Assessing Threats: Where You Stand Depends on Where You Sit—Remarks By Greg Thielmann to DIAA/DCOR

Assessing Threats: Where You Stand Depends on Where You Sit

By Greg Thielmann, Senior Fellow, Arms Control Association
Presentation to the Defense Intelligence Alumni Association and
the Diplomats and Consular Officers Retired

Washington, DC
July 27, 2016

In devising a title for my presentation today, I borrowed a phrase popularized by Graham Allison in his 1971 book on the Cuban Missile Crisis, Essence of Decision.  I suspect that every public policy student in the 1970s is familiar with it: “Where you stand depends on where you sit.”

As everyone in this audience appreciates, the “rational actor model” does not explain everything that happens inside government or between nation-states.  So even though the discipline of intelligence analysis is built on an ethos of objectivity, I don’t think I’m going out on a limb to argue that assessing threats is no exception to Allison’s aphorism.  I’d like to reflect on some threat assessments I’ve witnessed during my career in the executive and legislative branches of government. 

Backing into Threat Assessment

I backed into threat assessment in my first full-time job with the federal government.  Having just received my Masters Degree in Public Policy, I was hired as a budget examiner in the National Security Division of the Office of Management and Budget (OMB).  My task was to monitor a $4 billion line item, Navy Research and Development, and look for ways to save money on the function.  (Where I sat certainly influenced where I stood on assessing the threat!)

One of the things that I learned very quickly was that nearly all U.S. Navy R&D spending was then being justified by the need to stay ahead of Soviet military forces.  The Soviets were assessed to pose the only serious challenge to the U.S. Navy for control of the seas and for forcing entry on land.  I was told that Soviet warships were newer than America’s; they were armed with faster torpedoes and more capable surface-to-surface missiles; Soviet submarines were more numerous than their American counterparts; they could dive deeper and travel faster, and were double-hulled and hence less vulnerable to attack. Soviet anti-aircraft defenses were described as the most formidable in the world, and Soviet Backfire medium-range bombers could actually fly all the way to the continental U.S.

Exaggerating the Threat

As I gradually learned about the numerous advantages the U.S. and its allies enjoyed in the maritime realm over the Warsaw Pact, I became suspicious of the way the way the Soviet threat was being portrayed, and began to realize two things about DoD characterizations of foreign threats.

  • First, senior DoD officials and military commanders lean toward prudent worst-case analyses, and justifiably so because underestimating an enemy can be disastrous if deterrence fails and war occurs.
  • Second, OSD and the U.S. military have a vested, bureaucratic interest in exaggerating foreign capabilities to Congress, because doing so tends to increase appropriations for defense.

It was, therefore, unsurprising that the Defense Intelligence Agency (DIA) and the service intelligence agencies would paint the threat in more dire terms than the Central Intelligence Agency (CIA) or the State Department’s intelligence bureau, INR.  The latter two entities, I concluded, were consistently more objective in characterizing the threat.

After two years in OMB, I began a 25-year career in the Foreign Service, often working in political-military jobs at home and abroad.  In two of my last three tours working for the State Department, I served in INR, where assessing nuclear threats was my principal focus.

My exposure to the U.S. intelligence community, both at State and later on the Senate Intelligence Committee, contributed to my current view that national threat assessments, though dominated by the supposedly unbiased criteria of the CIA, were also skewed toward “worst case” outcomes rather than those judged “most likely.”

One of my “take-aways” from witnessing the threat estimation process up-close over the years was that even the CIA had a vested interest in exaggerating threats:

  • After all, the more dangerous the world, the more necessary it was to have a large and powerful intelligence establishment; and
  • An organization that warns against all manner of calamities is more likely to be able to claim “I told you so” if something bad happens.

Worst-Case Thinking

I think we can all agree that nasty surprises send huge shock waves throughout the national security establishment.  The mother of all such surprises was, of course, the Japanese attack on Pearl Harbor. But there have been other notable examples since then -- China’s intervention in the Korean War in 1950, the Soviet launch of Sputnik in 1957, the Tet Offensive in 1968, and the 9/11 attacks on New York and Washington in 2001.  Indeed, any intelligence analyst’s worst nightmare is failing to provide advanced warning of an adversary’s capability (or plan) to strike.  

And there hasn’t really been any bureaucratic penalty for over-estimating the threat, as the United States did during much of the Cold War – at least not until the invasion of Iraq in 2003.

The Iraq WMD Fiasco

The second U.S. war against Iraq was justified by the need to stop Saddam Hussein’s aggressive pursuit of nuclear and other weapons of mass destruction (WMD).  But, of course, we all now know that Saddam’s WMD had been eliminated before the invasion and his ability to reconstitute his WMD programs had been effectively stymied by the UN.

