Congressional Research Service ˜ The Library of Congress

CRS Report for CongressReceived through the CRS Web

Order Code RL32561

Risk Management and Critical InfrastructureProtection: Assessing, Integrating, and Managing

Threats, Vulnerabilities and Consequences

Updated February 4, 2005

John MoteffSpecialist in Science and Technology Policy

Resources, Science, and Industry Division

Risk Management and Critical Infrastructure Protection:Assessing, Integrating, and Managing Threats,

Vulnerabilities, and Consequences


The 9/11 Commission recommended that efforts to protect various modes oftransportation and allocation of federal assistance to state and local governmentsshould be based on an assessment of risk. In doing so, the Commission wasreiterating existing federal policy regarding the protection of all the nation’s criticalinfrastructures. The Homeland Security Act of 2002 (P.L. 107-296) and otherAdministration documents have assigned the Department of Homeland Securityspecific duties associated with coordinating the nation’s efforts to protect its criticalinfrastructure, including using a risk management approach to set priorities. Manyof these duties have been delegated to the Information Analysis and InfrastructureProtection (IA/IP) Directorate.

Risk assessment involves the integration of threat, vulnerability, andconsequence information. Risk management involves deciding which protectivemeasures to take based on an agreed upon risk reduction strategy. Manymodels/methodologies have been developed by which threats, vulnerabilities, andrisks are integrated and then used to inform the allocation of resources to reducethose risks. For the most part, these methodologies consist of the followingelements, performed, more or less, in the following order.

! identify assets and identify which are most critical! identify, characterize, and assess threats! assess the vulnerability of critical assets to specific threats! determine the risk (i.e. the expected consequences of specific types

of attacks on specific assets)! identify ways to reduce those risks! prioritize risk reduction measures based on a strategy

The IA/IP Directorate has been accumulating a list of infrastructure assets

(specific sites and facilities). From this list the Directorate is selecting assets thathave been judged to be critical from a national point of view. The Directorateintends to assess the vulnerability of all the assets on this shorter list. According toDirectorate officials, vulnerability assessments and threat information are consideredwhen determining the risk each asset poses to the nation. This risk assessment isthen used to prioritize subsequent additional protection activities. The IA/IPDirectorate’s efforts to date, however, raise several concerns, ranging from theprocess and criteria used to populate its lists of assets, its prioritization strategy, andthe extent to which the Directorate is coordinating its efforts with the intelligencecommunity and other agencies both internal and external to the Department. Thisreport will be updated as needed.


Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2IA/IP’s Responsibilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2A Generic Model for Assessing and Integrating Threat, Vulnerability,

and Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4Assessments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4Using Assessments to Identify and Prioritize Risk Reduction

Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11Status of DHS’s Implementation of Its Critical Infrastructure Protection

Effort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12Progress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Questions and Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15Identifying Assets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15Selecting High Priority Assets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19Assessing Threat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20Assessing Vulnerabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20Assessing Risks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Risk Mitigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22Prioritizing Protection Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

1 The Intelligence Reform and Terrorism Prevention Act of 2004 (S. 2845, P.L. 108-458),legislating some of the recommendations of the Commission’s report, included arequirement to develop a National Strategy for Transportation Security that includes thedevelopment of risk-based priorities.

Risk Management and Critical InfrastructureProtection: Assessing, Integrating, andManaging Threats, Vulnerabilities, and



As part of its chapter on a global strategy for protecting the United States againstfuture terrorist attacks, the 9/11 Commission recommended that efforts to protectvarious modes of transportation and allocation of federal assistance to state and localgovernments should be based on an assessment of risk.1 In doing so, theCommission was affirming existing federal policy regarding the protection of all thenation’s critical infrastructures. The Homeland Security Act of 2002 and otherAdministration documents have assigned the Department of Homeland Securityspecific duties associated with coordinating the nation’s efforts to protect its criticalinfrastructure. Many of these duties have been delegated to the Information Analysisand Infrastructure Protection (IA/IP) Directorate. In particular, the IA/IP Directorateis to integrate threat assessments with vulnerability assessments in an effort toidentify and manage the risk associated with possible terrorist attacks on the nation’scritical infrastructure. By doing so, the Directorate is to help the nation set prioritiesand take cost-effective protective measures.

This report is meant to support congressional oversight by discussing, in moredetail, what this task entails and issues that need to be addressed. In particular, thereport defines terms (e.g. threat, vulnerability, and risk), discusses how they fittogether in a systematic analysis, describes processes and techniques that have beenused to assess them, and discusses how the results of that analysis can informresource allocation and policy.

While the IA/IP Directorate has been given this task as one of its primarymissions, similar activities are being undertaken by other agencies under otherauthorities and by the private sector and states and local governments. Therefore,this report also discusses the Department’s role in coordinating and/or integratingthese activities.


2 Office of Homeland Security, National Strategy for Homeland Security, July 2002.3 Ibid. p. 33.4 Ibid. p. 64.


IA/IP’s Responsibilities

The Homeland Security Act of 2002 and other Administration documents haveassigned the Department of Homeland Security specific duties associated withcoordinating the nation’s efforts to protect its critical infrastructure. Many of theduties discussed below have been delegated to the Information Analysis andInfrastructure Protection Directorate.

The National Strategy for Homeland Security,2 anticipating the establishmentof the Department of Homeland Security, stated:

! “… the Department would build and maintain a complete, current,and accurate assessment of vulnerabilities and preparedness ofcritical targets across critical infrastructure sectors…[Thisassessment will] guide the rational long-term investment of effortand resources.3”

! “… we must carefully weigh the benefit of each homeland securityendeavor and only allocate resources where the benefit of reducingrisk is worth the amount of additional cost.4”

Among the specific tasks delegated to the Undersecretary for InformationAnalysis and Infrastructure Protection by Section 201(d) of the Homeland SecurityAct of 2002 (P.L. 107-296, enacted November 25, 2002) were:

! “… identify and assess the nature and scope of terrorist threats to thehomeland;”

! “… understand such threats in light of actual and potentialvulnerabilities of the homeland;”

! “… carry out comprehensive assessments of the vulnerabilities of thekey resources and critical infrastructures of the United States,including the performance of risk assessments to determine the riskposed by particular types of terrorist attacks within the United States….”

! “… integrate relevant information, analyses, and vulnerabilityassessments … in order to identify priorities for protective andsupport measures ….”

! “… develop a comprehensive national plan for securing the keyresources and critical infrastructure of the United States ….”

! “… recommend measures necessary to protect the key resources andcritical infrastructure of the United States ….”


