THE TWO FACES OF POLITICAL INSTABILITY

• Collection: FOIA Collection
• Collection: Declassified Articles from Studies in Intelligence: The IC’s Journal for the Intelligence Professional
• Document Number (FOIA) /ESDN (CREST): 0000620544
• Document Page Count: 9
• Document Release Date: July 30, 2014
• Case Number: F-2013-02322
• Publication Date: April 1, 1987

Approved for Release: 2014/07/29 000620544TITLE:TheTwo FacesofPolitical InstabilityAUTHOR:(b)(3)(c)VOLUME:31ISSUE:SpringYEAR: 1987Approved for Release: 2014/07/29 000620544pproved for Release: 2014/07/29 000620544STUDIES ININTELLIGENCEA collection of articles on the historical, operational, doctrinal, and theoretical aspects of intelligence.All statements of fact, opinion or analysis expressed in Studies in Intelligence are those ofthe authors. They do not necessarily reflect official positions or views of the CentralIntelligence Agency or any other US Government entity, past or present. Nothing in thecontents should be construed as asserting or implying US Government endorsement of anarticle's factual statements and interpretations.Approved for Release: 2014/07/29 000620544Approved for Release: 2014/07/29 000620544(b)(3)(n)iPotential and commotionTHE TWO FACES OF POLITICAL INSTABILITY(b)(3)(c)The analysis of political instability, and the collection of information on it,have become major missions of US intelligence. Instability that catchespolicymakers by surprise is apt to be labeled?perhaps publicly?as an-intelligence failure.- An agreed definition of political instability, however, hasbeen slow to emerge. Nor is there a clear and widely accepted concept ofexactly what the intelligence community should do to prepare the policymakerbetter.These conceptual shortcomings exist partly because political instability isa relatively untraditional intelligence topic in which friends and foes may bedifficult to identify, the arenas of action frequently shift, and the US interestsat stake often are latent rather than overt. It has grown in importance alongwith the growth of the post-World War II, post-colonial Third World. In theUnited States, it has received additional attention since the searing experienceof the Iranian revolution of 1979 and the subsequent hostage crisis.'The conceptual difficulties also reflect some general problems?chieflyconcerning how analysts set priorities, cope with uncertainty, and communi-cate that uncertainty to the consumer?that apply not only to instability butalso to other intelligence topics. This essay discusses several of those problemsand how they help to make political instability a difficult subject. But first,exactly what is the subject?Instability as PotentialInstability is a clear concept when used in a mechanical or physical sense.An overloaded vehicle is unstable if a small disturbance (such as hitting a bumpin the road) has the potential to trigger a sudden, major change (the vehicleoverturns). The term is often used in the same way to describe certain military,economic, and diplomatic situations. The nuclear balance between the super-powers, for example, is usually described as stable or unstable depending onwhether one believes that a single mistake such as an accidental nucleardetonation (i.e., a small disturbance) is apt to touch off a nuclear war (a sudden,major change).Similarly, political instability is the potential for?or probability of ?sudden, major change in a country's politics. The small disturbance that triggersthe change may be an assassin's bullet, the emergence of a charismaticopposition leader, the luck of a coup-plotting cabal, or an additional economichardship that becomes the last straw for a disaffected populace. The majorchange may be the overthrow of a regime by coup or revolution, but it mightalso be a sharp policy reversal by a ruler who stays in power.George Caldwell, ''The Mob Is in the Streets,- Studies in Intelligence, Winter 1985, 33-40.SECRET(b)(3)(n)Approved for Release: 2014/07/29 0006205445Approved for Release: 2014/07/29 000620544(b)(3)(n)InstabilityThe analysis of political instability, insofar as it involves the potential forchange, is not primarily a matter of predicting individual events. If intelligencecan provide warning of a coup or other event that may lead to bigger politicalchange' s, so much the better. But in most instances, US intelligence lacks suchspecific information, or the sources that can provide it. The great, vexingQuestions for most political analysts?particularly those who monitor ThirdWorld countries?hinge on events that are unknowable for most practicalpurposes: whether the gunman recruited to assassinate a head of state will shootstraight; whether the next military officer to join a coup plot will be the one tobetray the plan; whether it will rain hard enough to douse a riot on the day agovernment raises food prices. It is theoretically possible to predict such specificoutcomes, just as it is theoretically possible to predict the outcome of a roll ofdice (by noting the position of the dots when the player picks up the dice andobserving how he shakes them, the angle of his hand when he throws them, etc.)but it usually is futile to try. A more worthwhile use of our inherently limitedcapacity to gather and process information is to understand the characteristicsof the system that will generate the outcome (Are those dice crooked? Is thatgovernment shaky?) and to estimate the probabilities of different possibleresults.Intelligence analysts who cover other kinds of accounts also occasionallymust worry about potential big events depending on unpredictable small ones.Military conflicts, for example, sometimes are unstable in that a small turn ofevents can shift the momentum of a war (just as loss of the proverbial nailmeant the loss of shoe, horse, rider, and battle). But political