Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Appendix B Risk and Risk Management In this report the committee has adopted the conceptual foundation and nomenclature of the national- security community and defined risk in terms of a threat or hazard, the bad consequences that can arise from that condition, and the probability and severity of those consequences. Specifically, the committee has worked with definitions from national-security doctrine that are broadly applicable with minor adaptations.32,33 These definitions include the following: â¢ Risk: probability and severity of loss linked to hazards. â¢ Hazard: a condition with the potential to cause injury, illness, or death of personnel; damage to or loss of equipment or property; or mission degradation. â¢ Probability: the likelihood that an event will occur in an exposure interval (time, area, etc.), ranging from frequent or inevitable to unlikely or improbable. â¢ Severity: the expected consequences of an event in terms of injury, property damage, or other mission impairing factors, ranging from catastrophic to negligible. The committee has drawn from this doctrine, but paraphrased and adapted it for the reportâs purposes. This doctrine dates back to the 1980s239 and largely parallels the guidance of other federal agencies, including those with civilian and strategic concerns, for example, the Department of Homeland Security (DHS)240 and the Department of Transportation.241 DHS defines risk similarly as the âpotential for an unwanted outcome resulting from an incident, event, or occurrence, as determined by its likelihood and the associated consequences.â242 The committee chose to work with the national-security doctrine because of its maturity and utility. It pinpoints two concrete facets of riskâprobability and severityâ and ties them to a set of practical risk-management tools that can support qualitative analysis. Those tools are not without criticism, which the committee addresses in a later note in this appendix.243,244 For a review of the strengths and weaknesses of DHSâs approach to risk analysis, see the cited literature.245 Two aspects of the committeeâs use of this vocabulary merit attention. First, whereas the doctrine focuses on personnel and mission, the committee has adopted a broader lens that encompasses individuals, public and private institutions, and the environment. Second, the Department of the Army (DA) acknowledges the potential for losses from hazards (accidental), threats (tactical), acts of terrorism, suicide, homicide, illness, or substance abuse, but it refers to hazards holistically;32 for consistency, the committee also uses the term hazards in this appendix, but it highlights specific PREPUBLICATION COPY: UNCORRECTED PROOFS 127
128 Restricting Access to Explosive Precursor Chemicals concerns about potential losses from acts of terrorism. In the main text of the report, the committee also refers to threats, but does so broadly, to encompass a wide range of tactical and other concerns, including those of terrorism. Whether a hazard or a threat, the condition under consideration can lead to injury, illness, death, or damage, with a particular probability and severity. Generally speaking, a hazard can be viewed as a condition that might give rise to an event that can be more or less likely (probability) and entail more or less damage (severity). For example, a family might store an ignitable material, such as turpentine, near a gas heater in a basement. This storage arrangement poses a hazard that might lead eventually to a fire, entailing physical injury and property destruction. The odds of the fire could be high or low and the amount of damage could be marginal or catastrophic, depending on whether the basement is cluttered or family members are home. A public safety officer might be interested in helping families to reduce the probability of fire, by addressing the conditions under which the fire can occur; in this report, the committee is interested in identifying conditions under which terrorists can obtain chemicals to produce IEDs and means of reducing the odds that precursor chemicals fall into the wrong hands as fodder for a terrorist episode, which is the ultimate event of concern. To that end, one might consider whether a chemical can be or has been used in an IED and whether and how a terrorist might lay hands on it. Alternatively, the problem could be framed in terms of episode risk (RE) and precursor risk (RP). The former is a function of the probability (PE) and severity (SE) of the episode (Equation B-1), while the latter is a function of the probability of observing the use of