The Iraq War did not fail because the intelligence was wrong, though some of it was. It failed because the system that processes intelligence was designed to produce a specific output, and it produced that output faithfully. The people who knew better were in the room. Hans Blix was running inspections and finding nothing. Military planners were warning about occupation complexity. Intelligence analysts were flagging the weakness of the WMD case. Regional specialists were predicting sectarian fragmentation. These were credentialed people with direct operational knowledge of the specific questions at issue. They were not recognized as the relevant experts. The people who were recognized as the relevant experts were the ones whose models and conclusions fit the coalition that controlled the aggregation pipeline.
This is not a historical anomaly. It is a description of how national security expertise works as a system, and in 2026 it is running again in real time over Iran.
The Two-Sided Picture
The Max Boot case, a journalist elevated to Senior Fellow in National Security Studies at the Council on Foreign Relations without military service, operational experience, or a relevant academic credential, is commonly read as an indictment of individual credentialing. That reading is too narrow. Boot is not the story. He is the visible half of a two-sided system that only makes sense when both halves are examined together.
The Boot side of the system includes the people the institutions recognize: figures like Bret Stephens, whose authority rests on a BA in political science and a career in commentary; Walter Russell Mead, whose Yale degree was in English literature and who became a leading voice on American grand strategy through journalism and books; David Frum, a speechwriter and lawyer who became a primary architect of the moral language of the War on Terror; Ben Rhodes, whose MFA in creative writing preceded his role as Deputy National Security Advisor for Strategic Communications and chief narrative architect of Obama’s foreign policy; and a substantial cohort of think tank fellows, columnists, and senior advisors whose primary credential is fluency in the coalition’s frameworks and access to its networks. These are not stupid people or fraudulent ones. They are people whose specific skills, narrative construction, coalition alignment, and the ability to translate geopolitical complexity into moralized urgency, are precisely what the system selects for and rewards.
The other side of the system includes the people the institutions do not recognize, despite credentials that dwarf those of the recognized experts. John Mearsheimer holds a PhD from Cornell, served as an Air Force officer, and is by citation count one of the most influential international relations scholars alive. His offense-defense theory and great power competition framework predicted the consequences of NATO expansion and the Ukraine conflict with more accuracy than the interventionist consensus that dismissed him. He is not a Senior Fellow at CFR. Stephen Walt was dean of Harvard’s Kennedy School and has published foundational work in alliance theory and foreign policy analysis. His critiques of American grand strategy have been more empirically accurate than the consensus his institutional peers produced. He is not a primary voice in the policy pipeline. Andrew Bacevich spent twenty-three years in the Army, retired as a colonel, held a PhD from Princeton, and lost a son in the Iraq War he had spent years predicting would fail. Douglas Macgregor is a decorated combat veteran with a PhD in international relations whose operational analysis of the Ukraine conflict diverged from the institutional consensus in ways the subsequent two years partially vindicated. Neither is amplified by the institutions that amplify Boot and Stephens.
The pattern is not that one group has credentials and the other does not. It is that the recognized group has coalition alignment and the marginalized group does not. This is Turner’s core point made visible in personnel: expertise in this domain is a status conferred by institutions that select for narrative loyalty, not a qualification measured by demonstrable competence. The credential is the affiliation. The affiliation requires alignment. Alignment requires accepting the coalition’s models of which threats matter and which do not.
Why This Field Works This Way
The reason national security expertise has this structure rather than the structure of medicine or engineering requires precise explanation, because the difference is not about the character of the people involved. It is about the properties of the domain.
In medicine, reality filters experts. If a physician consistently misdiagnoses, patients deteriorate and die on a timeline short enough to attribute causation. Licensing boards revoke licenses. Malpractice suits create legal liability. The feedback loop between expert judgment and observable outcome is tight enough that the credential system is ultimately disciplined by the world.
In national security, experts filter reality. Outcomes unfold over years and decades. Causality is contested. The counterfactual, what would have happened under a different policy, is permanently inaccessible. The analyst who advocated for the Iraq War in 2003 could not be definitively proven wrong in 2005, or 2010, or 2015, because the outcome could always be attributed to implementation failure rather than conceptual error. The analyst who opposed the war could not be definitively proven right, because the alternative was never tested. In this environment, the credential system has no mechanism for self-correction through outcomes. It corrects, to the extent it corrects at all, through political pressure from outside.
This structural difference produces everything else. Because outcomes are ambiguous, models dominate. Because models are underdetermined, multiple internally consistent frameworks can fit the same observed facts. Because data is classified and selectively disclosed, outsiders cannot independently verify claims. Because accountability is reputational rather than legal, being wrong has no professional consequences within the coalition that recognized you as right. And because the field partly creates the problems it manages, defining threats is not separable from addressing threats: the expert who successfully argues that Iran is an existential danger has also created the professional context in which his own expertise is most necessary.
In this environment, the rational career strategy for an aspiring national security expert is not to maximize predictive accuracy. It is to maximize coalition alignment. The two strategies produce different people in positions of authority, and the people produced by the second strategy are systematically worse at the thing the title claims to certify.
