Best Defense

The Flynn report (IV): Cordesman’s take

The Flynn report (IV): Cordesman’s take

On the other hand, Anthony Cordesman of CSIS is one old school intel guy who likes the Flynn report. Here is his view of it:

Fixing Intel is one of the most insightful reports I have ever read on combatintelligence, and one that tracks all too well with the lessons that should be learned from Vietnam and Iraq.

I was Director of Intelligence Assessment in the Office of the Secretary of Defense at the end of the Vietnam War, and had to write a post mortem on the collapse. While it was a different war, and the collapse occurred under very different conditions, intelligence failures represented the same tendency to focus on the threat and ignore the range of politic-military and economic factors affecting South Vietnam. This was coupled to excessive classification and above all, to reporting systems that left military advisors largely in charge of assessing the ARVN while intelligence focused on heavily compartmented approaches to the threat.

This leads me to make some additional suggestions regarding the improvements that are need in both intelligence and the analysis of the war:

Intelligence has never really come to grips with the problem of net assessment. The intelligence community seems to have largely backed way from net assessment. So, however, have the US military or the Department of Defense. It is possible to argue that net assessment should be the function of plan, operations, and operations research and not intelligence. As Fixing Intel points out, however, intelligence must look beyond the threat and do so at every level. This sometimes may mean crossing the line into assessing the impact of plans, operations, and civilian activity. Yet, intelligence may well be the best place to conduct both net assessment and fusion analysis in a war that involve so many foreign actors involved in so many different activities.

Wars like Afghanistan are not red or threat side versus US or blue wars. They are dominated by the performance of threat versus host country forces, each of which is fragmented into  different regional, ethnic, sectarian, and tribal groups. Intelligence collection and analysis should be more capable of direct net assessment of these factions than any other element of the military and US analysis community. Friendlies and allies are never going to fully share US goals and interests, and key friendlies and allies will often be divided, suspect, and sometimes covertly hostile.

A realistic net intelligence assessment is needed to deal with the full mix of foreign activities and motives. It also is need to deal with problems like corruption and ties to power brokers, ANSF officers and personnel with ties to the enemy and separate interests, and the fact that the host country is and must be a major intelligence target or will be a critically understated problem in every aspect of operations and planning.

Compartmented analysis that separates intelligence on the threat and partner, mentor, and trainer analysis of host country forces We lost the war in Vietnam for many reasons, but one was that we never had truly honest assessments of ARVN forces, and direct comparisons of ARVN and threat forces done in detailed net assessment terms. In fact, our training and rating experience in Iraq and Afghanistan closely tracks with the mistakes of Vietnam (and our brief training effort in Lebanon in the early 1980s.) The CM readiness ratings for Afghan forces have so far been unreliable at best, and have been worst in providing credible ratings of the units described as CM-1 or the lead. They do not honestly reflect combat performance, actual leadership quality, and how units evolve or devolve over time seeming to ignore real world factors like attrition, overdeployment and combat fatigue, and actual equipment holdings and readiness. There often are other reports covering such issues as well as the loyalties of key officers and officials but often in forms that make direct net assessment and comparability impossible.

Someone is needed to provide an objective, outside view of host country and ANSF performance and loyalty, and this need to be done in a net assessment context. If this is not the task of the intelligence community, whose task is it?

Assessing Allies as an Intelligence Target:  This is equally true of assessments of the activities of our NATO/ISAF allies. Each pursues somewhat different goals, and sometimes operates under sharp national restrictions and caveats. There often is no unity of effort within a given allied country, and there are significant differences between some allied country military and aid efforts. These too are legitimate intelligence targets and the proper subject of net assessment. Moreover, as Fixing Intel points out, it is far too easy to focus on secrecy in this case host country and allied sensitivity. Here, it is important to point out that it took years to overcome this reality in NATO. Ironically, it was not until the MBFR exercise in the 1960s that it became clear just how damaging it was to ignore the fact that there were five different national concepts of how to fight a ground war, in the central region (ignoring two more in NORTHAG), and similar differences in planning for the air war plus two different internal NATO approaches to air operations. It took transparency and dialog not secrecy and fear of confrontation to deal with these issues.

A War for Political Influence and Control But Also a War of Political Attrition:  Fixing Intel also points out in a number of places that the war is political and that it is critical to know the in country situation down to the district level. The paper may well be error, however, in stating that, for the most part, that is where the war is being fought, which means,  unavoidably, that is where it is being won or lost.

All counterinsurgency is local in one sense, but it is increasingly international in another. Insurgents understand that they can conduct attacks and strategic communications designed to exploit US-host country government, internal ethnic and sectarian, and intra-NATO/ISAF fracture lines and sensitivities. Anyone who reads Mao, or the host of Chinese case studies of the war against the Kuomintang, realizes that Korea and Vietnam were scarcely the first wars in which insurgents attacked both the national government and outside advisors at the political and military level.