analysts who coverunstable Third World countries probably face this problem more often thanany of their colleagues. On the central questions of who will rule these countriesand what the rulers' policies will be, the political analyst may find it easy to begenerally right most of the time (by predicting continuity, because on most daysmost governments don't fall or overhaul their policies). He or she is morevulnerable, however, to the intelligence -failure- of being way off the markwhen a sudden, major change does occur.Actually, not predicting such an event should not be regarded as a-failure,- as long as the mission of intelligence in such situations is understoodto be the assessment of instability, and instability is defined as potential. By thisview, the track record of an analyst or an agency in covering instability can bejudged not by a few specific predictions but only over a long period, by lookingat how estimates of instability generally compare with the frequency and typeof subsequent political change.To say this is not to dodge responsibility or to rationalize mistakes.Policymakers frequently have to cope with uncertainty, on domestic as well asforeign issues, by hedging their bets. In the case of political instability in foreigncountries, bets tend to be difficult to hedge because often the possible scenariosare drastically different from each other and the ways of coping with them areincompatible. A close relationship with an incumbent regime, for example,may lessen the chance for good relations with an opposition group, and thus forUS influence in the country should that-group gain power. But this problemreflects the nature of the subject matter, not a failure of intelligence.Approved for Release: 2014/07/29 000620544(b)(3)(n)Approved for Release: 2014/07/29 000620544Instability(b)(3)(n)Indeed, intelligence ill serves the policymaker by offering specific predic-tions in the face of uncertainty. For one thing, this conveys a false sense ofconfidence and fails to alert the policymaker to the need to hedge bets, or atleast to try to hedge them. For another, it obscures what may be importantjudgments about changing probabilities of potential events. If the chance thata government will fall this year has gone from 10 to 30 percent, that is animportant development, but the best specific prediction would still be that itwill not fall. Finally, ungrounded specific predictions?some of which inevi-tably will be wrong?reduce confidence in, and reliance on, intelligence overthe long term.Instability as CommotionThe concept of potential for change is clear enough, but users of the term"political instability" often seem to have something else in mind. Intelligencepublications contain numerous references to the "potential for instability" in agiven country or the "prospects" for a government to be stable or unstable. Ifinstability is itself a type of potential, such phrases are redundant if notnonsensical. They raise more than a semantic or grammatical point, however;they point t6 a different concept of political instability.That concept includes the idea of possible major political change (andmore specifically, irregular or extra-legal change) but also includes actualevents such as demonstrations, strikes, riots, issuance of manifestoes, militarystates of alert, the emergence of hitherto underground opposition groups, andincreasingly vocal demands for a government to quit. What these events havein common is that they are overt, they get attention, and they seem to indicate,correctly or not, that the regime's grasp on power is weakening.This kind of commotion is often an important subject for US intelligence,for several reasons. If nothing else, civil unrest affects the climate for business,including US-owned business. Moreover, even if a regime stays in power, itsresponse to unrest may entail costs to itself and affect its image abroad and itsrelationship with the United States (e.g., when a crackdown violates humanrights). Most important, these kinds of disturbances sometimes really areprecursors to coups, revolutions, or other types of major political change, eitherbecause they reflect broader discontent or because they galvanize oppositiongroups into action.Instability-as-commotion is probably closer to most people's notion ofpolitical instability than is the concept of potential for change. For intelligenceto focus narrowly on it, however, entails drawbacks. By itself, political unrestis only an imperfect indicator of the probability of major political change. Thelevel of -normal- unrest is an aspect of political culture that varies greatly fromone country to another, even within the Third World. It also depends on whatthe regime does to contain it. Repression by a justifiably fearful governmentmay keep unrest low in a country where the chance of major change is high(e.g., South Africa until the disturbances in recent years, or possibly Iraq today).Conversely, a resilient democracy can also afford to be relatively rambunctious(e.g., Israel, or any of a number of Western developed countries).Overt unrest is especially unreliable as an indicator of major politicalchanges that do not entail the fall of a regime. Regimes sometimes revamp their(b)(3)(n)Approved for Release: 2014/07/29 0006205447Approved for Release: 2014/07/29 000620544Instabilitypolicies precisely to avoid such unrest?and to stay in power. For example, theIranian revolution has been the most dramatic consequence of the past decade'sresurgence of Islamic fundamentalism, but otherwise the main political impactof this resurgence has been the preemptive incorporation of fundamentalisttenets into official policy (as in Pakistan). The impact on how the governmentof an Islamic country does business with the United States may be significant,even if it is the same government. Where coups and military rule have been thenorm (as in much of Latin America), the greatest potential for major change(and thus the biggest political story to which intelligence can alert thepolicymaker) occurs where unrest sufficient to bring back the generals does notbreak out and a civilian government takes power constitutionally.Two Ways to Serve the PolicymakerWhich type of instability should intelligence officers monitor, study, andwrite about? The short, and generally correct, answer is -both.