lntrmsriond Journal of Forecasting 4 (1988) X1-141

%rth-Holland

OPERATIONS

Ahtrflct: Political risk forecasting has attracted considerable business and academic attention in the last decade. This paper starts with a review of the major findings and approaches to the assessment of political risk in foreign investment situations, particuI~rly in terms of the sptyification of relevant causctl relationships between the sources of risk and corpo- rate contingencies. Next. a comprehensive model of general applicability incorporating these findings is outlined. Finally. we conclude with :1 few comments on the dynamics of political change and corporate strategy that make any such model subject to constant review and evaluation.

In 1985. the world’s stock of foreign direct investment (dcfincd as those firms where 258 or more of the equity was in foreign hands) approached a value of $900 billion. If other assets owned abroad by non-financial corporations, such as bank deposits, securities, inventories and minority {less than 25%) interests, were added to this figure the total may very well exceed $1.500 billion. In addition, world trade in goods and services amounted to more than $2,000 billion in 1985. Given that a significant proportion of these assets are exposed to expropriation. war, terrorism or discriminant government inten.ention at any point in time. the implications for the management of global operations are rather sobering.

Surely there is nothing new in this. Trade and investment have been exposed to political risks ever

since the first caravans ventured across the Middle East several millennia ago. It is rather the magnitude of the exposure, and the much publicized losses associated with the nationalizations which followed regime changes in post-colonial Africa. Cuba. Iran and Nicaragua, that have thrust the issue of politicai risks to the forefront of business and academic concern.

Political change and instability per SC do not necessarily affect the international investor; change

may be abrupt and yet leave the fabric of society basically untouched, or gradual but with profound effects. Instead, what must concern the international investor is the impact that any environmental shock, whether the result of a violent change in political regime or of a gradual process of social and

l The authors wish to express their appreciation to Ms. Dororhea Bcnscn for her du;lhlc assistance in the prepnration of this

paper. WC arc also grateful to Farihon Ghadar. Skphrn Kobrin tnd Jeffrey Simon for thar critical commcnrs and

suggestions. while wc exonerate them from any responsibility for the final product.

016%Z!O70/88,43.50 ‘3 1988. Elsewcr Science Puhlishcrs B.V. (North-Holland)

political evolution is immaterial, may have on the value of its operations in that country. For this we can distinguish two generally different contingency losses. The first we may define as the involuntaq loss of control (generally meaning property rights) over specific assets located in a foreign country. typically without adequate compensation. Expropriation. nationalization. the ravages of civil war. or wanton destruction by terrorists are all examples of such losses. In fact. much of the literature and analysis on political risks has focused on specific instances of this type of loss, and particulxly on

the frequency and extent of expropriations. ’ A second. perhaps more important. yet less well understood contingency concerns a reduction in

the expected value of the returns from a foreign-controlled affiliate due to discriminatory actions taken against it. either because of its foreign nature or as part and parcel of ;I general tightening on free-market prerogatives. Some authors [e.g.. Kobrin (1983)] argue that the frequency of espropria- tion has declined dramaticrtlly since 1976, replaced almost universally by increased regulatory controls over investment behavior. Included here are various forms of discriminatory controls and restrictions often imposed by governments in times of domestic crisis [including foreign exchange and remittance restrictions]. limitations on access to factor markets [financial. labor or raw materials] and on outputs [e.g.. on prices or diversification possibilities]. ;1s kvell as changing rules on domestic value added. taxation or export performance requirements.

Another dimension of the exposure to political risks concerns the proximate c;1usc. One possibility is to distinguish between the actions undertaken by Icgitimate [if not necessarily representative or democratically elected] govcrnrnents in the cxtcrcisc of their national prcrogativcs. and those which arc the result of actions undertaken by actors outside the direct control of governmental ;luthoritics. ’ Fig. 1 summ;~rizcs the first level of classification. Altsrnativcly, one can dcfinc the horizontal dinicnsion of the matrix in terms of the indiscriminant or sclcctivc nature of the xtions. This diffcroncc bctwcon ‘macro‘ risks, that is. those which affect all foreign [and. for that matter. many national] corporations in the country in question. and ‘micro’ ri&s. dcfincd xi those which target ;i single or ;I few companies for intcrvcntion on the bais of specific argurncnts about their rcl:itivc contributions to n;itional wclfarc. is one that has ban ma& repcatsdly in the literature [e.g., Robock (1971). Kobrin (1979 and 1981). and Simon (19X2)].