In the case of Afghanistan, it is critical to know the degree to which the Taliban and Al Qaida explicitly seek to exploit such political differences in the host country political structure, in US strategic and political commitment to the war, and within NATO and ISAF countries. As the  paper points out, these are strategic and not tactical issues. They also, however, are at least as critical over time as understanding the political dimension of war at the district level. 

The Political, Prompt Justice, and Shadow Government Role of the Taliban and Other Insurgents: Looking at the Taliban and Insurgents Ideology, Political Tactics, and Shadow Structures.

This analysis also requires more focus on what each elements of the insurgency is doing in terms of promoting its ideology, it local and broader political actions, how it affects local economics and the justice system, and how it is establishing mixes of shadow governments, networks of influence, and stay behinds and infiltrators. Intelligence needs to look beyond identifying the bad guys, and examine how they operate in influencing and controlling the population as objectively as possible. This does not come from focusing on negative actions and abuses, important as these are to both understanding the Taliban, Haqqani, Al Qaida, etc. The history of insurgencies is usually a history of their ability to mix incentives with ruthless behavior, and the quality of their ability to change, co-opt, or destroy local power structures without alienating the population. In the past, far too much analysis portrayed insurgents only in negative terms and assumed they were unpopular. 

As analysts like Fred and Kim Kagan have pointed out, far too little effort went into establishing the size, character, and actions of the shadow governments and other insurgent actors in given areas. The same was true of the links between insurgents and contractors, officials, officers, and key aspects of the economy like narcotics, transportation, etc. If the Afghan insurgents are as rigid as AQI was during the critical phases of the Anbar campaign, this may not matter, but one should look at other movements that began with rigid approaches and adapted over time like the KDP, Shining Path, etc.  

Counterintelligence: Counterintelligence will be particularly important as progress is made in the Clear, Hold, and Build phases. Past insurgencies show that insurgent networks not only learn how to disperse and hide under tactical pressure, but become extremely sophistication in infiltrating, deploying sleepers and stay behinds.

The failure to honestly address these issues was critical in Vietnam. The community addressed part of the problem, but could never come to grips with just how many individuals in South Vietnam had ties to, provided services to, or were part of the threat. 

These issues will be even more complicated in Afghanistan than they were in Vietnam or Iraq. Dual loyalties or willingness to deal with any source power to protect, self/family/clan are more complex and more scattered. So is the impact of the fear the US will lose or leave, the need to deal with the Taliban to survive or prosper, the lack of reason to support the Afghan government, the impact of tribal and other differences, etc.

It is easy to talk about clear, but it is far from obvious what this really means, or how long it will take. It will be further complicated if amnesty programs take hold and provide a potential cover for insurgents. Even today, most vetting is more theory than practice in a young society with few real personal records, and so many divided loyalties.

Afghan and Allied HUMINT: The paper does an excellent job of highlighting the need for grassroots intelligence and HUMINT at the battalion level and below. It does not, however, address the need to develop more systematic efforts to create joint Afghan, allied, and US intelligence fusion and tie HUMINT together at higher levels.

Transferring Intelligence Responsibility I do not find plans to make significant transfers of responsibility to the ANSF as early as mid-2011 to be credible. I do find plans to do this as soon as possible to be absolutely essential. This means that creating truly effective host country intelligence structures is a critical aspect of Fixing Intel.

Population Centers and Stability Operations Information Centers: The paper may focus too much on the term District, and be driven too much by the fact that the fighting to date has been largely tactical and outside the main population centers. Virtually every element of Shape, Clear, Hold, Build, and Transfer will have to focus on key population clusters or centers to be effective.

Like the operations involved, intelligence and other assessments will generally cross district lines and/or be more urban (in an Afghan sense) than rural. The Helmand River Valley and Khandahar are just two key examples. They also illustrate the fact that different intelligence collection, analysis, and metrics will often be needed to deal with each major population cluster, and that standardizing them on a national level will be dysfunctional.

Fixing Intel focuses on ready access and transparency analyzing district level stability  operations, but this may well be even more critical in terms of population clusters — where efforts to cut across any stovepipes and administrative lines and do so in net assessment terms may be even more critical.

As the paper points out, this will be equally critical at the Regional Command level. Regional boundaries often do not lend themselves natural to population centric operations. They were not designed for this purpose. Regional Commands are, however, the logical Fusion Centers to look at both the entire structure of Shape, Clear, Hold, Build, and Transfer not simply the red side and do so both by cluster and in comparative ways. Such analysis is likely to be far more productive than region-wide analysis that cannot be tied to the operational and strategic goals of providing successful operations focused on key populations.