5 Office of Homeland Security, The National Strategy for the Physical Protection of CriticalInfrastructures and Key Assets, February 2003.6 Ibid. p. 23.7 Homeland Security Presidential Directive Number 7, Critical Infrastructure Identification,Prioritization, and Protection, December 17, 2003.8 The Clinton Administration referred to these as Lead Agencies in its Presidential DecisionDirective Number 63 (PDD-63, May 1998). HSPD-7 supercedes PDD-63 in those instanceswhere the two disagree. 9 The Department did not meet this deadline. A draft plan is still in review. TheDepartment intends to release elements of the plan in 2005. See, See CQ HomelandSecurity, Jan. 28, 2005, “Still Waiting: Plan to Protect Critical Infrastructure Overdue fromDHS,”at []. This sitewas last viewed on February 4, 2005. It is available only by subscription.10 Just as one example, the 9/11 Commission Report (released July 22, 2004, see page 396)when discussing the basis upon which federal resources should be allocated to states andlocalities, stated that such assistance should be based “strictly on an assessment of risks andvulnerabilities.” Later, in the next paragraph, it stated “the allocation of funds should bebased on an assessment of threats and vulnerabilities.” In the next paragraph it stated that


The National Strategy for the Physical Protection of Critical Infrastructure andKey Assets 5 stated:

! “DHS, in collaboration with other key stakeholders, will develop auniform methodology for identifying facilities, systems, andfunctions with national-level criticality to help establish federal,state, and local government, and the private-sector protectionpriorities. Using this methodology, DHS will build a comprehensivedatabase to catalog these critical facility, systems, and functions.6”

Homeland Security Presidential Directive Number 7 (HSPD-7)7 stated that theSecretary of Homeland Security was responsible for coordinating the overall nationaleffort to identify, prioritize, and protect critical infrastructure and key resources. TheDirective assigned Sector Specific Agencies8 the responsibility of conducting orfacilitating vulnerability assessments of their sector, and encouraging the use of riskmanagement strategies to protect against or mitigate the effects of attacks againstcritical infrastructures or key resources. It also gave the Secretary to the end ofcalendar year 2004 to produce a comprehensive, integrated National Plan for CriticalInfrastructure and Key Resources Protection.9 That Plan shall include a strategy anda summary of activities to be undertaken to: define and prioritize, reduce thevulnerability of, and coordinate the protection of critical infrastructure and keyresources.

The terms “vulnerabilities,” “threats,” “risk,” “integrated,” and “prioritize” areused repeatedly in the documents cited above. However, none of the documentsdefined these terms or discussed how they were to be integrated and used. Also, inhearings, articles in the press, and other public discourse these terms are used loosely,clouding the intent of what is being proposed or discussed.10 What might seem trivial


10 (…continued)resources “must be allocated according to vulnerabilities.”11 Roper, Carl. A. Risk Management for Security Professionals, Butterworth-Heinemann.1999.

differences in definitions can make a big difference in policy and implementation.The following section provides definitions and a generic model for integrating themin a systematic way.

A Generic Model for Assessing and Integrating Threat,Vulnerability, and Risk

Many models/methodologies have been developed by which threats,vulnerabilities, and risks are integrated and then used to inform the cost-effectiveallocation of resources to reduce those risks. For this report, CRS reviewedvulnerability assessment models or methodologies, including some developed andused, to varying degrees, in certain selected sectors (electric power, ports, oil andgas). These are listed in the Reference section of this report. In addition, this reportdraws upon information contained in a book by Carl Roper entitled Risk Managementfor Security Professionals.11 Essential elements of these models/methods have beendistilled and are presented below. They may provide some guidance in overseeingDHS’s methodology as it is developed and employed.

For the most part, each of the methodologies reviewed consist of certainelements. These elements can be divided into: assessments per se; and, the use of theassessments to make decisions. The elements are performed, more or less, in thefollowing sequence:

Assessments! identify assets and identify which are most critical! identify, characterize, and assess threats! assess the vulnerability of critical assets to specific threats! determine the risk (i.e. the expected consequences of specific types

of attacks on specific assets)Using Assessments to Identify and Prioritize Risk Reduction Activities ! identify and characterize ways to reduce those risks! prioritize risk reduction activities based on a risk reduction strategy


Identifying Assets and Determining Criticality. The infrastructure of afacility, a company, or an economic sector, consists of an array of assets which arenecessary for the production and/or delivery of a good or service. Similarly, theinfrastructure of a city, state, or nation consists of an array of assets necessary for theeconomic and social activity of the city and region, and the public health and welfareof its citizens. The first step in the process is to determine which infrastructure assetsto include in the study. The American Chemistry Council, the Chlorine Institute, andthe Synthetic Organic Chemical Manufacturers Association, in their Site Security


12 American Petroleum Institute and the National Petrochemical and Refiners Association,Security Vulnerability Assessment Methodology for the Petroleum and PetrochemicalIndustries, May 2003, p. 4.

Guidelines for the U.S. Chemistry Industry, broadly define assets as people, property,and information. Roper’s Risk Management for Security Professionals (and DOE’sEnergy Infrastructure Risk Management Checklists for Small and Medium SizedEnergy Facilities) broadly define assets as people, activities and operations,information, facilities (installations), and equipment and materials.

The methodologies reviewed do not provide a definitive list of such assets butsuggest which ones might be considered. For example, people assets may includeemployees, customers, and/or the surrounding community. Property usually includesa long list of physical assets like buildings, vehicles, production equipment, storagetanks, control equipment, raw materials, power, water, communication systems,information systems, office equipment, supplies, etc. Information could includeproduct designs, formulae, process data, operational data, business strategies,financial data, employee data, etc. Roper’s examples of activities and operationsassets include such things as intelligence gathering and special training programs.Many methodologies suggest considering, initially, as broad a set of assets as isreasonable.

However, not every asset is as important as another. In order to focusassessment resources, all of the methodologies reviewed suggest that the assessmentshould focus on those assets judged to be most critical. Criticality is typicallydefined as a measure of the consequences associated with the loss or degradation ofa particular asset. The more the loss of an asset threatens the survival or viability ofits owners, of those located nearby, or of others who depend on it (including thenation as a whole), the more critical it becomes.

Consequences can be categorized in a number of ways: economic; financial;environmental; health and safety; technological; operational; and, time. For example,a process control center may be essential for the safe production of a particularproduct. Its loss, or inability to function properly, could result not only in adisruption of production (with its concomitant loss of revenue and additional costsassociated with replacing the lost capability), but it might also result in the loss oflife, property damage, or environmental damage, if the process being controlledinvolves hazardous materials. The loss of an asset might also reduce a firm’scompetitive advantage, not only because of the financial costs associated with itsloss, but also because of the loss of technological advantage or loss of uniqueknowledge or information that would be difficult to replace or reproduce. Individualfirms, too, have to worry about loss of reputation. The American Petroleum Instituteand the National Petrochemical and Refiners Association (API/NPRA) in theirSecurity Vulnerability Assessment Methodology for the Petroleum and PetrochemicalIndustries also suggested considering the possibility of “excessive media exposureand resulting public hysteria that may affect people that may be far removed from theactual event location.12”


While the immediate impact is important, so, too, is the amount of time andresources required to replace the lost capability. If losing the asset results in a largeimmediate disruption, but the asset can be replaced quickly and cheaply, or there arecost-effective substitutes, the total consequence may not be so great. Alternatively,the loss of an asset resulting in a small immediate consequence, but which continuesfor a long period of time because of the difficulty in reconstituting the lost capability,may result in a much greater total loss.

Another issue which decision makers may consider is if the loss of a particularasset could lead to cascading effects, not only within the facility or the company, butalso cascading effects that might affect other infrastructures. For example, the lossof electric power can lead to problems in the supply of safe drinking water. The lossof a key communications node can impair the function of ATM machines.

The initial set of assets are categorized by their degree of criticality. Typicallythe degree of criticality is assessed qualitatively as high, medium, or low, or somevariation of this type of measure. However, even if assessed qualitatively, a numberof methodologies suggest being specific about what kind of consequence qualifiesan asset to be placed in each category. For example, the electric utility sectormethodology suggests that a highly critical asset might be one whose loss wouldrequire an immediate response by a company’s board of directors, or whose losscarries with it the possibility of off-site fatalities, property damage in excess of aspecified amount of dollars, or the interruption of operations for more than aspecified amount of time. Alternatively, an asset whose loss results in no injuries,or shuts down operations for only a few days, may be designated as having lowcriticality.