Consider the four types of risks illustrated in fig. 2. Type A consists of the potential for massive expropriations of foreign properties typically related to drastic changes in government, such as those following decolonializution or the triumph of ;I IMarxist revolutionary rebel force. There are often associated with closed systems in developing countries. which e.xplode after a long period of national frustration or discontent. followed by growing repression and a collapse of authority. All foreign investors are subject to the same treatment with rare exceptions as they are closely identified with the

previous regime and offer a ready source for the re-birth of national pride. Type B risks. on the other hand, can occur more gradually and in both industrial or developin, 0 economies. They consist of sector-specific or company-specific exposure which can be traced to the particular characteristics of

the industry in question [e.g.. its degree of technological sophistication or its oligopolistic nature] as

There i\ no dcny~ng (hc ~mport;lnce of (hex Iosscs. AS summarized hy Burton and Inouc (19X4). [h c’ cxpropnar~un nsk ha.5

been variody estimated 10: affccr 1% per year of the numhcr of foreign affihdtcs xrire in ICM dwclopd coun~nes [LI>Cs[

durmg the 1960-77 prnd; rcprcxcn~ cumulatlvcly ovrr the lY56-71 pcrwd ;1 ~OIA of 1X.X% of the SIW~ of forclgn d~rccl

invcs[mcnt in 1Y72 plus the vduc of the cxpropriatcd aswts [William (IY75)[: amount IO 1.6% of rhc 1o1;11 vduc of U.S.

inveslmcn(s in LDCs during 1960-74 [Hufbaucr ad Brigs (lY75)]: ad xcoun~ for ;! cumulative [lY60-761 4 4% of the

1976 smck of wholly and p;trGdty owned firms in LDCs plus the vdw of rhc wrcd AUCIS [Kohnn (IYXO)[.

Simon (19X2) makes a slmlls distinction hc~wccn ‘soae131 forces [i.e.. [box that cmcryc from gcnurai w&l phcnomcna[

;md govcmmcnl-lnspirrd axions. 1~ is also imporranl IO make ;L distinction hclwccn sudden changes in govcmmcnlr.

govcrnmcnl pohcics. or entcrnally induced CVCIIIS. and ;1 grdud evotuuon along a more or less prcdicrahlc xx-x-pollrical

pactcm. For more on this xc Kohnn (1979).

Contingencies may incllude:

The Lnvoluntary loss of control over specific assets without adequate compensation

A reduction in the value of a stream of benefits expected from the foreign- controlled affiliate

Loss may be the result of:

The actions of legitimate government a-thoritles

Events caused by actors outside the control of government

-Total or partial expropriation

-Forced drvestiture -Confiscation -Cancellation or unfair calling of performance bonds

-War

-Revolution

-Terrorism

-Strikes

-Extorslo"

-Non applicability of -Nationalistic "national treatment- buyers or suppliers

-RestrIction in access -Threats and to financial, labour disruption to or material markets operations by -Controls on prices, hostile groups outputs or activities -Externally induced -Currency 6 remittance financial constraints restrictions -Externally imposed

-Value-added and limits on imports export performance or exports requirements

Loss continq_ncLcs

An involuntary loss of control over specrflc assets without odcquatc compensation

Heduction in the expected value of the benefits to be derived from the forelqn offlliate

Type A:

ML1.SRLVC expropriations

Type B:

Solectlvc nationalizations

Type C:

General deterioration of the investment climate

Type D:

Restrictions tarqeted to key sectors

Yacro risks: Micro risks:

Sudden convulsive chances that threaten most of :hc population of foreiqn direct investors within the country.

Interventions generally motivated by specific consideration closely related to the economic and social conditions prevailing at the time, and to specific industry and firm characteristics.

uell as to the current situation in the host country with regards to the saliency of the investment relative to national priorities and objectives. Thus. drastic changes in governments or political retimes mav or mav not result in massive expropriations. just as the gradual evolution of national goals and industry characteristics may bring certain sectors of economic activity under sudden scrutiny by non-radical government authorities eager to bring key economic sectors under national control for what appear to be legitimate objectives. ’

On a less dramatic level. a change in government orientation. either in the course of a freely-held election or in response to external threats. can result in increases in taxation. new requirements in terms of domestic equity participation. lower remittance allowances, etc.. all or any or which can substantially alter the post-tax. home-currency present value of the benefits realized by the firm’s foreign affiliate [Type C risks]. Similarly. Type D risks entail considerable potential for loss. As LDC governments strive for a greater share of the fruits derived from foreign company operations in their territory, case-by-case analysis and lengthy negotiation on domestic value added. export perfor- mance, ownership limitations and the like will be more and more the rule than the exception. Of course. the evaluation of the potential benefits of the investment and the split of the spoils will be greatly dependent on the sector and the company involved. and not only on what the country has to offer. as has been argued elsewhere [de la Torre (19X1)]. It is this latter category that we believe may be most prevalent in the future.