The Iraq Proof of Concept
The Iraq War is the cleanest available historical demonstration of how the system fails, because the failure is now fully documented and the alternative models that were suppressed are now visible.
The dominant coalition in 2002 and 2003 controlled the aggregation pipeline: the senior policymakers, the aligned think tanks, the media amplification, and the intelligence presentation. It defined the admissible reality: Iraq likely had weapons of mass destruction or was close to having them, the risk of inaction was catastrophic, and postwar transition was manageable. It used safety language in precisely the way this project has mapped: safety meant preventing a WMD threat from materializing, and questioning that framing was classified as naivety or complacency. The experts who appeared on the Sunday shows, testified before Congress, and wrote the op-eds that shaped public support were the ones whose models fit this framework.
The marginalized coalition included Blix and his inspection team, who were finding no evidence of active weapons programs. It included State Department regional specialists who produced the Future of Iraq Project’s detailed assessments of postwar governance challenges. It included military planners who estimated that occupation would require far more forces than the administration intended to deploy. It included academic realists whose models predicted that regime change would produce sectarian fragmentation rather than democratic transformation. These people had credentials that exceeded those of many of the recognized experts. They were not amplified. Their safety claims, that the primary risk was self-created instability rather than Iraqi capability, were not admitted into the decision-making pipeline.
The outcome is documented history. No stockpiles. Prolonged insurgency. Regional destabilization. Hundreds of thousands of deaths and trillions of dollars in costs. The risks the dominant coalition backgrounded became the dominant reality. The risks it elevated turned out not to exist. The credentialed dissenters were right. The narrative-aligned recognized experts were wrong. No career consequences followed for the wrong. No vindication followed for the right. Boot still holds his fellowship. Mearsheimer is still outside the policy pipeline.
The Iran War in Real Time
On February 28, 2026, the United States and Israel launched large-scale strikes on Iran’s nuclear and military infrastructure. Iran retaliated across the region. The Strait of Hormuz was disrupted. Global oil markets moved sharply. The escalation pathways multiplied. The same structure that produced the Iraq War is visible in real time, and the two competing expert systems are producing their competing realities from the same facts.
The escalation coalition defines admissible reality as follows: Iran was a near-term strategic threat whose capability required denial, the strikes restore deterrence credibility, and the primary risk is an Iran that interprets restraint as weakness and accelerates its nuclear program. Safety means preventing a stronger Iran. The risks this coalition backgrounds are escalation spiral, regional war expansion, economic shock, and the historical pattern of military interventions producing long-term instability rather than the security improvement they promise.
The restraint coalition defines admissible reality differently: escalation itself is the primary threat, Iran can impose asymmetric costs that exceed the benefits of capability denial, regime collapse is unlikely, and the Hormuz disruption represents exactly the kind of systemic shock that a war with Iran was supposed to prevent. Safety means avoiding a regional war. The risks this coalition backgrounds are Iranian nuclear breakout, deterrence erosion, and the long-term consequences of allowing Iran to interpret de-escalation as a pattern it can exploit.
Both coalitions are using the same public data. Both are applying internally coherent frameworks. Neither can demonstrate its conclusion is correct because the relevant counterfactual, what would have happened without the strikes, is permanently inaccessible. Which conclusion becomes policy depends on which coalition controls the aggregation pipeline at the moment the decisions are made.
The early warning signs of Iraq-style failure are visible. Justifications have already shifted from nuclear threat to deterrence to credibility. Dissenting voices are being framed as naive or insufficiently serious about security. Worst-case scenarios dominate the dominant coalition’s public arguments. Second-order risks, economic shock, regional war expansion, long-term occupation dynamics, are present in the conversation but not driving it. Narrative certainty exceeds evidentiary certainty, which is the signature of a coalition managing its reality rather than tracking it.
This does not mean the strikes were wrong. It means the system producing the expert consensus supporting the strikes has the same structural features as the system that produced the expert consensus supporting the Iraq invasion. Whether the outcome will be similar depends on facts that are not yet known. What is known is that the excluded experts, the restraint coalition’s scholars and practitioners, are making specific predictions about escalation dynamics that should be tracked against outcomes with the same rigor we now apply retrospectively to Iraq.
Three Audiences, Three Wars
The system that produces these competing expert realities delivers them to different audiences who are not experiencing the same conflict.
The high-tier audience, the policy professionals, finance sector leaders, senior journalists, and institutional actors who consume prestige media and think tank briefings, experiences the Iran war as a structured policy problem with legible tradeoffs between deterrence and escalation, uncertainty acknowledged but bounded, and the assumption that institutional actors can contain the outcomes. Their definition of safety is system stability. Their blind spot is the gap between managed complexity and lived consequence.
The mid-tier audience, the educated professionals who consume a mix of prestige and networked sources, experiences the Iran war as a narrative that does not quite stabilize, where experts disagree more than they admit and the risk feels larger than the official framing presents. Their definition of safety is accurate understanding. Their characteristic experience is the anxiety of sensing contradictions they cannot resolve, of oscillating between competing models without a way to adjudicate between them.