For those sectors not vertically integrated, ownership of infrastructure assetsmay span a number of firms, or industries. Whoever is doing the analysis may feelconstrained to consider only those assets owned and operated by the analyst oranalyst’s client. For example, transmission assets (whether pipeline, electric, orcommunication) may not be owned or operated by the same firms that produce thecommodity being transmitted. Both the production assets and the transmissionassets, however, are key elements of the overall infrastructure. Also, a firm may relyon the output from a specific asset owned and operated by someone else. The usermay consider that asset critical, but the owner and operator may not. Some of themethodologies reviewed encourage the analyst to also consider (or at least accountfor) the vulnerability of those assets owned or operated by someone else that providecritical input into the system being analyzed. These “interdependency” problemshave been talked about, mainly in the context of inter-sector dependencies (e.g thereliance of water systems on electric power), but they may also exist intra-sector.The interdependency issue is both a technical one (i.e. identifying them) and apolitical/legal one (i.e. how can entity A induce entity B to protect an asset).

Identify, Characterize, and Assess Threat. Roper and the API/NPRAdefine threat as “any indication, circumstance or event with the potential to cause loss


13 American Petroleum Institute, op. cit., p. 5. 14 Roper, op. cit. , p. 43.15 This quote is taken from the Government Accountability Office testimony, HomelandSecurity: Key Elements of a Risk Management Approach, GAO-02-150T, before theSubcommittee on National Security, Veteran’s Affairs, and International Relations, HouseCommittee on Government Reform, October 21, 2001. It is used in several of the othermethodologies reviewed.

or damage to an asset.13” Roper includes an additional definition: “The intention andcapability of an adversary to undertake actions that would be detrimental to U.S.interests.14”

To be helpful in assessing vulnerability and risk, threats need to be characterizedin some detail. Important characteristics include type (e.g. insider, terrorist, military,or environmental (e.g. hurricane, tornado)); intent or motivation; triggers (i.e eventsthat might initiate an attack); capability (e.g. skills, specific knowledge, access tomaterials or equipment); methods (e.g. use of individual suicide bombers, truckbombs, assault, cyber); and trends (what techniques have groups used in the past orhave experimented with, etc.).

Information useful to characterizing the threat can come from the intelligencecommunity, law enforcement, specialists, news reports, analysis and investigationsof past incidents, received threats, or “red teams” whose purpose is to “think” likea terrorist. Threat assessment typically also involves assumptions and speculationsince information on specific threats may be scant, incomplete, or vague.

Once potential threats have been identified (both generically, e.g. terrorists, andspecifically, e.g. Al Qaeda) and characterized, a threat assessment estimates the“likelihood of adversary activity against a given asset or group of assets.15” Thelikelihood of an attack is a function of at least two parameters: a) whether or not theasset represents a tempting target based on the goals and motivation of the adversary(i.e. would a successful attack on that asset further the goals and objectives of theattacker); and, b) whether the adversary has the capability to attack the asset byvarious methods. Other parameters to consider include past history of such attacksagainst such targets by the same adversary or by others, the availability of the assetas a target (e.g. is the location of the target fixed or does it change and how would theadversary know of the target’s existence or movement, etc.). The asset’svulnerability to various methods of attack (determined in the next step) may alsoaffect the attractiveness of the asset as a target.

As an example of a threat assessment technique, the U.S. Coast Guard, using anexpert panel made up of Coast Guard subject matter and risk experts, evaluated thelikelihood of 12 different attack modes against 50 different potential targets (i.e. 600scenarios). Attack modes included “… boat loaded with explosives exploding alongside a docked tank vessel,” or “… tank vessel being commandeered and intentionallydamaged.” The Coast Guard also considered scenarios where port assets could bestolen or commandeered and used as a weapon or used to transport terrorists orterrorism materials. Potential targets included various types of vessels (includingferries), container facilities, water intakes, utility pipelines, hazardous materials


16 Roper, op. cit., p. 63.17 American Petroleum Institute, op. cit., p. 5.18 Federal Register, Department of Homeland Security, Coast Guard, Implementation ofNational Maritime Security Initiatives, Vol. 68, No. 126, July 1, 2003, p. 39245.

barges, etc. The panel of experts judged the credibility of each scenario. Forexample, using a military vessel for transporting terrorists or terrorism materials wasjudged not to be credible given the inherent security measures in place, but anexternal attack on a military target was considered credible. Each credible scenariowas assigned one of 5 threat levels representing the perceived probability (likelihood)of it occurring, after considering the hostile group’s intent, its capabilities, priorincidents, and any existing intelligence.

The Electricity Sector’s methodology uses a checklist which asks for thespecific attack mode (such as the use of explosives, truck bomb, or cyber attack) andwhether it is likely that such an attack would be carried out by: a) an individual; orb) by an assault team of up to five members. In this case, the analyst is to identifylikely targets for each type of attack scenario and the objective that the adversarywould achieve by such an attack.

Likelihood can be measured quantitatively, by assigning it a probability (e.g. an85% chance of occurring), or qualitatively, such as “Very High Threat Level,” whichmight mean there is a credible threat, with a demonstrated capability, and it hashappened before. As with criticality, a number of methodologies suggested specificcriteria be used to define what would constitute varying threat levels.

A threat assessment need not be static in time. Threats (i.e. the likelihood thatan adversary may attack) may rise and fall over time, depending on events,anniversary dates, an increase in capability, or the need for the adversary to reassertitself. Intelligence may detect activity that indicates pre-attack activity or a lull insuch activity, or an explicit threat may be made.

Assess Vulnerability. Roper defines vulnerability as a “weakness that canbe exploited to gain access to a given asset.16” The API/NPRA expands thisdefinition to include “… and subsequent destruction or theft of [the] … asset.17” TheCoast Guard defines vulnerability as “the conditional probability of success giventhat a threat scenario occurs.18”

Weaknesses, like criticality, can be categorized in a number of ways: physical(accessibility, relative locations, visibility, toughness, strength, etc.), technical(susceptible to cyber attack, energy surges, contamination, eavesdropping, etc.),operational (policies, procedures, personal habits ), organizational (e.g. would takingout headquarters severely disrupt operations), etc.

Existing countermeasures may already exist to address these weaknesses. Avulnerability assessment must evaluate the reliability and effectiveness of thoseexisting countermeasures in detail. For example, security guards may provide acertain degree of deterrence against unauthorized access to a certain asset. However,


19 Roper, op. cit., p. 73.20 Federal Register, op. cit., p. 39245.21 American Petroleum Institute, op. cit., p. 3.

to assess their effectiveness, a number of additional questions may need to be asked.For example, how many security guards are on duty? Do they patrol or monitorsurveillance equipment? How equipped or well trained are they to delay or repulsean attempt to gain access? Have they successfully repulsed any attempt to gainunauthorized access?

Vulnerabilities are assessed by the analyst against specific attacks. API/NPRAidentifies three steps to assessing vulnerabilities: 1) determine how an adversarycould carry out a specific kind of attack against a specific asset (or group of assets);2) evaluate existing countermeasures for their reliability and their effectiveness todeter, detect, or delay the specific attack; and 3) estimate current state of vulnerabilityand assign it a value. Specific types of attacks can be informed by the precedingthreat assessment.

The Coast Guard measured vulnerability of potential targets for each attackscenario in four areas: 1) is the target available (i.e. is it present and/or predictableas it relates to the adversary’s ability to plan and operate); 2) is it accessible (i.e. howeasily can the adversary get to or near the target); 3) what are the “organic”countermeasures in place (i.e. what is the existing security plan, communicationcapabilities, intrusion detection systems, guard force, etc.); and, 4) is the target hard(i.e. based on the target’s design complexity and material construction characteristics,how effectively can it withstand the attack). Each of these four vectors wereevaluated on a level of 1 to 5, with each level corresponding to a assigned probabilityof a successful attack. By comparison, the electricity sector process measuresvulnerability as a probability that existing countermeasures can mitigate specificattack scenarios (e.g. probability of surviving attack = 80%).