Political risk can then be defined as ‘the probability distribution that an actual or opportunity loss will occur due to the exposure of foreign affiliates to a set of contingencies that range from the total scizurc of corporate assets without compensation to the unprovoked intcrfcrencc of cttcrnal agents. with or without governmental sanction. with tho normal opcratinns and pcrformancc expcctcd from the affiliate’. Whcthcr the loss is caused by Icgitimatc govcrnmcnt acts or not. whcthcr it is the result of forces acting internally within the host country or cmanatin g from the home or global cnviron- mcnt, and whcthcr all foreign companies arc equally affcctcd or not. are important rncthociologic~II questions. The next section reviews some of the major findings and approaches to the assessment of political risk, particularly in terms of the specification of rclcvant causal relationships between the sources of risk and corporate contingencies. Ncnt, a comprchcnsivc model of gcncral applicability which incorporates these findings is outlined. Finally. we conclude with a few comments on the dynamics of political change and corporate strategy that make any such model subject to constant

review and evaluation.

2. Empirical evidence and practice

Empirical data on socio-political events and their impact on international business are hard to come by [for recent summaries see Kobrin (1982). Simon (1982) and Robock and Simmonds (19X4)]. First. the collection of data over long time periods and covering a sufficiently large number of countries presents substantial problems of accuracy, validity and comparability. Second, while the

most dramatic impacts are readily reported in the press [e.g.. major nationalizations and confiscation of assets), the quality of the reporting is not always homogeneous or reliable. Third. there arc no time series or data banks that report on the multitude of minor inconveniences and obstacles imposed on foreign companies on a daily basis by governments and other environmental actors. In fact, this information is seldom collected systematically by individual corporations. thus limiting the possibili-

ties of case study methodologies. Fourth. the application of quantitative methodologies to political phenomena is a relatively new development. as attested by Gillespie and Nesvold (1971). Armstrong (1978). Choucri and Robinson (1978). Heuer (1978) and Simon (1982). Finally. and perhaps most critically, there has not been sufficient theoretical work until recently that allowed for the specifica- tion of causal relationships between political. social. and economic data. and the contingencies faced by firms.

-7. I. Expropruflon

There have been a number of serious studies attemptin g to explain the incidence of expropriation across countries. time and industries [see Burton and Inoue (1984)]. Truitt (1970 and 1974) and Hawkins et al. (1976) concluded, inter alia. that the extractive and service sectors were more vulnerable to expropriation. that certain organizational characteristics. such as size and ownership structure. were associated with a higher frequency of takeover. and that economic motivations. and not the ideological rhetoric designed for public consumption. dominated public policy as expropria- tions were directed at controlling economic activities vital to the nation. Bradley (1977) cast doubts on the widely held belief that joint ventures reduced political risks, and showed that very high and. surprisingly, very low levels of technological complexity stem to be a deterrent against cxproprintion. Bradley’s data also indicated that those nffiliatcs which were highly integrated into a multinational system were less likely to suffer expropriation. particularly if cutting them off from the parent

company network would rcndcr them of littlc value. Four more rcccnt studiss have covcrcd sonic of the same ground as those ahovc itsing larger data

hascs. and have tcstcd for further hypothcsi/.cL! relationships among country. corporate and con- tingcncy variables. Jodicc (1980) found littlc cvidcncc that willingness to cxpropriatc was associatccl with the level of economic dcvclopment of the host country, but instead confirmed that the ‘capacity’ of the state [as measured by the ratio of central govcrnmcnt revcnuc to GDP] was strongly correlated with the incidcncc of expropriation. In addition. hc established that governing elites tend to USC expropriation as a mean of distracting attention from their own shortcomings in times of increasing political turmoil. A Kobrin (1980). using the same data base, focused on industry and corporate specific factors associated with the incidence of forced divestment. Hc found that in most instances of ‘selective’ (as opposed to wholesale) expropriations, countries acted mainly in the more highly sensitive sectors, that is, those such as agriculture. mining and petroleum where national priorities and sensibilities were the largest. 5 He also confirmed that technological complexity and global integration of the subsidiary help reduce vulnerability to takeovers, while the level of industry maturity encourages it in the case of manufacturing affiliates.

Burton and Inoue (1984) examined an even larger data base consisting of 1857 cases of expropriation which included for the first time the experience of Japanese investors. Their findings related sectoral patterns of expropriation to the country’s stage of economic development. Their

’ The Jodice (1980) and Kobrin (1980) paper5 have a wealth of d;tu on the history and regional distrihutwn of expropriatwns

in rcccn~ years. They show. among other things. that some of the smaller countries. e.g.. Iraly. the Nerherlands. Belyum and

C;rnada, have suffcwd a much higher ratio of share of expropriations to share of FDI than the United States. Bncain or

Fmnce. Also. Afnca and the Middle Est show a grcz~ccr relative propensity IO expropriate than Asia or [the luwe~~!] Lattn

America.

’ An interesting dtvcrgencc between the Kobrin (19X0) and Burton and lnoue (1984) [discussed hclow] studies concerns this

pant. While the former argues that mass expropriations account for slightly over 10% of 311 takings m his sample. the latccr.

using csscnridly the same CICIILI base. conclude that Ixge sccllc nalionalizations xcount for mow than 70% of all firms taken.