The low-tier audience, the working-class and less institutionally embedded people who consume primarily networked media, talk radio, and direct social media feeds, experiences the Iran war as immediate visible consequence: rising fuel prices, supply chain disruption, images of violence, and a sense that the people in charge do not have the situation under control and that ordinary people will absorb the costs. Their definition of safety is daily stability. Their blind spot is limited access to strategic context. Their advantage is unmediated access to the consequences the high-tier audience models in the abstract.
The policy is made in the first reality. It is debated in the second. It is paid for in the third. The experts whose credentials are recognized by the institutions that make the policy live entirely in the first reality. The people who pay the costs live in the third. This is the optionality gap applied to foreign policy: the people who design the interventions can exit the consequences, and the people who cannot exit the consequences have no recognized voice in the design.
The Media Tiers That Transmit the Gap
The three audience realities are produced and maintained by three media tiers that interact but do not do the same work.
Prestige media, the Times, the Post, the major networks, set the baseline legitimacy of the expert consensus. They define what is serious, which analysts are credible, and what the acceptable range of debate contains. In the Iran context, they stabilize the dominant coalition’s framework as the default reality to which policymakers must respond, while filtering out interpretations that would collapse the framework entirely. They did not invent the expert consensus. They certify it.
The think tank and policy ecosystem generates the interpretive frameworks that prestige media then amplifies. CFR, Brookings, AEI, FDD, and their aligned institutions produce the op-eds, reports, and briefings that supply the language, models, and authorized conclusions that shape what policymakers treat as the menu of serious options. This is the tier where Boot’s fellowship does its structural work: not in any specific argument Boot makes, but in the way that the CFR’s institutional authority certifies the interventionist framework as the default serious position from which deviation requires justification.
Networked media, the Substack writers, podcasters, X accounts, and YouTube analysts who constitute Gurri’s revolt of the public, provides the real-time challenge to the stabilized consensus. It amplifies dissenting experts who lack institutional recognition. It surfaces OSINT and footage that the managed consensus cannot easily absorb. It makes visible the gap between what the recognized experts predicted and what is actually happening. It does not have consistent standards or gatekeeping, which is a real limitation. But it has something the upper tiers lack: it cannot be managed by the coalition that controls the aggregation pipeline.
The three tiers interact in a direction that the upper tiers have historically controlled but can no longer fully contain. Think tanks feed prestige media, which legitimizes certain experts, which shapes what policymakers treat as real. But networked media now visibly contradicts the upper tiers in real time, exposes the gap between expert prediction and observed outcome, and amplifies the credentialed dissenters the upper tiers exclude. The result is the simultaneous existence of multiple incompatible expert realities, each backed by people with plausible claims to authority, none capable of achieving the stable consensus the system was designed to produce.
The Full Picture
The complete picture that all of this produces can be stated in a sequence of propositions that follow from each other.
In national security, expertise is not primarily a technical credential. It is a coalition position. Institutions confer expertise by appointing people to roles that carry institutional authority. The credential is the affiliation.
Coalition alignment is the primary selection criterion because the domain’s structural features, delayed outcomes, restricted data, underdetermined models, and reputational rather than legal accountability, make it impossible to measure expertise by accuracy. The coalition selects for narrative competence and framework loyalty because these are what it can measure.
The system therefore produces two groups: people with high institutional recognition and moderate technical credentials who are aligned with the dominant framework, and people with high technical credentials and direct operational knowledge who are marginalized because their models challenge the framework. The first group defines admissible reality. The second group’s reality is inadmissible regardless of its accuracy.
The safety claims of the recognized group determine what counts as danger. The safety concerns of the marginalized group are the ones that tend to materialize as outcomes, because they are the risks the recognized group’s frameworks are designed to background.
The costs of the resulting failures are distributed through the three audience tiers in inverse proportion to participation in the decision. The high-tier audience that makes the policy can exit the consequences. The low-tier audience that cannot exit the consequences has no recognized voice in the decision.
Gurri’s information revolution has disrupted the upper tiers’ capacity to stabilize their version of reality, but has not replaced it with a system that is better at tracking accuracy. What it has done is make the gap between managed reality and lived reality visible in real time to the people who pay for the gap. That visibility is the precondition for the kind of accountability that the system was designed to prevent, and whether it produces accountability or merely produces noise is the open political question of the present moment.
Turner’s framework suggests the answer cannot come from within the system, because the system selects against the people who would produce it. It can only come from outside, from a democratic public that is large enough, coherent enough, and sufficiently informed about the pattern to withdraw recognition from a credential system that has demonstrated its disconnection from the thing it claims to certify.
That is the Max Boot lesson, stated fully. He is not the problem. He is the indicator. The problem is a system that produces him as its representative expert while excluding Mearsheimer, Bacevich, and Blix. And the solution is not better credentials. It is accountability for outcomes, applied by people who cannot exit the consequences of being wrong.
