 2002 and 2007 the CIA repeatedly claimed that al-Qaeda member Abu Zubaydah provided information on terrorist suspect Jose Padilla after being subjected to the CIA's enhanced interrogation techniques. In fact, he had provided this information while being interrogated by the FBI using rapport-building techniques, before being handed to the CIA. When questioned by the Senate, the CIA drafted an answer saying that the legal office of CTC "simply inadvertently reported this wrong", but then refused to send it. Similarly, the CIA admitted that it had "mischaracterized the impact of the reporting [the CIA] acquired from detainees on the Karachi plots" against the U.S. Consulate. Instead, the plots had been thwarted by Pakistan's arrest of the operatives.

The Senate's report on the CIA's detention and interrogation program showed that CTC officers understood that their budget and reputation depended on the interrogation program's perceived effectiveness. In 2005 the CTC's Deputy Director told his colleagues that "we either get out and sell [the interrogation program], or we get hammered, which has implications beyond the media. Congress reads it, cuts our authorities, messes up our budget." In their attempt to "sell" the program, CTC officers gathered information in a way that was hardly conducive to dispassionate analysis. In 2004 the CTC Legal Office asked CTC officers for "a list of specific plots that have been thwarted by the use of detainee reporting that we acquired following the use of enhanced techniques." That office planned to write a report "emphasizing that thousands of innocent lives have been saved as a result of our use of those techniques" and warned that "the future of the program, and the consequent saving of innocent lives, may depend substantially upon the input you provide."

Warned that they would have had innocent blood on their hands had they not contributed to selling the program it is not surprising that CTC analysts succumbed to the pressure. Had there been a separation between analysts and operations officers, it would have been easier for analysts to assess whether the interrogation program really was saving thousands of innocent lives, as CTC claimed.

Ironically, the CIA more than any other agency should be aware of the perils of compromising its objectivity. Initially a weak organization competing with larger bureaucracies like the FBI and military intelligence, the CIA rose to prominence by investing in what the founding father of the CIA's analytical branch Sherman Kent called in a collection of essays published by the CIA in 1994, "disinterested objectivity." Whereas other intelligence agencies tailored their analyses to fit their bureaucratic needs, Kent argued that to be truly "Central" the CIA had to build a reputation for speaking truth to power.

In order to speak with disinterested objectivity, Kent believed that the Agency had to reduce covert operations, especially paramilitary ones, to a minimum. The more the CIA engaged in covert operations, the more it developed an interest in their continuation, creating an incentive to exaggerate their effectiveness. Second, analysts and operations officers had to work separately, giving analysts the space they needed to look at the larger picture without any pressure to "sell" the Agency's operations.

If Brennan's suggested reform comes to pass, any pretense of disinterested objectivity in the CIA would disappear. The CIA would become a largely paramilitary agency with no separation between analysts and operators, the opposite of what Kent had envisioned.

The likely consequence would be a multiplication of the biased assessments like the ones on the drone and interrogation programs. In the long run, the CIA would lose its reputation for objectivity, its ready access to senior policy-makers and its centrality in the policy process.

Sign up to get The Weekly Wonk, New America's digital magazine, delivered to your inbox each Thursday here.