The discrepancy arises from the use of 3~1s [by Kohrin] versus firms [by Burton and Inouel. although in both analyses 11

appcns that selcctiwty is on the rise.

analysis. however. is limited by the fact that no relative intensity of expropriation can be determined

by the lack. of bass data on the stock of investment by region and sector. and by the distorting effect

of several large scale takings in Cuba and Africa. Finally. a study by Juhl (1985). whose sources and

data base are not specified. supported many of these findings. particularly the view that vulncrabilit>

increases with the host countrlr_‘s capacity- to assume responsibility for the affiliate.

The drxth of e\-idrnce on contingencies other than e.xpropriation make it practically impossible to

confirm the existence of c;1us;11 relatIonships bvith the confidence of the studies cited above. The

literature on foreign investment in developin g countries, ho\vever. provides ample basis for specifying

hypothetical relationships [e.g.. Vernon (1971 and 1977). Reuhrr (1973). Robinson (1976). Penrose

(1976). Lnll and Streeten (1977) and Frank (19SO)]. T~vn basic models can he used to do this. First,

there is the series of propositic)ns deriving from the rtrlative bargaining powxx model and its corollary.

the obsnlescent bargain paradigm. The second concerns the dcpendrncia mndcl and its assumptions

about rclativc gains and losses to the host country from dependency on foreign investment. The

formtx attempts to judge optimal policy on the basis of social cost-benefit analysis. suhJect to the

txistcncc of both firm- and cnuntrv-sptxific xivantagcs and to the opportunities for internalizing

transactions within the firm. The latter, on the other hand. plnctx ;I premium on non-txonomic

factors such as national identity and self-rcliancs.

Fclur rcccnt studies bring to hear sornc of thcsc hypothcscs to narrowly dcfincd arcas in the

rclation+ip bctwccn host country and nillltin~~tion~~l investor. xnd do so in ;I way bvhich is consistent

Lvith the ;I~OC.C conclusions. Fagrc 2nd Wells (19S?,) focusccl on the o\vncrship pcAicics of multin;l-

tional cornpanics in Latin /\rncricx They conclutlcd that technology [I<& I>/ s;llc.\]. product diffcr-

cnti:iIion [;itlvcrti.\inS/sal~s]. 2nd market xccss [both intra-corporate tran\fcr5 and clport volume]

;trc signific:lntly corrclatcd Lvith corporate bar gaining pokvcr. I’ovntcr (19SZ) cxamincd 3 sample of

10-t foreign suhsidiarics opcratin g in Tan/.;lnia. Zambia. IndoncGa and Kenya, which hxi apcri-

cnccd g~xcrnriicnt intcrvunti~ln ranging from expropriation to minor f<xms of harassment hctwecn

1970 and 1975. this findings confirm the propositions that control over sourcin g of production inputs

~tnti wlcs to associated companies arc a dctcrrcnt to hat government intervention. A high Ievcl of

operational and managerial complouity of the suhhidiary also stems to provide insurance against

intcrfcrcncc. On the other h:lnd. large firm\ opaltin, 17 in strategically important fields [to the hobt

country] \vcrc: found to txpcricncc aboveavcrqc intt’rvcntion. Finally. f’oyntcr discovcrcd that

managers of foreign firms who pursue aggrasivc policies of lobbyin, 0 for their C;ILISC’S. not only wt’rc:

better informed of the political winds. but ~~~cctuicd in lowering the IcvtA of intervention by the

government.

Lccrau (19M) trstcd the impact of firm-specific ad\,antagrs [technology Icadtxship, advertising

intensity. asset size and export intensity] as well as country-specific advantages [market attrxtivcnr’5s

and industry competition] on three sets of dcpcndent variables: actual equity ownership held by the

foreign firm. bargaining SLICC~‘SS [a firm- and country-corrected ownership vari;lble] and ‘cffcctixr

control‘. HL’ confirmed and extended the previous findings on the impact of unique corpordtr:

rcx)urccs on bargaining strength. but went further in concluding that the same firm-specific

advantages xc positivtzly correlated with the cxrcise of effcctivo control [that is. control by thtz

p:lrent ovc‘r key ;l.~pc’cts of the venture] cvcn in the abscncc of maljority ownership. Furthermore, hc

found a strong linear relationship bctwtxn the success of the vcnturr: [;L composite variable which

included profitability. manclgement satisfaction with results and pcrformancc rclativo to other

companies in the same industry and country] and the degree of effective control the parent txerciscd

over the affiliate. In contrast. the relationship hetwccn ownership and SLICC~‘SS was J-shaped. with

50/50 arrangements faring the worst. Finally, Kim (1985) e.xtended Poynter’s analysis with a detailed

look at the level of industry competition and the ‘political responsiveness’ of the subsidiary relative

to the degree of government intenention in the firm’s operations.