Alternatively, the analyst can value vulnerability qualitatively. For example,a “highly vulnerable” asset might be one that is highly attractive as a target, for whichno countermeasures currently exist against a highly credible threat. An asset withlow vulnerability might be one that has multiple effective countermeasures.

Assess Risk. Risk implies uncertain consequences. Roper defines risk as the“… probability of loss or damage, and its impact …19” The Coast Guard refers to arisk assessment as “ essentially an estimate of the expected losses should a specifictarget/attack scenario occur.20” “Expected” loss is determined by multiplying theestimated adverse impact caused by a successful threat/attack scenario by theprobabilities associated with threat and vulnerability. API/NPRA defines risk as “afunction of: consequences of a successful attack against an asset; and, likelihood ofa successful attack against an asset.21” “Likelihood” is defined as “a function of: theattractiveness of the target to the adversary [based on the adversary’s intent and thetarget’s perceived value to the adversary], degree of threat [based on adversary’s


22 Ibid.23 Consequences, too, could be uncertain. For example, it may be determined that in theabove scenario, a successful attack may cause a distribution of possible deaths between zeroand 500 people.24 This formulation assumes that the uncertainties in this case are independent, which inmany cases is not accurate. The attractiveness of a target (an element in determining threat)may very much depend on its vulnerability. Likewise, the consequence of an attack mayalso depend on a target’s vulnerability. This complicates the calculation.

capabilities], and degree of vulnerability of the asset.22” An important point is thatrisk, as defined here, is a discounted measure of consequence; i.e. discounted by theuncertainty of what might happen (see the example given below).

As noted in the first step, impact can be categorized in a number of ways.Impact or consequences may be measured more precisely at this point in the process,however, to better inform the prioritization of risk reduction steps that follows.

The Coast Guard considered six categories of impact: death/injury; economic;environmental; national defense; symbolic effect; and secondary national securityissues. Each target/attack scenario measured the potential impact in each of thesecategories on a severity scale from 1 to 5 (from low to catastrophic). The assignedscale value was based on benchmarks. The API/NPRA, which used a similarconstruct, suggested the following benchmarks for its severity scale. The severityof death and injury varied from high to low depending on whether they occurred off-site or on-site, and whether they were certain or possible. The severity ofenvironmental damage again varied from high to low depending on whether it waslarge scale (spreading off-site) or small scale (staying on-site). The severity offinancial losses or economic disruptions were valued on threshold dollar amounts andtime-frames.

The analyst can also try to measure risk quantitatively. For example, for aspecific target/attack scenario, the analysis may determine that there is a 50/50chance (i.e. we don’t know) that the adversary will try to attack a particulargovernment building. But, if they did, there is a 75% chance that they would use atruck bomb (i.e we are pretty sure that if they attack they would try to use a truckbomb). If they try use a truck bomb, the vulnerability assessment determined thatthey would have a 30% chance of succeeding (i.e. if they try, there is a good chancethat the current protective measures will prevent them from getting close enough tothe building to bring it down). The consequences of a successful attack (bringing thebuilding down) could be 500 people killed and $300 million in property damage.23

The risk associated with this scenario would be:

expected loss = (consequence) x ( probability that an attack will occur) x(conditional probability that the attacker uses a truck bomb) x (theconditional probability that they would be successful)24, or

(500 people killed + $300 million in damage) x (.5) x (.75) x (.3), or


25 Note: the risk in this scenario is not 500 people dead, but 56 expected deaths. That is notto say that if an attack were actually successfully carried out only 56 people might die. Infact, in this scenario, it has been judged that 500 people would likely die. If one chose touse the 500 potential deaths in subsequent decisions, they would be called risk averse in thisconstruct. Taking a risk averse position is a legitimate policy option. See further discussionon risk aversion below.26 Again, dependencies between threat, risk, and consequences need to be considered.

risk = 56 expected deaths and $33.8 million in expected damages.25

Risk is often measured qualitatively (e.g. high, medium, low). Since

consequences may be measured along a number of different vectors, and threat andvulnerability have been measured separately, a qualitative measure of risk must havesome criteria for integrating the number of different qualitative measures. Forexample, how should the assessment decide what risk rating to give a medium threatagainst a highly vulnerable target that would have a low death/injury impact, amedium environmental impact, but a high short-term financial impact? Does thisscenario equal a high, medium, or low level of risk?

Using Assessments to Identify and Prioritize Risk ReductionActivities.

Identify Ways to Reduce Risk. Risks can be reduced in a number of ways:by reducing threats (e.g. through eliminating or intercepting the adversary before hestrikes); by reducing vulnerabilities (e.g. harden or toughen the asset to withstand theattack); or, by reducing the impact or consequences (e.g. build back-ups systems orisolate facilities from major populations). For each potential countermeasure, thebenefit in risk reduction should also be determined.26 More than one countermeasuremay exist for a particular asset, or one countermeasure may reduce the risk for anumber of assets. Multiple countermeasures should be assessed together todetermine their net effects. The analyst should also assess the feasibility of thecountermeasure.

The cost of each countermeasure must also be determined. Costs, too, aremultidimensional. There may be up-front financial costs with associated materials,equipment, installation, and training. There are also longer term operational costs ofthe new protective measures, including maintenance and repair. There may also beoperational costs associated with changes to overall operations. Costs also includetime and impact on staff, customers, and vendors, etc. Expenditures on theprotection of assets also results in opportunity costs, i.e. costs associated with notbeing able to invest those resources in something else.

Prioritize and Decide In What to Invest. Once a set of countermeasureshave been assessed and characterized by their impact on risk, feasibility, and cost,priorities may be set. Decision makers would have to come to a consensus on whichrisk reduction strategy to use to set priorities.

Most of the methods reviewed suggest a cost-effective selection process (i.e.implementation of the risk-reduction method(s) should not cost more than the benefit


27 Roper, op. cit., p. 88.

derived by the reduced risk). Cost-effectiveness could also imply that the countryinvest in risk reduction to the point where the marginal cost to society equals themarginal benefit. Alternatively, given a fixed budget, cost-effectiveness might implyinvesting in protections that maximize the benefits for that investment.Countermeasures that lower risk to a number of assets may prove to be most cost-effective. Also, focusing attention on those assets associated with the highest risksmay yield the greatest risk reduction and be one way to implement a cost-effectiveapproach.

While cost-effectiveness is usually the recommended measure for settingpriorities, decision makers may use others. For example, decision makers may berisk averse. In other words, even if the chance of an attack is small, or the potentialtarget is not particularly vulnerable, the consequences may be too adverse tocontemplate. In this case, decision makers may wish to bear the costs of additionalprotection that exceed the “expected” reduction in risk. Roper notes, however, that,in general, protection costs should not exceed a reasonable percentage of the totalvalue of the asset.27

Another measure by which to select protective actions might be to favormaximizing the number or geographical distribution of assets for which risks arereduced. Alternatively, decision makers might want to focus efforts on reducing aspecific threat scenario (e.g. dirty bombs) or protecting specific targets (e.g. eventswhere large numbers of people attend).

The electric utility checklist states that the ultimate goal of risk management isto select and implement security improvements to achieve an “acceptable level ofrisk” at an acceptable cost. The concept of acceptable risk is mentioned in a numberof methodologies, and it needs to be determined by decision makers.

After selecting which protective measures to pursue, programs, responsibilities,and mechanisms for implementing them must be established. Many of the reviewedmethodologies conclude with the recommendation to revisit the analysis on a regularbasis.