- Resources foranalysis and collection are limited, however. Moreover, the number ofjudgments,that intelligence can present to policymakers, highlight as beingimportant, and expect policymakers to absorb is also limited. A focus on unrestand other overt behavior thus might mean inadequate attention given to other,less visible but still important, developments.This raises the larger question of how much emphasis intelligence shouldgive to topics already on the policymaker's agenda, versus forcing the agendaby emphasizing other issues that analysts believe may ultimately have greaterimpact on US interests. Instability-as-commotion, when it breaks out in an evenmoderately important pro-US country, will definitely be on the policymaker'sagenda. It will probably receive attention in the press and in Congress.Consumers of intelligence in the executive branch will necessarily give itattention as well. For intelligence not to have forecast the unrest that causessuch attention would commonly be regarded as an intelligence -failure.- Incontrast, important changes occurring underneath the surface of a country'spolitical and social fabric will not automatically get on the policymakingagenda, simply because they are less visible and the policymaker is not forcedto react to them.To cover both types of instability adequately, intelligence officers need tomake a more conscious effort to study the potential for change than they do toreport on commotion and unrest. Because instability-as-commotion will be onthe policymaker's agenda, it will necessarily be on the intelligence community'slist of tasks as well. Instability-as-potential, in contrast, needs special care andfeeding as an intelligence topic. This is not necessarily a matter of tellingpolicymakers what they don't want to hear. Commotion in foreign countriesoften is bad news for US policy. Conversely, a potential for change may be goodnews (in a country currently unfriendly to the US). Even when the main newsis not good, as long as the issue is the potential for change there may still be anopportunity for US policymakers?guided by US intelligence?to do somethingabout it, such as by shoring up a friendly regime or by cultivating oppositionmovements in the hope that a new regime will be friendly as well.(b)(3)(n)Approved for Release: 2014/07/29 000620544Approved for Release: 2014/07/29 000620544Instability SECRET(b)(3)(n)Challenges in Studying InstabilityIn addition to conceptual confusion, there are several other challenges incovering political instability. One is the extraordinary range of expertiserequired to make sound judgments of instability in a foreign country. It isdifficult enough to assemble the right mixture of backgrounds and specialties ina group of analysts to monitor instability in country X; it is even more difficultto find such breadth of knowledge in an individual analyst (and judgments areultimately made by individuals, albeit subject to -review by other individuals).The study of instability is profoundly interdisciplinary. The judgmentsinvolved concern political developments, but economic, military, and sociolog-ical factors typically are critical ingredients. The most destabilizing factor inmany Third World countries today is economic scarcity, together with thequandary of how to put an economy back on track without alienating thepopulace through harsh austerity measures. In many of these same countries,the most likely alternative rulers are in the military, making an understandingof military morale, and the effect on it of war or insurgency, another key toassessing instability.Judgm'ents about political instability also require both familiarity with aspecific country and broader knowledge about patterns of political changeacross many countries. On one hand, the assessment of instability is partly anintuitive process in which there is no substitute for an intimate, longstanding,seat-of-the-pants feel for country X and its political culture. On the other hand,certain patterns of loss of legitimacy, rise of opposition movements, andrevolutionary political change have recurred through history in differentcountries. Social scientists have studied these patterns and generalized aboutthem. The generalizations are imperfect, but they contain some useful insightsand lessons. The trick for the analyst is to exploit the lessons while rememberingthat each case will be different, probably in some important ways, from anyprevious one.A further difficulty is how to deal with uncertainty, and specificallyuncertainty about possible events?namely, the overthrow of governments?that are unlikely but potentially very significant for US interests. In some cases,intelligence officers may hesitate to -make the call- that a regime is likely tofall, because the fall of a regime is a rare event, or at least rarer than thecontinuation of a regime in office for another day. Probably in more cases,however, there will be a bias in the opposite direction (i.e., toward crying-wolf"), because of the intelligence officer's awareness that not to predict arevolutionary change that does occur would be regarded as a bigger intelligence-failure- than to warn repeatedly of one that does not. Rotation of intelligenceofficers from one assignment to another encourages this latter bias; none ofthem wants to