2.3. Palirical assessmenr models

Until very recently. most international firms limited their analysis of the political climate in a

country to casual observations by ‘local experts’ or corporate ‘old hands’ sent in for this purpose.

and to such occasions when a particular new investment or financial commitment was being

considered. If management perceived political risks to be high. the investment would be cancelled or

postponed. or a ‘risk premium’ would be added to the calculations to account for the higher

probability of loss. Seldom was this exercise conceived as an ongoing proposition: unless a major

catastrophe occured, the country’s political rating was unlikely to be reassessed.

In one of the earliest surveys of corporate practice in this area, Stobuugh (1969) reported a

prevalence for the ‘go/no go’ or ‘premium for risk’ methods involving little quantification or

sophistication. Root (1968) also showed the lack of systematic approaches to risk assessment by U.S.

multinationals. as did Marois (1981) for French companies. Finally, Rummel and Heenan (1978)

confirmed the use of casual obsenation by trusted corporate officials as a preferred method of

assessment. It goes without saying that these idiosyncratic/impressionistic approaches suffer from

excessive subjectivity that can be dangerous and misleading. Old stcrcotypos of foreign xxi&es.

rooted in either the corporate or individual mind, can play a vital and often distorting role in the

decision-making process.

Much formalization :lnd modclling of macro political and economic risks has occurred in the Iaht

fifteen years. ’ The major international banks. spurred by dramatic incrcascs in lcntling to less

dcvclopcd countries since 1974. were primarily conccrncd with what they called sovereign country

risk, essentially the prospects for default or rescheduling of cxtcrnal debt by the borrowing nation.

The specific nature of the risks involved lend itself to systematic analysis of macroeconomic data.

although there was gencrul recognition that some subjective of judgmental elcmcnts needed to be

included as well. Van Agtmarl (1976) reported on some early efforts in this direction involving both

quantitative and qualitative measures. More recently, Nagy (1979 and 1984) proposed a ‘structured

qualitative approach to the quantification of country risk’ that combines an assessment of the size of

loss with the probability of occurrence for different type of borrowers over time in a discounted

present value model. Krayenbuehl (1985) suggests that the global assessment be divided into a

political component [i.e., the will to honor external obligations] and a ‘transfer’ risk consisting of a

solvency and a liquidity measure of the ability to pay. Finally. Mascarrnhas and Sand (1985). in a

comprehensive review of U.S. bank practices. identified four major organizational approaches to

country risk assessment which varied in technical and structural sophistication and which produced

significantly different results.

A second group of these general surveys consist of a number of ‘expert assessments, typically

obtained as the end product of a multi-stage consultation process that may or may not involve

Delphi methods. Some of these reports might include econometric data as well. but their major

characteristic is the progressive rankin, 0 of a lrtrgs number of countries according to ~1 more or less

explicit logic of analysis. The BERI [Business Environment Risk Index] service is the oldest of these.

and consists of a rating system that ranks countries on the basis of four subcategories highlighting

political. operational. financial. and nationalistic factors. Judgments on 48 countries are made by ;1

panel of experts located throughout the world. processed and sent back for Jnother iteration. BERI

also produces detailed forecast reports for certain countries and a lending risk rating evaluating a

countn’s credit worthiness over the follo\ving five years.

Competing rating systems utilizing similar methodology have been developed by Frost Sr Sullivan

[the LVorld Political Risk Forecast]. Business International and Data Resources Inc. [Policon]. hlost

of these are available to users on-line and. at least in the case of Policon. users may alter the weight

of different variables or include their o\vn judgmental information whenever considered superior to

the model’s. T~vo financially oriented rating systems worth noting are the Institutional Investor’s

Count? Credit Rating and Euromoney’s Country Risk Index covering 109 and 116 countries

respectively. The latest entry to the ‘expert’ assessment rating field is by the Futures Group: their

Political Stability Prospects reports combine observational data in formal models with e?cpert

generated opinions to produce ;1 stability inde?c on n probabilistic distribution.

These and other similar techniques have the advantage of permittin, 0 rank ordering of different

environments on a fairly comparative busis. They also allo~v. in some c;~sL’s. for a significant degree of

flexibility since the weight associated Lvith the various criteria can hc modified to suit different

circumstances. However. the ranking can hc only 3s a flood as the judgments which go into their

comp<xicnts. and scvcral observers h;lvc notcd the tcndsncy to utili1.c ‘cstahlishmcnt’ private sector

cxpcrts that may not ncccssarily view cvcnts dispassionately. Furthcrmcxc. thcsc ratings arc st:itic by

definition: they rcprcscnt ;I view of past cvcnts and conditions that may hear no rcl;ltionship to the

future. The most serious criticism in this scnsc is that as long as the relationship hctwcen

socio-ccorloillic factors and p~~litical risk rcmain5 implicit in the cxpcrts’ minds no cv:tlu2tion of the

rating’s utility for ;I specific ;1pplic;rtion can be mailc.