Status of DHS’s Implementation of Its Critical InfrastructureProtection Effort

Programming. The IA/IP Directorate’s FY2005 budget justificationdocument provided a glimpse into how DHS intends to implement itsresponsibilities in critical infrastructure protection. Below is a list of budgetactivities that most relate to the topic of this report.

! Threat Determination and Assessment: This activity involves thedevelopment of tools and techniques to help model terroristorganizations, to develop a terrorist capabilities baseline, and tofacilitate collaboration and fusion of intelligence information and thecoordinated analysis of that information.


! Critical Infrastructure and Asset Identification: This activityinvolves identifying critical infrastructure and assets, assessingpotential risk of successful attacks to those assets, and prioritizingprotective measures.

! Critical Infrastructure Vulnerability and Field Assessments: Thisactivity supports specialized teams of experts that are sent to highpriority sites to assess their vulnerabilities.

! Infrastructure and Key Asset Protection Implementation: Thisactivity supports an array of services offered by the IA/IP Directorateto assist state, local, and private stakeholders in risk mitigationmeasures for high priority assets. These services include training,protection and response planning, pilot programs, technologytransfers, and sharing of best practices.

! National Infrastructure Risk Analysis: This activity supports thedevelopment of comprehensive risk and vulnerability analyses on anational scale. These analyses are cross-sector in nature, focusingon problems affecting multiple infrastructures. The goal is toprovide timely, actionable information that is more meaningful toindustry. It is not clear how this activity differs from the CriticalInfrastructure and Asset Identification activity mentioned above.

! Threat/Vulnerability/Asset Databases: This activity supports thedevelopment, operation, and maintenance of an integrated datawarehouse of threat, vulnerability, and risks posed to specificfacilities and assets (including the probability of attack andassociated consequences). The description of this activity impliesthat data would be located in disparate places, and technology wouldbe developed to allow access and integration of the information.

! Competitive Analysis and Evaluation: This activity supportsactivities that evaluate the effectiveness of IA/IP products andprocesses. This could involve “red teaming” to test, via exercises,gaming, and alternate hypothesizing analyses, IA/IP assessments andprotective actions.

Progress. Testimony given by the Undersecretary for Information Analysisand Infrastructure Protection, Frank Libutti, before the House AppropriationsCommittee’s Subcommittee on Homeland Security, April 1, 2004, provided anotherglimpse into the IA/IP Directorate’s process and progress in integrating threat,vulnerability, and risk and using that information to set help set priorities. In thattestimony, the Undersecretary stated that the Directorate had 28,000 sites or facilitiesin a national infrastructure database. A few days later, this number had grown to


28 The article cited below makes reference to testimony by the Undersecretary before a jointhearing of the House Select Committee’s subcommittees on Cybersecurity, Science andResearch & Development and Infrastructure and Border Security held on April 21, 2004.See, [], “DHS needsasset info.” This site was last visited on February 4, 2005.29 See, “Terror Target List Way Behind,” USA Today, December 9, 2004. p.A1.30 It is not clear if this number, too, has grown since the testimony was given.31 Report language in the Senate Appropriations Committee’s report accompanying itsFY2005 Department of Homeland Security Appropriations bill (see reference below) statedthat 150 of these were expected to be completed in FY2004, with another 400 expected tobe completed in FY2005. This only would be about one third of the 1700 assets whichIA/IP had said it would like to assess by the end of the calender year, 2004.

33,000.28 By the end of 2004, the number of assets in the database has reportedlyreached 80,000.29 According to the testimony, these are sites, facilities, and assetsthat state Homeland Security Advisers have identified as being critical. They includepublic and private assets.

According to the above testimony, the Directorate, working with state, local, andprivate stakeholders, had identified 1,700 assets from the larger list on which itintends to conduct, or lead, vulnerability field assessments.30 Presumably, theseassets have been identified by IA/IP as being the most critical. According to thetestimony, IA/IP intended to lead vulnerability assessments teams to each of these1,700 assets in 2004.31 Based on these vulnerability assessments and the associatedrisk (the Undersecretary did not elaborate on how risk would be determined), theUndersecretary stated that IA/IP intends to follow-up with asset owners/operators tohelp them develop “operational plans” for improving the security (both physical andcyber) of these assets, if needed.

A statement by the Director of the Protection Services Division (a divisionwithin the IA/IP Directorate), James McDonnell, before the House GovernmentReform Committee’s Subcommittee on Technology, Information Policy,Intergovernmental Affairs, and the Census on March 30, 2004 offered someadditional insight. According to his statement, the IA/IP Directorate identified thosesites or assets that may be most attractive as targets and have sent or plan to send ateam of specialized security experts to the site to assess its vulnerabilities and howthose vulnerabilities might be exploited. The IA/IP Directorate then maps threat andvulnerability information to determine risk. Base on the risk information, theDirectorate then prioritizes the implementation of protective measures that addressvulnerabilities. No where in this statement are the criteria that govern these decisions(assessing target attractiveness, mapping threat and vulnerability to determine risk,prioritization) mentioned.

According to the Director, protective measures focus on near-term technical andprocedural changes that can harden the site against attack. The FY2005 budgetrequest referred to the type of protective measures that IA/IP may recommend. Theseincluded 1) including the detection of weapons of mass destruction in thedevelopment of protection plans; 2) measures to disrupt attack planning by makinginformation gathering and surveillance by terrorist difficult; and 3) ensure counter-


32 U.S. Congress, Senate, Appropriations Committee, Report accompanying S. 2537, S.Rept.108-280, June 17, 2004.33 Stephen McHale, Deputy Administrator, Transportation Security Administration,Statement before the Subcommittee on Infrastructure and Border Security, House SelectCommittee on Homeland Security, May 12, 2004. Also, Rear Admiral David S. Belz,Assistant Commandant for Operations, U.S. Coast Guard,, Statement before the HouseSelect Committee on Homeland Security, May 5, 2004.34 See, “Terror Target List Way Behind,” USA Today, December 9, 2004. p. A1.

assault capabilities. The Senate Appropriations Committee’s report on DHS’sappropriation bill32 made reference to Buffer Zone Protection Plans (BZPP) whichthe Protective Services Division assists stakeholders in developing. The Director, inthe statement cited above, described the BZPP as being a community-basedapproach, incorporating local law enforcement and emergency personnel into asecurity plan that extends beyond the fence (i.e. outside the property limits of the siteor facility).

In addition to the above mentioned efforts, and as required by HSPD-7, theIA/IP Directorate is in the process of developing a national plan for protecting criticalinfrastructure. The national plan will use as its foundation sector-level plans, to bedeveloped cooperatively by Sector Specific Agencies and representatives of theirsector. According to testimony by DHS officials responsible for the transportationsector,33 these sector level plans are in progress (at least within the TransportationSecurity Administration and the Coast Guard). The IA/IP Directorate has preparedguidance for developing Sector Specific Plans that contain many of the stepsdiscussed in the generic model above, and which require Sector Specific Agenciesto be specific about methodology, criteria, etc.

From the above discussion it appears the IA/IP Directorate is in the process ofcarrying out the fundamental steps identified in the preceding discussion. However,some questions and issues remain.

Questions and Issues

Identifying Assets

By the end of 2004, the IA/IP Directorate has compiled a list of 80,000 assets.Policy makers, from both sides of the aisle, however, have raised several concernsabout this list.34 From where does the information for this list come? Is there asystematic approach to generating this list? How comprehensive is it? How accurateis the data? In what form does this information exist?