the precursor (PP) given concern for the episode and its associated risk (Equation B-2). PP will be dependent on constituent factors, such as the ease of extracting a precursor chemical from a product or formulation and the attendant personal dangers of handling it. = , (B-1) = | , (B-2) DA juxtaposes probability and severity in a risk matrix that can be used to evaluate the risk level of a particular hazard and inform, if not establish, mitigation priorities; the doctrine addresses both severity and probability to distinguish the one hundred-year storm, the tsunami, and seasonal flooding. Therefore, the possibility of a moderately harmful but likely event might generate as much concern as the possibility of a catastrophic but unlikely event. Quantitative data, for example, rates of occurrence, illness, injury, death, or damage, should be used to support the assessment, if available, but are not absolutely necessary. In the committeeâs simplified rendering of the standard matrix (Figure B-1), risk levels reside in four quadrants, range from high (red) to low (yellow) and depend equally on probability and severity, but other more-sophisticated formulations are possible. The version used by DA is asymmetric, provides more categories for probability and severity and, in some instances, gives slightly more weight to severity than probability.32,33 Non-linear permutations are also possible.237 PREPUBLICATION COPY: UNCORRECTED PROOFS
Appe endix B 129 F FIGURE B- Simplified risk matrix H: high, M: medium; L low. -1 d x; M L: The matr presents an initial bas for consid rix a sis dering, if no establishin priorities across scen ot ng, narios, c chemicals, an policy res nd sponses and can be used in at least th ways. O could st a risk analysis d hree One tart f from the end of a narrativ with a set of scenario d ve t osâthat is, tyypes of terro orist episode of particul es lar c concern becaause of their riskinessâa back out a set of che and t emicals of in nterest; one c could start th he a analysis from the beginn m ning with a se of chemic and track them forw et cals k ward to scenaarios, by aski what ing r risks each pr resents under what condi r itions, with what likeliho and seve w ood erity; or one could appro oach the a analysis from both direct m tions and iterate toward convergence e. Prelimina to any su approach the matrix can be used to establish whether an episode is of ary uch h, x d h n p particular concern, depen nding on wh hether it is su ufficiently ris sky. For exaample, a scen nario involvi a ing l larger-scale explosive de e evice, such as a truck bom could en a mb, ntail substan ntially more damage than an s scenario invo olving a sma aller-scale, person-borne device, suc h as a back-pack bomb, but could be far less p e e l likely; thus, the risk of ei t ither scenari might rate concern. If taking the s io e f scenario-by-s scenario appproach, o could ide one entify the ch hemicals that terrorists ca use to pro t an oduce each ttype of devic and the co ce onditions ( (e.g., relating to availability and othe factors) un g er nder which t they can obta them; de ain evelop strategies to r reduce the od of terror dds rists getting access to the chemicals; and, ultimat a e tely, lessen t risk of ei the ither s scenario. By limiting the analysis to a set of scen e narios, one c focus atte can ention on a ssingle dimennsion of p precursor risk, namely, th probabilit of observing the use o the precur he ty of rsor, given cconcern for thhe e episode (Eq. B-2). Figure B-2 redraw the matrix as a contin e ws x nuum withou strict rankings in whic the ut ch r risks of vehic and pers cle- son-borne ep pisodes are at similar lev and pote a vels entially reduucible. PREPUBLIC P CATION CO OPY: UNCOR RRECTED P PROOFS
130 Restrictin Access to Explosive P ing o Precursor C Chemicals F FIGURE B- Simplified risk matrix as a continu -2 d x uum with tw scenarios; P: episode involving pe wo ; erson- b borne device V: episode involving vehicle-born device. e; e v ne However the matrix, even with elaboration, cannot prov definitiv guidance for establish r, e vide ve hing p priorities, in part, becaus it is silent on the feasibility and co of mitig se osts gation option If relying only on ns. t matrix, a policy-mak might ina the ker advertently iggnore low-hhanging fruit or suggest d dedicating in nordinate r resources to the impossib a low ris might hav an easy an cheap fix whereas a m ble: sk ve nd x medium risk might k h have no feasible or affordable remed dy. Research offer a number of cr hers n riticisms of qualitative to and met q ools thods of anal lysis, includ ding risk m matrices. Wh acknowl