TWX) models dcvclopcd in the 197Oh xc based on such explicit causal rclation.ships anJ rely

prim;lrily on cconomctric and other objcctivc datx Perhaps the bat known of thchc is the Political

System Stability Index first described by Iixnclcl et al (1975) and kiter clahor;~tccl in ~Iarndcl (1979).

By measuring directly ;I scriss of discrctc components of the political and social cnvironmcnt [e.g.,

number of riots, ethnolinguistic fragmentation. and Icgislative effectiveness. among others], the

resulting index is claimed to be free of judgmental inferences or distortions. One cannot escape,

however. the model’s implicit assumption that it accurately represents reality in both its structure and

the choice of variables. In this sense. ;I major innovation of the model was the addition of confidence

estimates which were assigned to the index scores for each component and each country. The second

model, the Knudsen (1974) ‘ecological’ approach, is based on the notions first put forth by Gurr

(1971) that a high level of national frustration will exist whenever there is ~1 gap between the

aspirations of ;1 people and their welfare. both dynamic concepts. If combined bvith ;I visible

foreign-owned sector, such frustration may Icad to intervention or expropriation. 3s foreign firms

serve as useful scapegoats to the failure of the existing political order to satisfy the economic and

political yearnings of the people.

Regardless of the thoroughness of the model’s specifications or the accuracy of its measures. all

these methods of estimating environmental risk share two unavoidable drawbacks. First. they are

macro-risk oriented and largely ignore the need of the individual firm for custom-tailored mcasure-

ment of project-specific risks. Although useful as n first-order indicator of the potential dangers

threatening ;I given investment [a ‘red flag’ function as Kobrin calls it]. exclusive reliance on broad

measures of risk would tend to overstate the threat to specific projects that may be immune to

intervention under most circumstnnccs. and may fail to anticipate the partial losses that would result

from a gradual tightening of operating freedom facing foreign firms in many developing and

developed countries. Second, these models are based on historical data that may be totally or partially irrelevant for

future conditions. For example. recent high levels of political turmoil leading to a radical change in government may appear under various quantitative indices as evidence of a high degree of political

instability. While this may be undeniable for the immediate past. does it signify that instability will continue into the future? Or is the new government more likely to address the root causes of past instability and lead the nation to a new era of prosperity and tranquility? Obviously. no time series analysis can answer these questions adequately. Furthermore. to the extent that the data fed into the analysis are not entirely current. there will be a potentially significant gap between the last period for which data were available and current conditions. Given the rate of change of political and social phenomena in the less developed countries. and the difficulties and commensurate delays in generating reliable data in many of them, this is not a trivial problem.

The environmental turbulence that characterized most of the 1970s. culminating with the fall of the Iranian monarchy in 1979. gave extraordinary impetus to the development of in-house capabili-

ties in political and economic assessment among the world’s largest international corporations. A survey conducttxt on behalf of the U.S. Conference Board [Blank et aI. (1 YSO)] confirmed the rise in corporate interest in political risk analysis durin, 0 the dccnde. It concluded. however. that most of thc>c efforts consisted of intuitive and unsystematic atkmpts to translate vague notions of the ‘quality of the invcstmcnt climate’ into rucorllmendations for invc.\tmcnt policy.

.A good txamplc of an cxtcnsivc corporate model is tht ESP [for txonomic. social and political] bylrtcrn dcvclopcd by Dow Chemical for their Latin American operations [Miclucl (IY7X ;~nd IYSO)]. Ck~\v’s approach has the advantages that it is .\pccifically tailored to their needs and that it involves lint :uid senior corporate officials in the ;isscs5mcnt, thus assuring lht the results will bc taken rclativcly seriously. It docs not deal, however. with the need for clear specification of causality. relying instc;kcl on the cxpcriencc of the Icaders of the asx~~mcnt tcamh to interpret events correctly and consi~tcntly. It ako implies high costs and ;I large time commitment. thus limiting its applicability to ;i few countries per year at best.

Other companies have mudr ue of xxnurio mrthodologics in an attempt to deal with socio-politi- ~‘31 projections [Kaubitschck (1933) and Wxk (1985)]. .A relarrd approach. much cited in the litcr;Lturc, is that developed by Shell Oil to assess the probability that contracts for tht: exploration. development and production of oil in a certain country will bz muintaincd on an equitable basis for a period of up to ten years. As described by Bunn and Mustsfaoglu (1978) and G&&in. Pearson and Silbrrgh (1978). the Shell approach and its subsequent variants [e.g., the models developed by Risk Insights. Inc. of New York] include a formal specification of th e relationships involved, expert opinions constrained in a fashion designed to limit judgment errors. and ;I sophisticated statistical algorithm to combine the results of both aggregate econometric data and individual assessments. As Kobrin (1981) views it. this is one of ‘the most sophisticated and effective approaches to political risk

asxssmcnt that existed 31 that time. Its major limitation is again the cost issue. To apply thz methodology to a large number of countries, or. for that mattt‘r, to ;1 number of industries with diffcrsnt characteristics and risk profiles, would be extrrmsly costly and cumbersome