According to the testimony of the Undersecretary discussed above, the list ispopulated by assets identified by state Homeland Security Advisors. The process bywhich the states have provided this information is not discussed in the testimony.However, according to Democratic members of the House Select Committee on


35 The letter, dated August 3, 2004, is available on the Democrat’s Select Committee onHomeland Security website: see []. The letter was lastviewed at this site on February 4, 2004 and can be found on the Critical Infrastructure link.36 The Office of Domestic Preparedness (ODP) was transferred from the Department ofJustice to the Department of Homeland Security by the Homeland Security Act of 2002.The ODP has since been integrated with the Office of State and Local GovernmentCoordination, established by the Homeland Security Act, into the Office of State and LocalGovernment Coordination and Preparedness (OSLGCP). The OSLGCP now administersthe grants programs. 37 According to the guidance, threat assessment determines the relative likelihood of aknown potential threat element attempting to attack using a weapon of mass destruction.A potential threat element includes any group or individual in which there are allegationsor information indicating a possibility of the unlawful use of force or violence, specificallythe utilization of weapons of mass destruction. Threat factors include evidence of theexistence, violent history, intentions, motivations, targeting, and weapons of massdestruction capability of a known potential threat element. For each potential target, thevulnerability assessment is to consider factors such as target visibility, its criticality to the


Homeland Security, in a letter to the Secretary of Homeland Security,35 DHS’sefforts to coordinate submissions by the states’ Homeland Security Advisors have notincluded specific guidelines or a specific reporting mechanism, and in some cases,state officials contend they were never contacted even though assets from their statesappear on the list.

There is a mechanism in place by which the IA/IP could received state and localinformation on critical infrastructure assets. The Office of State and LocalGovernment Coordination and Preparedness (OSLGCP) administers two grantprograms that give states the opportunity to identify critical infrastructure assets: theState Homeland Security Grant Program and the Urban Area Security InitiativeGrants, both of which are to be submitted through a state’s designated HomelandSecurity Advisor. While both of these grant programs focus primarily on the needsof first responders, they both also support activities related to critical infrastructureprotection, including the purchase of equipment such as chemical, biological,radiological, nuclear, and explosive (CBRNE) detectors, physical security equipmentsuch as surveillance cameras, fences, cybersecurity hardware and software,interoperable communications equipment, etc.

Allocation of funds through the State grant program is based on a formuladetermined by Congress in the U.S.A. PATRIOT Act (P.L. 107-56). All states, theDistrict of Columbia, and U.S. territories receive funds. States must develop a StateHomeland Security Strategy before they can spend their funds.

Urban Areas grants (to which have been added Port Security Grants and TransitSystem Security Grants) are allocated to cities, selected by DHS, based on a formuladeveloped by the Office of Domestic Preparedness (ODP).36 While the grantapplication guidelines do not elaborate, the formula considers current threatestimates, critical assets within the urban area, and population density. Accordingto the application guidelines, grantees must provide a risk assessment for review.The risk assessment must also include threat and vulnerability assessments.37 The


37 (…continued)jurisdiction, its impact outside the jurisdiction, the potential access of a threat element to thetarget, the target’s population capacity, and the potential for mass casualties. Note that theguidance on vulnerability assessment mixes vulnerability and consequence considerations.38 DHS published its interim rule on the procedures associated with the sharing andhandling of information designated as Critical Infrastructure Information in February 2004.Federal Register, No. 69, Vol. 34, pp 8074-8089. February 20, 2004.

risk assessment informs a capabilities and a needs assessment also required of thestates. The states must submit these assessments to qualify for the grant and tojustify their expenditures of grant resources.

Assets eligible for protection in either grant program include drinking watersystems; primary data storage facilities, stock markets and major banking centers;chemical facilities located near large populations; major power generators generatingin excess of 2 gigawatts of power and whose disruption would affect a regional grid;hydroelectric facilities and dams that produce in excess of 2 gigawatts of power andcould result in catastrophic loss of life if breached; nuclear power plants; electricsubstations critical to service areas in excess of one million people; rail and highwaybridges and tunnels and whose loss would cause catastrophic economic loss and/orloss of life; natural gas pipelines with throughput in excess of 3000 billion cubic feet;liquified natural gas facilities; major petroleum handling facilities includingpipelines, ports, refineries, and terminals; and mass transit systems.

Nothing in the grant guidelines or in the testimony cited above indicates thatthe IA/IP Directorate was involved in the development of the criteria for these grantprograms. Apparently the IA/IP Directorate did not generate its list using these grantapplications. Nor is it clear what role, if any, the IA/IP Directorate’s subsequentvulnerability and/or risk assessments have informed the allocations of grantresources. However, the Senate Appropriation’s Committee, in the appropriationsbill report cited above, stated that it was encouraged by the initial cooperationbetween the IA/IP Directorate and ODP, and expected it to improve and for the IA/IPDirectorate to be a full partner in the grant award process.

There are other potential sources of information for identifying assets. Title II,Subtitle B of the Homeland Security Act of 2002 established a new category ofinformation called Critical Infrastructure Information. Critical InfrastructureInformation is defined as “information not customarily in the public domain andrelated to the security of critical infrastructure ….” It includes a broad range ofinformation, including information related to actual, potential or threatenedinterference or attacks that could compromise or incapacitate a critical infrastructure,the ability of a critical infrastructure to resist interference or attack, and any plannedor past operational problems. Such information, if voluntarily provided to the DHSstate, local, or private entities, either directly or indirectly through another agency,is accorded a number of protections.38 The purpose for encouraging owners/operatorsof critical infrastructures to supply this information is to help DHS assessvulnerabilities and threat, mitigation strategies and to monitor the operational statusof the infrastructures. It is not known how much critical infrastructure informationhas been submitted to DHS, and to what extent, if any, it has helped to populate the


39 For example, the Department of Energy for electric utilities, the Environmental ProtectionAgency for drinking water, and Treasury for financial services.40 Based on a personal communication with industry official, September 29, 2003.41 See, CQ Homeland Security. July 29, 2004. []. This site was lastviewed on February 4, 2005. It is available by subscription only.

list of 80,000 assets. Nor is it clear how DHS uses the information it receives or ifit is helpful at all to IA/IP in accomplishing its mission.

Another source of asset information could be other agencies, especially theSector Specific Agencies, that have been tasked with helping their sectors assess theirvulnerabilities and to encourage them to use risk management techniques to setpriorities. While Sector Specific Agencies are apparently in the midst of developingSector Specific Plans for submission to the IA/IP Directorate, many of theseAgencies have already engaged in facilitating vulnerability assessments within theirsectors.39 It is not clear what the connection is between the specific sector planningprocess and the on-going assessments associated with the list of 1,700 assets. Noris it clear to what extent, if any, the list of 80,000 assets generated by the IA/IPDirectorate, or any of the subsequent vulnerability assessments led by the IA/IPDirectorate, reflect or duplicate information accumulated by these Agencies. Doesthe IA/IP Directorate use these outside analyses in lieu of doing their ownvulnerability assessments or does the Directorate use the information to help directits own separate analyses?

Other potential sources could be public (open source) and private databases. Forexample, there are private databases used by U.S. industry that map and monitor theoperations of the nation’s systems of electric grids. Much of this information maybe proprietary.

Regardless of where DHS’s data comes from, the accuracy of the data isimportant. For example, DHS compiled a preliminary list of critical infrastructurein electric power and circulated that list to certain infrastructure owners for comment.Among utilities operators, there was some confusion as to why certain assets wereincluded in the list, since some were not currently in use, while others which wereconsidered critical by industry standards were not on the list.40 The previouslyreferenced letter from the Democrats to Secretary Ridge also stated that IA/IP’s listincluded sites that no longer existed. Reportedly, DHS initially had Disneyland inCalifornia located in Los Angeles County, not Orange Country.41 For criticality andpotential consequences to be assessed accurately, location and utilization must beknown.