hile ledging the popularity of matrices, th warn ag p f hey gainst a lack of scientific c v validation an the drawb nd backs of usin them indis ng scriminately 243,244 For e y. example, cor rrelations bettween s severity and probability can invert re c esults. Lundbberg and Wi illis offer rem marks on the imperfectio of e on v various analyytical option 246 âWhile some articles have expl ns, e lored the reaasons to be ccautious abou the ut u of qualita use ative risk ass sessment too there are also reasons to be cautio about th use of qua ols s ous he antitative e estimatesâ; th latter poin is further explored els he nt sewhere.245 The decision to start or stop mitiggating risk also requires consideratio of risk tolerance and the a on a acceptability of residual risk. The for y rmer refers to the level o risk below which a ha t of w azard does no ot w warrant any expenditure of resources on mitigati and the l s ion latter refers t the amoun of risk tha to nt at r remains after mitigation. If the initial level of risk is low eno r ough to be toolerable abse any mitig ent gation, t then the resid risk would equal the initial risk. By implica dual ation, risk m mitigation nee not amoun to risk ed nt elimination. In some inst e tances, it mig be prefer ght rable to acce some am ept mount of riskââinitial or residualâan develop a response an recovery plan.247 Such preferences typically a r nd nd p h arise when th costs he o addressing the risk ou of g utweigh the anticipated benefit. a b PREPUBLIC P CATION CO OPY: UNCOR RRECTED P PROOFS
Appe endix B 131 The matr is also sil on causa rix lent ality, but the cause or ca e auses of a hazzard can bea directly on policy ar n m making. Abs a thorou understan sent ugh nding of the underlying reasons for a hazard, a p policy maker might r c choose a risk mitigation strategy that doesnât imp k t prove condittions or mak them wor 247,248 Th same kes rse. he c be said of any attemp to control risk, including attempts to control r associate with prec can o pt s risk ed cursor c chemicals. Fortunate the risk matrix is jus one tool fo managing risk, and na ely, st or g ational-secur doctrine nests it rity in a continuo five-step risk-manag ous p gement proce that throu its cycle (Figure B-3 addresses these ess ugh e 3), s a other con and ncerns, potenntially inclu uding those pertaining to operational dynamics an behavioral p nd adaptations. A particular hazard might present a low-level risk in one mo a r oment and a high-level r in risk a another and, even if polic makers have a firm grasp on caus cy g sality, efforts to control r can enta risk ail unintended consequences that give ri to new an different risks. Takin action in o arena mi u c ise nd ng one ight r result in disp placement an merely pu the risk elsewhere. T five-step risk management proce nd ush e The p ess a allowsâand even requir d resâdeeper, ongoing con nsideration o risks, con of onsequences and ntrols, and co p provides exp plicit means to incorpora new infor t ate rmation and address chan nges in circu umstances. WWhereas t matrix is inherently static, the fiv the s ve-step proce which fo ess, orms an unen nding loop, is inherently y d dynamic. Th meets the fourth princ his ciple of risk managemen âapply ris manageme cyclicall and nt, sk ent ly continuously 32,33 c yâ. F FIGURE B- Five-step risk-manage -3 ement proces Recent n ational-secu ss. urity doctrine includes a much- e s gure.247,248 simplified fig In Figure B-3, steps 1 and 2âwh include hazard iden e hich ntification an address ca nd ausality, probability, a severityâ and âconstitute risk assessm ation.247 Fro DA:32 ment, and ste 3 through 5 constitut risk mitiga eps h te om Steps 3 through 5 are the ess s sential followw-through a actions to m manage risk eeffectively. I In these steps, leade balance risk against costs and t ers t take approprriate actions to eliminat s te unneccessary risk and accept residual ri at the a k t isk appropriate level. Durin execution ng n, leade continuou ers usly assess the risk to the overall m t t mission and to those inv volved in thhe task. Finally, lea aders and inddividuals evvaluate the e effectiveness of controls and provid s s de lesson learned so that [they and] others may benefit from the ex ns o a m xperience. PREPUBLIC P CATION CO OPY: UNCOR RRECTED P PROOFS
132 Restricting Access to Explosive Precursor Chemicals Steps 4 and 5 generate experience and feedback, which can either validate the current approach or suggest the need for a change of course or a procedural refinement, whether applying the process in a military or civilian setting. PREPUBLICATION COPY: UNCORRECTED PROOFS