The multitude of studies and models described above arc indicative of the complexity and multidimcnsionality implied in measuring political risk which are specific to the foreign activities of

‘30

GENERAL (KACRO 1

ORIENTATION

Grand Tours

L______________

:Putures croup:

ci’

I

COUNTR C ES COUNTHIes

ASPROjSPAIR

ESP

individual firms across many countries. The various analyses of the expropriation cxpcrience of

foreign invators have yizldtxl significant akkncL: of the importance of conG&xations rood in both

the national cnvironmcnt [cultural. politicd. social 2nd economic] LLS well as in industry, firm. and

project [structural] characteristics. The latter have been confirmd by the mow recent studies on the

dctcrminants of bargaining powr. The gcncral thrust of prxticc in the field. howcwr. tends to be

polarized between those models ad tcchniqucs aimed dt measuring macro politicul risks on a

comparative basis for a large number of subject countries, and those which arc spccific to ;t firm’s

needs but which xc limitd in their gtxgraphic scope. Fig. 3 illubtrata this dichotomy, whcrc the

search for the i&xl Jpproach obviously Ic;ds to the bottom right-hand corner of the figure without

much succt’~s to date.

I\-lacro models must play an important role in this scorch. V.‘hile it is true that political stability k

no guarantc’c of the abscncc of potential r’cposurc to loss. nor is instability ncccssarily associated

with the probability of loss. it remains that 755 of all instances of expropriation have ken linked to

‘31

stage 1:

Macro Risks

CTUNTRY CHARACTERISTICS

2

ECONOMIC, SOCIAL AND f-c

POLITICAL FORCES LS ZZ

POSSIBLE WEB-I'S

EL. ~1, tL

E2, ~2, t2 ------

. . . . * . * * . En, P", t,

-,

PROJECT CHARACTl?RISTlCS POSSIBLE OUTCOttE

. Industry factors

. Corporate factors

_ Structural factors

Ox/L. Px/L* tx/1

0x/2. Px/2* tx/2

. nanaqerial Eactors OX/". Px/n* tx/n

rcgimc changes. The prcfcrrcd approach niuht thcrcforc have the cafxlhility of measuring rii;1cro rihks

;Incl ffic dhility to ititcrprct thcni in terms of project-qccific consitlcrations.

M~tii~)clolo~icall~, all cxidln g rriocid~ have certain 5trcngth5 and liniilation~. fixpcrt-had sy5tcni5

c;tn hc criticizccl for not alwrtys making c~us:1l rcl:Itiordiips c?cplicit and for their potcnli:kl bias in the

~u~lgnicnts of it5 nicmhcrs. Econometric niodcls often suffer from the difficulty of securing current

wurccs of data for many of the important inclcpcnclcnt variables nccc~x~ry for the analy.\is. In-house

methods can lx espensivc. timr: consuming and of limited geographic covcragc. It follous that what

ih n~dc’d i.\ an cckctic approach that combines thr: best each method has to offer and incluclcs both

mxro ad micro judgments on the risks ILI~L’CI by specific foreign affili:lte~. ’

Since risk is both ;t country and project drprndent concept. any model &signed to forecast the

probability of IOU must encompass both clrments. and assess alI contingcncics thtit can result from

changes in mltional policy. It is this ability to predict ;~n emergin g situation before it is fully

manifcstd. in order to circumvent the crisis, that is essential for the survival ad prosperity of many

foreign operations.

Fig. 4 ~ummarires the structure of the wcommcnded framework for awlysis. Its logic is rather

dmplc 2nd str:lightfonwrd. although its implcmcntation is another matter. It begins by e.xamining ;I

scrims of natiotxl characteristics - awnomic. social and political forces at work - that may or may

not be critical to the issue of political stability. Whether such is the case or not will depend on the

importance of the particular factor. on its relationship to others and on the magnitude of any

changes or discountinuitirs involved. Furthermore, the source of trouble could be internal [e.g..

political repression] or external [e.g.. a drastic fall in commodity export prices]. It is important to

note. honever, that these forces may, whsn activated. huvr dramatically different impacts depending.

to a large extent. on the maturity and absorptive capacity of national institutions. A country where

political parties. the press. the educational establishmtznt. the financial system. etc.. have achieved

high levels of development should be able to withstand grater shocks to the system without

precipitating drastic change in the social fabric or in the institutions themselves. The institutional

framework. therefore. acts as ;L filter for the environmental forces in softening their impact on

events. ’ An xcur;lte judgement on the qualities of this filter ib then as necessary to good political

risk forecasting 35 an understanding of the underlying forces themselves. In the end. any number of

politiccll events can be the result of these conditions. Mach with its o\vn probability distribution. as

uell ;1s a probable timetable.