Nor is it clear in what form the IA/IP Directorate has collected its information,whether it is being entered into a single database, or if it exists in multiple databases.If the latter, is it formatted consistently and is it accessible to all those who need touse it? Also, it is not clear if the IA/IP’s current list of 80, 000 assets, or its refinedlist of 1,700 assets, constitute the foundation of the integrated data warehousesreferenced in IA/IP’s budget justification above.


Selecting High Priority Assets

On what basis did the IA/IP Directorate select, in consultation with otherstakeholders, the 1,700 assets for further attention? According to the Undersecretary,in his testimony referenced above, the 1,700 assets were ones with a crediblepotential for loss of life and loss of citizen confidence and that these impacts wouldbe felt nationally. He described these assets as ones the nation cannot afford to lose.

Roper, and other methodologies reviewed for this report, recommended thecriteria for assessing the level of criticality be specific. For example, at what pointis the impact of an attack felt nationally versus one felt primarily locally orregionally? How many casualties rise to the level of having a national impact? What level of economic impact or what measure of reduced confidence would rankan asset as nationally critical? Again, the answers to these questions would probablyrequire a consensus among decision makers.

The list of assets that qualify for Urban Area Security grants provides a littlemore specificity for some of the assets. For example, power generating plants inexcess of 2 gigawatts, and whose disruption would affect a regional grid or electricsubstations critical to service areas in excess of one million people, qualify for grants.Presumably these thresholds were arrived at based on some assessment of the relativeimpacts the loss of those assets would have. Other qualifying assets are less wellcharacterized, such as rail and highway bridges over major waterways that, ifdestroyed, would cause catastrophic economic loss. What constitutes catastrophic?

An example of an analysis that provides more detail as to what might beconsidered nationally critical can be found in a white paper entitled InteragencyPaper on Sound Practices to Strengthen the Resilience of the U.S. Financial System.The authors of the white paper, the Federal Reserve Board, U.S. Security ExchangeCommission, and Office of the Comptroller of the Currency, determined that adisruption in the services of certain “core clearing and settlement” organizationscould, by virtue of their market share, present a systemic risk to the smoothoperations of the financial markets they service. The paper defined “systemic risk”as the risk that failure of one participant to meet its required obligations will causeother participants to be unable to meet their obligations when due, causing significantliquidity or credit problems and threatening the stability of financial markets. Thewhite paper identified a threshold market share, above which a firm’s plansassociated with back-up capacity, geographic location, and recovery would be subjectto review by the appropriate agency.

Another important issue is how to adjudicate disagreements about what shouldand should not be on the list. What the IA/IP Directorate might, through careful andaccurate analysis, determine to be critical (or not) at the national level may notnecessarily agree with what other stakeholders or policy makers consider to becritical or not (and even need not agree as long as the data analyzed is consistent,accurate, and comprehensive).


42 Congressional Research Service. Homeland Security: Intelligence Support. CRS ReportRS21283, by Richard Best.

Assessing Threat

The Homeland Security Act assigned to the IA/IP Directorate the responsibilityof integrating all-source information in order to identify and assess the nature andscope of terrorist threats against the homeland and to detect and identify threats ofterrorism against the United States. However, shortly after the act was passed, theBush Administration, in January 2003, established the Terrorist Threat IntegrationCenter (TTIC), within the intelligence community. Many observers felt that theTTIC assumed many of the same responsibilities of the IA/IP Directorate. TheHomeland Security Act designated DHS a member of the intelligence communityand it has a seat at the TTIC. The issues and concerns associated with the divisionof labor between TTIC and the IA/IP Directorate is beyond the scope of this report.For information on this, see CRS Report RS21283.42 There are, however, two keyquestions that are relevant to this report. Is there a consistent characterization of thethreat used throughout the intelligence community and made available to the IA/IPDirectorate and beyond to other stakeholders? Is that characterization usedconsistently to inform the teams sent out to do vulnerability assessments or thoseagencies and other stakeholders tasked with assessing the vulnerabilities of thesectors for which they are responsible?

In a written response to a list of questions submitted by the House SelectCommittee on Homeland Security, after its joint hearing with the JudiciaryCommittee on July 22, 2003, the DHS gave a few examples of threat information thathas been passed on to stakeholders to inform their decision-making. These includedan advisory to the oil and gas industry that discussed recent terrorist behavioroverseas, including general characteristics of terrorist surveillance; and, an advisoryon how nitrocellulose could be used by terrorists to create a bomb. The answer didnot discuss if or how this information may have influenced subsequent vulnerabilityor risk assessments.

Another issue is whether the IA/IP Directorate values all threats equally. Forexample, Al Qaeda has demonstrated capabilities in a number of attack modes (e.g.bombs, hijacking and piloting planes). But, their capability in other attack modes arenot necessarily as well developed. How does IA/IP consider this in their threatassessments?

Assessing Vulnerabilities

The testimony of IA/IP officials implies that the IA/IP Directorate will eitherperform or lead vulnerability assessments in the field. However, much of the workbeing performed by the IA/IP Directorate is being done by contractors or details fromother agencies until the IA/IP Directorate is more fully staffed. Also, as mentionedearlier, it is not clear if the IA/IP Directorate uses the vulnerability assessmentsperformed by other agencies or stakeholders in lieu of doing their own. A keyquestion is whether or not contractors, details, or other agencies and stakeholdersfollow a similar protocol in doing their vulnerability assessments? As discussed


43 For example, the Coast Guard considered six categories of consequences, includingdeath/injury, economic, environmental and symbolic impacts, all equally weighted, andassigned a value of 1 to 5 to each of these, based on severity. An overall level of risk wasdetermined by the sum total value.

above, there are many models by which to assess vulnerabilities. It is not necessarilyimportant that all vulnerability assessments follow the exact same methodology.However, Congress might want to ensure that certain general considerations areincluded. For example, what vectors of vulnerability are examined? Aredependencies on assets beyond the control of the immediate owner/operatorconsidered? As discussed above, is the threat sufficiently characterized?

Assessing Risks

As discussed above, risk is a function of threat, vulnerability, and consequences.What consequences does the IA/IP Directorate consider when assessing risk? TheUndersecretary of Information Analysis and Infrastructure Protection, in thetestimony discussed above, mentioned that the criticality of an asset was measuredin part by loss of life and loss of citizen confidence, and the IA/IP budget justificationalludes to forecasting national security, economic, and public safety implications.

HSPD-7 lists the types of attacks that animate national critical infrastructurepolicy. These are attacks that could: cause catastrophic health effects or masscasualties; impair federal agencies’ ability to perform essential missions; underminethe ability of state and local governments’ to maintain order and provide essentialservices; damage the orderly function of the economy; or undermine the public’smorale or confidence.

One may assume that the IA/IP Directorate considers these factors whendetermining risk. But, are they all considered together? How are differentconsequences integrated into an overall risk rating for a given scenario?43 Does IA/IPweigh each category of consequence equally? HSPD-7 stated that the Secretary ofHomeland Security, when identifying, prioritizing, and coordinating the protectionof critical infrastructures, should emphasis those infrastructures that could beexploited to cause catastrophic health effects or mass casualties comparable to thosefrom the use of a weapon of mass destruction. In this case, might preventing anattack on the Super Bowl take precedent over an attack on one of those core clearingand settlement facilities mentioned above, the destruction of which mightsignificantly disrupt national financial markets? To what extent, if any, is theDirectorate risk averse? The grant application guidelines for the State and UrbanAreas Security grants imply a risk aversion when it focuses on weapons of massdestruction, even though the threat of this might be lower than the threat ofconventional attacks.

