CIA Caught Between Operational Security and Analytical Quality In 1953 Iran Coup Planning

By Danielle Siegel

The National Security Archive posted the most recently-reviewed version of the partially-declassified internal CIA history of the 1953 coup in Iran, “Zendebad, Shah!” this February. This version, written by CIA historian Scott A. Koch and released by the agency in late 2017, includes a number of previously-secret passages about the planning and execution of the coup—known internally as Operation TPAJAX. Arguably its most notable inclusion is an appendix titled “CIA and Operation TPAJAX: The Tension Between Analysis and Operations.”

The appendix paints a picture of a divided CIA in which analysts were denied information about the operational details on the ground in Iran, a dynamic that was encouraged by high-level agency officials and impacted the accuracy of intelligence briefings received by the Eisenhower administration. However, there are a few more interesting points that Koch emphasizes in this appendix that warrant further exploration—most notably what TPAJAX illustrates about the tensions that exist between operational security and analytical quality in the planning and execution of covert operations, the factors driving that dynamic, and what lessons he thinks ought to be learned from the operational-analytical divide.

Koch notes that in any covert operation, “Preparation must balance the need for fully informed decisionmaking with the need for strict operational security,” ultimately concluding that “TPAJAX was planned and executed with far greater concern for operational security than for ensuring that the planners had all relevant information” [113]. This “philosophical tension,” to use Koch’s words, between operational security and informed analysis manifested itself in a systemic lack of communication between operatives and analysts throughout the planning and execution of TPAJAX. He notes that “There is no evidence that the [   ] in Kermit Roosevelt’s NEA Division consulted the Office of National Estimates (ONE) or the analysts in CIA’s Office of Current Intelligence (OCI) at any stage of the operation,” and adds that the Office of Strategic Operations often declined requests from ONE for information to guide its long-term strategic estimates—estimates that were used, in the early stages of the operation, to brief President Eisenhower [113, 114, 116].

Furthermore, the appendix highlights how traditional methods of intelligence analysis were subverted during TPAJAX.  Usually, unless there is a time-sensitive crisis and the executive needs immediate information, analysts put raw intelligence in context and consider the long-term strategic implications that ought to inform policies and decisions. Koch points out how, in the months leading up to the ousting of Mosaddeq, “rather than going to analysts, the ‘best clandestine reports were being hand-carried by top clandestine services people over to senior people in the White House, the State Department, and the Pentagon,’” while analysts saw “’mostly inconsequential scraps of information about foreign personalities’” [117].  Thus, filtering intelligence through the analytical wing of the agency was the exception rather than the norm during the planning and execution of TPAJAX.

John H. Waller, one of the senior Directorate of Operations officers involved in TPAJAX, whose interviews with the author of Zendebad, Shah! provide the basis for much of the analysis about the tensions between the operational and analytical wings of the agency.

One explanation for these tensions, Koch contends, lays “in differences between overt and covert employees” [117]. There was a sense among operatives (the covert employees) that analysts (the overt employees) may be more likely to leak classified information. Yet beyond concerns about the security of classified information, Koch discusses an intellectual and bureaucratic divide that he argues perpetuated the tendency to exclude analysts from the planning process of TPAJAX. He cites a 1995 interview with John Waller, one of the senior Directorate of Operations officers involved with the coup, in which Waller explains the divergent approaches to intelligence during TPAJAX: “…most Iranian specialists in the DDP were OSS veterans who spent substantial amounts of time in the Middle East. They had acquired their knowledge from practical experience and thought that knowledge acquired this way was superior to the academic knowledge many Directorate of Intelligence (DI) analysts prized” [117-118].

Koch also suggests a divide existed at an institutional level, writing that “DDP officers may have thought that if the DI were included in covert action planning, analysts would begin to challenge DDP’s preeminence in covert operations. Similarly, analysts may have feared that DPP operators would question their analytical preeminence and that close association with a covert action would raise questions about their intellectual objectivity” [118]. Thus, there seems to have been a set of bureaucratic norms about the roles of different actors within the agency that reinforced a culture of distrust between the analytical and operational arms and discouraged internal cooperation.

Despite these deficiencies, Koch argues that analytical exclusion had “negligible” consequences for the outcome of TPAJAX [120]. He points out that many analysts concluded that a coup attempt would fail, but that “…did not dissuade the President, the Secretary of State, and the DCI from executing TPAJAX” [120].  The administration, in Koch’s assessment, was committed to ousting Mosaddeq whether analysts thought it wise or not. If anything, he suggests that analytical exclusion harmed the intelligence product “because it prevented analysts from basing their judgments on complete information” [120]. Koch more specifically argues that if analysts were informed about the extent of the U.S. role in the developing political situation in Iran, they would have been “more inclined to recognize the operation’s potential for success” [120]. For example, “Mossadeq arrested Col. Nassiri, and the military challenge melted away. Headquarters wanted to call off the operation. Had the planning taken into account the possibility—even the likelihood—that segments of the Iranian military would react this way, DDP could have had contingency plans in effect instead of relying on Roosevelt’s improvisation” [120].

Shah Mohammed Reza Pahlavi of Iran and President Dwight D. Eisenhower – all smiles in the early years following the ouster of Prime Minister Mohammad Mosaddeq. (RFE web site)

In other words, coordination between analysts and operatives could have made the success of TPAJAX less random, inadvertent, and contingent upon the luck and circumstances of individual agents. This assessment is particularly notable in light of the events that took place in Guatemala just one year later.  In CIA historian Nick Cullather’s internal history of that operation, called PBSUCCESS, he concludes that the ousting of the Guatemalan leader Jacobo Arbenz was largely a consequence of good timing and good luck as opposed to the actual operational tactics and strategies employed by the CIA. It is interesting to think about how improving interagency coordination could have enhanced the execution of both TPAJAX and PBSUCCESS. In the case of Iran, Koch suggests that “The DI’s more scholarly and detached perspective and its methodology for assessing a dynamic situation perhaps could have helped NEA clarify the assumptions upon which TPAJAX was based, and how those assumptions might affect the operation” [120]. Perhaps normalizing communication and increasing coordination between operatives and analysts could have made the operation in Guatemala more direct and purposeful, as well. A major lesson Cullather imparts in his recounting of PBSUCCESS is that assumptions drive the tactical choices agents deem available to them, so if analysts had the most accurate intelligence, they may have come to more sound conclusions and consequently made operational choices that had a more direct impact on the desired outcome.

Koch concludes the appendix with a warning about the increasingly dangerous consequences of subverting the analytical process in intelligence.  He writes that “advances in technology have given today’s analyst access to an almost bewildering array of sources inconceivable to his colleague of 44 years ago,” and that the most up-to-date technologies are usually available to analysts but not operational planners [120].  In other words, planners rarely see all of the intelligence that is collected and must rely on analysts to filter through and contextualize information due to the sheer volume of intelligence made available through technology. Ultimately, Koch argues, these changes make “the consequences of ignoring analysis more serious today than was the case in 1953” [120].