be accused of being caught by surprise as country X is -lost- ontheir watch; they are less likely to be concerned about the long-term effect onan intelligence agency's credibility if it cries -wolf- too often.In addition to these analytical biases, there is the problem of communi-cating uncertainty to the policymaker and of choosing the right message toemphasize when describing an uncertain, unstable situation. Does one need tostress that the government of Ruritania, although it will probably retain power, (b)(3)(n)Approved for Release: 2014/07/29 0006205449sccno-Approved for Release: 2014/07/29 000620544(b)(3)(n)Instabilityhas a significant chance of falling? Or does the policymaker need instead to bereminded that a continuation of the current Ruritanian government is still themost likely scenario? To answer such questions, intelligence analysts need tostudy their consumers?and their biases and blind spots?as closely as theystudy the countries they are responsible for monitoring.An additional challenge in covering political instability is deciding howmuch coverage to give to individual countries. How collection and analyticalresources should be distributed depends not only on the current importance ofeach country but also on how important each one might become if it underwenta major political political change. This is a difficult calculation. A nation's size,resources, and location determine its importance to some degree, regardless ofwho governs it. Qadhafi, for example, has been able to make so much troubleand get so much attention partly because of Libya's oil, which was known evenbefore Qadhafi came along. But what about, say, Castro's Cuba? It lacks acomparable resource, and even its location close to US shores has not?since themissile crisis of 1962?been the principal reason it has commanded so much ofthe US government's attention. Cuba's intrinsic, permanent characteristics, inother words; would have been insufficient grounds?given what was known inthe 1950s?to divert a large amount of resources from other topics to study thestability of the Batista regime.Cuba illustrates how the dynamics of a revolution (and a revolutionaryregime's internal political needs that are served by, for example, sending troopsto fight in guerilla wars overseas) can make a country an important intelligencetarget. It also demonstrates how the importance to US policymakers of politicalchange depends heavily on what kind of relationship a new regime strikes upwith the USSR. This, in turn, is apt to depend less on the permanentcharacteristics of the given country than on such harder-to-calculate factors asthe prejudices and cultural affinities of the new regime's leadership, thenumber of other revolutionary opportunities that open up to the Sovietselsewhere, and how the Soviets apportion their resources as they try to bolsteror nurture clients in return for military access rights or other concessions.US intelligence's coverage of political instability must incorporate a globalview, both because the problem is, to a degree, part of a worldwide US-Sovietgame and because different analysts have different ways of expressing uncer-tainty. The Political Instability Quarterly, which CIA's Directorate of Intel-ligence initiated in 1983, helps to provide such a global view. As a compendiumof separate judgments by country analysts, however, it is designed more to tracechanges in individual countries than to make cross-national comparisons. Amore direct effort to make such comparisons and to measure instability in alarge number of countries with a single yardstick has been NIE 7-81 2 and itssuccessor Estimates. Such worldwide, comparative assessments are audacious inscope and difficult to produce, but they probably are the best basis yet forestablishing priorities and distributing resources for monitoring instability.2 Political Instability and Regional Tensions, 14 September 1981.10  (b)(3)(n)Approved for Release: 2014/07/29 000620544Approved for Release: 2014/07/29 000620544Instability SECRET(b)(3)(n)Doing Even BetterThe intelligence community, and CIA in particular, have made substantialprogress over the last several years in studying political instability. 3 The topic'srelatively recent recogition as a distinct and important subset of intelligenceproblems (even though it has actually been an intelligence problem, underdifferent labels, for much longer) has a lot to do with this. In striving for stillmore progress, the following points bear emphasis.First, both kinds of relevant expertise?detailed knowledge of a singlecountry and broad understanding of cross-national patterns of politicalchange?need nurturing. Just as important, the two types of experts need tocommunicate with each other well enough to exploit fully each other's insights.Second, analysts and collectors need to look at a wide variety of indicatorsof potential change, including ones that have nothing to do with instability-as-commotion. The Political Instability Quarterly has already done a majorservice in this regard.Third, intelligence officers need to be precise as to which type of instabilitythey are talking about. The policymaker is apt to be interested in bothcommotion and the potential for change. To warn of one type of instabilitycarries a different message, however, with different policy implications, than towarn of the other type.Finally, intelligence officers need to be frank about their inability toprophesy the specific events that will trigger major political changes. Consum-ers of intelligence have always preferred certainty to uncertainty and alwayswill. But intelligence is not crystal-ball-gazing, and consumers should bereminded often of how what we don't know can hurt us.This article is classified SECRET(b)(3)(n)(b)(3)(c)(b)(3)(n)Approved for Release: 2014/07/29 00062054411