Another question is how are these consequences measured? Are potentialdeaths based on experiential data or models or best estimates? How is confidenceor morale, and the impact on morale measured? Are economic models used to


44 The Senate Appropriation Committee, in its FY2005 appropriations bill’s report,recommended continued funding for risk analysis activities that include evaluating second-and third-order cascade effects associated with market interdependencies. 45 See, Office of Homeland Security, The National Strategy for the Physical Protection ofCritical Infrastructures and Key Assets, February 2003, p. 1.

determine economic impact? How are cascading effects due to interdependenciesdetermined? How far down the chain of reactions does IA/IP consider44?

Risk Mitigation

The risk associated with a specific attack on an asset can be reduced by reducingthe level of threat to it, by reducing its vulnerability to that threat, or by reducing theconsequences or impact of an attack should it happen. This parallels the BushAdministration’s overall strategy for homeland security—1) prevent terrorist attacks,2) reduce America’s vulnerability to terrorism, and 3) minimize the damage andrecover from attacks that do occur.45 The Department of Defense, the CentralIntelligence Agency, the Federal Bureau of Investigations, elements of DHS’s Borderand Transportation Directorate, and other law enforcement and intelligence agencies(including the Information Analysis side of the IA/IP Directorate) have the primaryrole of reducing threat, by disrupting, finding, detaining, or eliminating individualsthat threaten the United States. The DHS’s Emergency Preparedness and ResponseDirectorate is principally responsible for trying to mitigate the consequences of anattack, through rapid response and quick recovery. The IA/IP Directorate’s primaryrole is to reduce an asset’s vulnerability. As discussed above, it is doing so mainlyby hardening the asset against attack, by improving the ability of those protecting theasset to deny access to the asset and to improve their ability to repulse an attack. Thisbegs the question, however, of whether or not, and by what mechanism, are thevarious efforts to reduce threat (prevent), vulnerability (protect), and consequences(prepare) coordinated both within DHS and between DHS and other agencies and towhat extent, and by what mechanism, are the allocation of federal resources to thesethree areas influenced at all by comparing the risk reduction achieved by each? Thiswould likely require a level of risk management currently beyond the IA/IPDirectorate’s mandate.

Prioritizing Protection Activities

According to the statement of Director McDonnell, the IA/IP Directorateconducts a risk assessment, mapping threat and vulnerability information. The riskinformation is then used to prioritize the implementation of protective measures. Itis not clear, however, how the risk assessment is used by IA/IP to prioritizesubsequent activity. The implication is that the IA/IP Directorate ranks further its listof 1,700 assets based on the risk. But several question remain. Does the IA/IPDirectorate rank assets as high, medium, and low risk and focus it subsequent effortson those ranked as high? Does it estimate the risk reduction associated with itsrecommendations? Does it seek to maximize the risk reduction of the highest at-riskassets or does it seek to maximize the risk reduction across all of the assets it hasidentified as critical?


46 The National Commission on Terrorist Attacks Upon the United States, The 9/11Commission Report, W.W. Norton and Company, 2004, p. 392.

HSPD-7 gives some guidance in addressing these questions. According to theDirective, the Secretary will “identify, prioritize, and coordinate the protection ofcritical infrastructure and key assets with an emphasis on critical infrastructures andkey assets that could be exploited to cause catastrophic health effects or masscasualties comparable to those from the use of weapons of mass destruction.” Thisimplies that greater weight should be given to loss of life consequences thaneconomic impact, or continuity of government services, or loss of national morale orconfidence.


The IA/IP Directorate has been tasked with a very complex problem. Securityoriented risk management is typically done at the site or facility level or at thecorporate level. The IA/IP Directorate is being asked to do this at the national level,assessing and comparing at least 1,700 disparate sites and facilities it has judged asbeing nationally important.

The IA/IP Directorate is also asked to consider not only economic impacts andloss of life, but the possible impact on national morale and the ability of state andlocal governments to maintain order and deliver essential services. None of these areeasy to measure and all are difficult to trade off one against the other, should theanalysis come down to that. To determine the economic impact of the loss of anasset is more difficult than determining the effect on a company’s bottom line. TheIA/IP is being instructed to determine economic impacts two to three levels throughthe supply chain. It is not clear how the Directorate can or intends to measure theimpact on national morale associated with the loss of an asset, especially a culturalicon. Comparing the potential loss of life in one scenario with the potential loss oflife in another scenario, while sensitive, presents a direct comparison. However,comparing the importance of an asset whose loss may result in a relatively small lossof life with another asset the loss of which might result in a large economic impactis much harder.

The exercise will be less than perfect and probably less than objective. TheBush Administration and Congress are allocating resources in any event, so thesechoices are getting made implicitly. If such processes were more transparent,Congress could better oversee them and offer guidance if necessary.

The 9/11 Commission, in discussing a need for a layered security system forpublic transportation systems, stated that the Transportation Security Administrationshould be able to identify for Congress the array of potential terrorist attacks, thelayers of security in place, and the reliability provided by each layer.46 Expanding onthis, the IA/IP Directorate should be able to tell Congress what criteria it used toselect assets of national importance, the basic strategy it uses to determine whichassets warrant additional protective measures, and by how much these measurescould reduce the risk to the nation. It should also be able to tell how much these


additional measures might cost. Who pays for these measures is another issue,beyond the scope of this report. The national plan called for by HSPD-7 could be agood vehicle for presenting this information. Alternatively, the IA/IP Directoratecould develop a written protocol that outlines specifically the steps taken in the riskassessment and risk management process and the assumptions, criteria, and tradeoffsthat are made. Such a protocol could not only help keep Congress informed, butcould also ensure consistency in carrying out assessments and in making decisions.Some critics of this approach might suggest that each sector and each asset isdifferent and “one size can’t fit all.” However, while each sector and each assetpresents their own unique situations, a common set of consequences that measurerisk and by which to measure risk reduction will act to normalize the analysis acrossassets.

Finally, Congress may choose to offer its guidance to the IA/IP Directorate onsome of these criteria or tradeoffs. To do so with the same systematic approach thatthe IA/IP Directorate has been asked to do, the different committees with jurisdictionover different infrastructures may want to consider coordinating their advice.



Carl Roper, Risk Management for Security Professionals, Butterworth-Heinemann,1999.

U.S. Coast Guard, Implementation of National Maritime Security Initiatives, FederalRegister, Vol. 68, No. 126, July 1, 2003, pp 39240-39250.

American Petroleum Institute and the National Petrochemical & RefinersAssociation, Security Vulnerability Assessment Methodology for the Petroleum andPetrochemical Industries, May2003.

U.S. Department of Energy, Office of Energy Assurance, Vulnerability AssessmentMethodology, Electric Power Infrastructure (Draft), September 30, 2002.

National Communications Systems, Office of the Manager, Public Switched NetworkSecurity Assessment Guidelines, September 2000.

Association of Metropolitan Sewerage Agencies, Protecting WastewaterInfrastructure Assets: Asset Based Vulnerability Checklist for Wastewater Utilities,2002.

Government Accountability Office, Homeland Security: Key Elements of a RiskManagement Approach, GAO-02-150T, October 12, 2001.

American Chemistry Council, the Chlorine Institute, and the Synthetic OrganicChemical Manufacturers Association, in their Site Security Guidelines for the U.S.Chemistry Industry.

Argonne National Laboratory, et al., prepared for the Office of Energy Assurance,U.S. Department of Energy, Energy Infrastructure Vulnerability Survey Checklists,February 22, 2002.

Is this the question you were looking for? Place your Order Here