Political Pandering

I realized then that I had underestimated the willingness of the Director of Central Intelligence and other intelligence officials to support the political desires of the White House.  Daily access to President George W. Bush by the DCI and the rapport between them was important to George Tenet.  The numerous visits of Vice President Dick Cheney to CIA had an impact on the agency’s product.  So did the establishment of a DoD Undersecretary for Intelligence, which created a mechanism for bypassing the intelligence community to deliver unvetted intelligence information to the president.

One of the keystones to the Bush administration’s case that Saddam was reconstituting his nuclear weapons program came from intercepting a delivery of high-strength aluminum tubes to Iraq.  The CIA said these tubes were intended to be used for centrifuges in the production of highly enriched uranium, a critical material used for nuclear weapons.  The Department of Energy, which operated U.S. nuclear weapons labs, correctly assessed that the tubes were not suitable for centrifuges.  (They were actually being used to manufacture artillery rockets.)  INR concurred.

The CIA also assessed that Saddam was importing “yellow cake” from Niger that was needed to manufacture the uranium hexafluoride, which would be fed into the centrifuges.  DOE and INR again challenged the validity of these reports, based on technical analysis and country expertise.

The judgments of the best experts in the U.S. Government on these issues were key to INR’s critical dissent in the 2002 National Intelligence Estimate on Iraq’s WMD Program.  But when the day arrived to finalize (“coordinate”) the estimate, DOE sided with the 15 other members of the intelligence community, delivering what the White House wanted -- the conclusion that Iraq was reconstituting its nuclear weapons program.  INR was thus left to stand alone in disagreeing with the most important judgment in the estimate.  And INR registered it in a conspicuous dissent on the bottom of page one of the Executive Summary at the bottom of the  one-page summary of the estimate that went to the president.

Congressional Cowardice

The U.S. Congress and the press failed to rigorously examine the intelligence on which the administration’s allegations were based.  (The head of INR was never asked to explain his bureau’s dissent in hearings on the NIE held by Congress.)  By mid-October of 2002, following a brief debate, the Congress authorized the use of force against Iraq – an action many Members of Congress hoped would persuade Saddam to allow the return of UN inspectors.  This pressure, backed by a UN Security Council resolution in November, was successful.  The inspectors returned at the end of the month, resumed on-cite inspections, and began filling in the information, which had been absent since their expulsion in 1998.  But their findings over the next three months were spurned by the administration and largely ignored by Congress and the press. 


The invasion was launched in March 2003, but no WMD was found.  The baleful legacy for the Middle East and the U.S. in terms of lives lost and treasure squandered is continuing to be felt, both at home and abroad.  And the decision to invade has now been labeled by people from both political parties as one of the biggest foreign policy blunders in post-Cold War history.

Although the systemic failures ultimately led to a significant reform of the U.S. intelligence community structure and process, those administration officials who were culpable for willfully distorting an already flawed intelligence product were never held to account – either at the polls or in their privileged government positions.

So too did most in Congress escape retribution for their dereliction of duty in failing to revisit the previous year’s authorization to use military force before the invasion.  (In my admittedly biased view, it should have been clear in February 2003 to anyone paying close attention that every evidentiary pillar of support for the administration’s case was collapsing.) 

The Fantasical Thinking of Star Wars

Another insight about threat estimates I’ve gained over the years was the inability of many politicians and military officials to put themselves in the shoes of their adversaries.  This means that decisions made on military measures to meet the threat frequently do not adequately consider either the adversary’s perspective or likely response.

The classical case for me is the U.S. decision in 2002 to withdraw from the Anti-Ballistic Missile Treaty.  That treaty, then 30 years old, had created specific limits on defenses against strategic offensive ballistic missiles, a necessary condition for securing and sustaining reductions in U.S. and Soviet (or Russian) nuclear weapons.  The ABM Treaty had facilitated the limits and reductions achieved by the SALT and START processes.

Moreover, negotiators had agreed on a protocol to the ABM Treaty, which established a dividing line between the strategic missile defense interceptors limited in the treaty, and the theater and tactical missile defense interceptors, which would not affect the strategic balance.  Hence, Patriot and THAAD interceptors, which are used to blunt the effectiveness of ballistic missile use against troop formations, port facilities, air bases, and other point targets, could be produced and deployed without limits under the ABM Treaty.

But when the U.S. revoked the ABM Treaty, the Russians refused to ratify the terms of START II – a treaty on strategic offensive systems, which, among other things, would have eliminated all multiple warheads on Russian ICBMs.  Russia reacted as the U.S. did when faced with Soviet strategic defenses in the 1960s; it increased its strategic offensive efforts to compensate and safeguard the viability of its nuclear deterrent.

The intelligence community has been very reticent to include in its threat assessments the obvious point that strategic missile defense deployments lead to strategic missile offense deployments.  There is no empirical evidence that strategic missile defenses discourage proliferators as its proponents assert – whether it is Russia, China, Pakistan, North Korea, or Iran.  And yet, the policy community moves forward with what one missile defense proponent this month termed: “the long-running principled opposition to ‘reject any negotiated restraints’ on missile defenses. 