Dealing with the Threat from the State Department and Aid Community To be successful, however, the analysis of Shape, Clear, Hold, Build, and Transfer will have to address issues that the State Department, AID, and intelligence community have so far failed to address in any convincing form: how to measure success in Shape, Clear, Hold, Build, and Transfer at the political, economic, and rule of law levels.

Neither the narratives nor the metrics developed to date in either Afghanistan or Iraq have much credibility. Far too often they are what David Kilcullen refers to as input metrics: statements that money has been spent, a project has been started or completed, or that actions have had some positive effect without any reference in terms of how well it meets a validated requirement or the scale of impact on a given area or part of the population. In Iraq’s case, this has been partially balanced by shifting from input metrics to survey data in the maturity model, but the end results have uncertain focus and credibility. The entire aid community US, UNAMA, allied, and NGO has become a nightmare of stovepiped programs without meaningful validation of requirements, proper financial transparency and auditing, and measures of effectiveness. A lack of basic management effort is complicated by national branding, rapid rotation, and the use of contractors and officials that are corrupt and/or linked to the Taliban at least to the extent of buying it off to complete projects.

There is a need for far better intelligence on aid activity than exists to date, and this needs to be tied to narratives and metrics that focus on what aid does to win the war and do so at the scale and with lasting effect necessary to deal with entire population centers. In other cases, national econometric data are generated but without any convincing break down by key element of the population, credible analysis of unemployment and underemployment as it impacts on perceptions and the willingness of young men to act as insurgents, or data on key factors like income distribution. There is a chronic lack of sectoral analysis covering key areas like agriculture. 

Broad stoplight color coding is often used to describe the quality of governance and justice system often emphasizing positive trends and with little regard to the seriousness of corruption or whether service good or bad are provided to broad parts of the population. Any positive trend is reported as progress, regardless of how much of the population it actually affects and how given elements of the population see the trends in Afghan politics, governance, and rule of law/prompt justice.

Key issues like corruption, and equity of effort, and the role of power brokers are dealt with only in generalities if at all although understanding the impact of the networks of corrupt  officers, officials, contractors, and power brokers will be as important to overall success as understanding the networks of insurgents.

This again raises questions about the function of intelligence in providing the kind of fusion needed to win the war. Some independent structure is needed that can integrate assessment of the military situation with the political, economic, demographic, and ideological trends that determine the success of Shape, Clear, Hold, Build, and Transfer in each key population center. These efforts need to be independent, honest self-critical of progress, and cut across  stovepipes and bureaucratic lines. Intelligence may not be the place to do this, but if not, small Strategic Assessment Groups need to be established for each major population objective in the campaign that include intelligence, plans and ops, and civil representatives.

Pakistan and Iran: Good as Fixing Intel is, it tends to perpetuate one of the key problems in the intelligence effort throughout the war. Pakistan, Iran, and other neighboring states are key part of the conflict. Furthermore, no amount of polite, politically correct rhetoric, can disguise the fact that Pakistan is an ally only to the limited point where it perceives its interest coincide with ours, it is incentivized to cooperate, or it is pressured to cooperate. Dancing around this reality has essentially made analysis of Pakistan an entirely separate and compartmented area. The role of Iran presents similar problems. Effective fusion cannot ignore the impact of the other half of the war.

Strategic Communications: As a final comment, Fixing Intel repeatedly focuses on the need for internal transparency and to fight the tendency towards overclassification and compartmentation. This reflects a valid concern, and a tacit recognition of the fact that one never knows whether one is better off shooting the enemy, or ones own public affairs and security officers. All three actions generally have the same positive effect.

More seriously, the last eight months have seen a major deterioration in confidence that the war can be won, and in support for the conflict by both the US and its NATO/ISAF allies, This has been compounded by a sharp drop in the President’s popularity, and the weak handling and mistakes surrounding the drafting and supporting efforts for his new strategy in Washington. It may not be tactful to point out just how much the popular war has moved towards calls for an exit strategy, and how serious the level of Congressional and media doubt has become. The fact is, however, that the country team must now demonstrate competence, unity, and progress or  lose the war.

This raises a key issue not addressed in Fixing Intel. How can the release of unclassified assessments and metrics reverse this situation and help win. Until the recent release of new unclassified metrics by USCENTCOM, no element of the US military or Executive Branch began to address this issue. The fact is, however, that intelligence should be a key element of a process of strategic communications that helps to correct the mistakes made in presenting and supporting the President’s speech, that reinforces the broad themes raised in the testimony to Congress that followed, that establishes broad credibility, and shapes as much of the reporting on the war and perceptions of its progress as possible. This is as critical a part of Fixing Intel as any addressed in the paper.