The Perfidious Persians

In my third and most recent historical example, I find evidence of both assessment problems I’ve wanted to highlight.  In describing  Iran’s ballistic missile program over the last two decades, the U.S. Government has both overestimated the technological threat and underestimated Iran’s determination to thwart U.S. policy on ballistic missiles. 

Ballistic Missile Balderdash I

I go back aways on this issue.  One of my most interesting professional endeavors was to participate in coordination of a 1999 National Intelligence Estimate on the Foreign Ballistic Missile Threat.  This estimate constituted an official reaction to the extremely influential report on the ballistic missile threat in the previous year of the Rumsfeld Commission.  That commission essentially predicted in the spring of 1998 that, within five years, North Korea, Iraq, and Iran would be able to develop ICBMs (a missile capable of delivering a nuclear warhead 5,500+ km).  An attempted North Korean launch of a satellite a few months later seemed to many to validate Rumsfeld’s dire forecast.

The 1999 NIE assessed that a North Korean ICBM could be launched at any time, an Iranian or Iraqi ICBM could be launched between 2005 and 2010.  In terms of likelihood, the estimate judged that, by 2015: a North Korean ICBM was most likely, an Iranian ICBM was probable; an Iraqi ICBM was possible.

The Rumsfeld Commission report and the NIE helped kill the ABM Treaty and, as collateral damage, the START II agreement.  They helped solidify the congressional view of U.S. missile defense requirements that has lasted 17 years.  They were also spectacularly wrong.  None of these three countries has fielded an ICBM.  When Iraq was invaded in 2003, it was working on a 200 km range ballistic missile. North Korea has come closest, with several launches of a large space rocket that could be used as the basis for developing an ICBM, but the only type of “ICBMs” it has paraded have never flown.  Iran has never tested a missile with a range over 2,000 km. The U.S. military now says that Iran could field an ICBM as soon as 2020, but it is unlikely to have a nuclear warhead for 10-15 years, because of the Iran nuclear deal.

Of course, having squirmed in the uncomfortable position of dissenting from the majority estimates during my years as an intelligence analyst, I jumped on the chance to remind the public in the summer of 2003 (after retiring from the State Department) that Rumsfeld’s five years had expired, asking: where are the ICBMs?  I also enjoyed authoring an Arms Control Association blogpost in 2013, facetiously asking: “What Kind of Glasses Do You Need to See Iranian ICBMs?”

Ballistic Missile Balderdash II

As if exaggerating Iran’s missile prowess was not enough, we now seem to be implying that the Iranians are acting irresponsibly and deceitfully in improving their short- and medium-range ballistic missile force.  This failure to understand and predict Iranian behavior may be more appropriately blamed on the Congress and the press, but both have been aided and abetted by the intelligence community’s silence. 

In the 8-year war that started with Iraq’s invasion of Iran, Saddam Hussein had weapons superiority over Iran across the spectrum: more modern aircraft, missiles, and tanks – and in greater quantities; and an arsenal of chemical weapons along with a willingness to use them.  It was Iraq, which initiated the ballistic missile attacks against Iranian cities.  It was Iraq, which violated the 1925 Protocol against chemical weapons use.  Yet the West supported Iraq in the conflict.

To expect Iran not only to accept the Iran nuclear deal, but also a ban on all medium-range conventionally-armed ballistic missiles, is extremely naïve.  Such missiles are Iran’s only direct deterrent against nuclear-armed Israel.  Moreover, they serve to counterbalance -- although only in small measure – the much larger and more modern air forces of potential enemies Israel, Saudi Arabia, and the U.S.

Yet the tenor of discussion in Washington about the Iran nuclear deal gives the impression that this agreement between Iran and the P5+1 includes limits on ballistic missiles.  Politicians assert incorrectly that Iran is “violating” the deal and/or acting irresponsibly as a regional power to even test missiles with ranges over 300 km.  Was anyone in the intelligence community expecting the Iranians to abandon their medium-range missile program?  Did anyone in the IC fail to understand the compromise resulting in watered down language on ballistic missiles in the new UN Security Council resolution?   Well, I guess the answers to these rhetorical questions are classified, but I think the answer to both is “no.”


My point in these case studies is not to suggest that prognostication is easy or that I have a perfect track record.  (In fact, I joke about my good luck that some of my correct calls have been publicized, while my mistakes remain classified.)

But I would advise that, before one relies on threat assessments, it is wise to ask about the track record of the assessor, to ask for the kind of evidence on which projections are based, and the probability attached to whatever horrendous outcome the assessor says is possible.  Finally, one must look realistically at how the adversary will respond to U.S. actions.  Otherwise, if we’re not careful, we will prepare for a future that is very unlikely – and in a way, which leads to an outcome that leaves us worse off than before.