Stagt: I of political risk forecastin, ~7 consists. therefore. of an asst’ssment of the forces at work in

the nation and without. the institutions temperin, 0 their effcctivrncss. and the likely events that they

may precipitate. This is the realm of ‘country’ risk analysis. kIuch of it is ;mlen;thlr to txonomctric

mndelllng and to the ust’ of e.xptxt methods similar to those dtxcrihcd carlicr. In the corresponding

sccticxi hclow we shall illustrate the variahlcs that ought to he included in this analysis. The output

should c~tahli41 ~mt: basic mcxsurcs of rcfcrcncc compx-ahIt: xross countries. it should identify

current trends and any potential brxxks in them. and it should delimit the ;Lrc;Is of concern that ma)

harbor the seeds of potcnti:ll threats to foreign invcstcxs.

Xot all uvcnts. howcvcr. will have Gmilar co~~scqucncus for diffcrcnt proJcct.\. As cstabli~hucl in the

above rcviav. industry and ccxpcxatc factors. as ~vcll AS ;1 numhcr of characteristics of the structure

of the invcstmcnt. will ploy ;1 dctcrminant role in the likclihoc>d and prob;tblc cstcnt ~)f any potential

IO\KY. Thlls. for each project or affiliate, cvcry pos.siblc event an bc xcn ;IS having ;I set of outcomca

or c~mxq~~c~~ccs which arc unique to it. Thcx altcrnativc outcomes have each ;I ccrt;lin probability

dihtributic>n ~nci time horizon. Stage I I of the political risk forecast consist5 of bringing into sharper

rclicf the fcatura of the: project that cithcr incxlsc of dimini.\h the possible negative conxqucnca of

cxh cvcnt. In this scnss it is akin to estimating the conditional probabilities of various contingcncics

given ;L number of possible events. Since the latter may or may not hc mutually c.xuclusive. final

etirn:lres of the likelihood of loss can bc arrived ;~t by the proper mathcmaticul manipulation of the

probability c’~tim:~ttx

What f<>llows is not ;L specification of such ;1 comprehe:nGvc model, .since ohvioubly that would

require considcrahle more space than availahlc ha-c UKI an intimate: knowlcdgc of the circumstanocs

appllcahk to ;I specific investment project. ” Fig. 5 summarizcb the approach. The country a.\sociatcd

riA analysis involves ;1 total of 11 sets of vuriahlcs or compo.site factors that mubt hc monitored on

an ongoing basis. yieldin g estimates of possible t’vcnts. their prohahility of occurr‘nct’ and the

exp~tcd timetable. The project associated risk examines an additional 16 variables th;Lt arc likely to

have ;I major impact on the consc’qur‘nces of political change on a given foreign affiliate.

3.1. Counrr?_ Associated Rusk

An arbitrary but useful distinction can be made between economic and socio-political factors on the one hand. and between internal and external sources of risk on the other. The division is arbitrary because developments in all four areas have interactive effects on the other variables under scrutiny. It is important for the analyst to be sensitized to this web of hidden relationships and not follow the simple structure blindly. Also. it should be noted that many external influences. particularly in the economic sphere. can be systemic to the world. It is in such cases that the institutional elements of the analysis can serve to distinguish between countries in terms of expected impacts. Economrc Fucfors - fnfernul. The analyst must first understand the basic components of the host nation’s economy. its rate of development. and its vulnerability in order to anticipate factors that can affect the general business environment. Six major headings may be useful to organize the data as illustrated in fig. 6a. After this stage of the process. the analyst should have a reasonable picture of Lvhich economic variables are critical for continuity in the country’s current economic development strategy. the vulnerability of the strategy to failure in any of these critical links. the likelihood of such failures and, as a result, the probability that performance will fall short of expectations. The objective is not economic analysis per se:. but a search for what one might call the potential for trouble. Econontrc Furors - E.rrerrrtrl. In order to gain a better understanding of the nation’s economic conditions, the analyst must next turn to its external payments position. What are the country’s international obligations. the extent of foreign indebtedness and servicing requirements, its level of dlvcrsification of export earnings, the exposure to commodity price fluctuations. the rigidity of import requirements. etc. Five headings would bc helpful in organizin g the analysis of thcsc issues [fig. 6a].

This set of questions sewes to determine to what extent external constraints will dictate domestic economic policy. A high degree of dependency and instability together with external debt scrxwng difficulties will substrlntiall> increase the r&k of host gwrrnment interfrrencr with foreign investors in the county. both in terms of expropriation and convertibility. In .LIozambiqur, shortly after the revolution, the government. faced bvith severe external payment difficulties, nutionalized those

: .-.,,_w. F,., - _

--

snwprisss that consumed significant ~rn~unt~ of forsIgn rxchaqs. Likrwisr in Xicargua. the grtlvt:

shortage of foreign eschange after Somoza’s ouster prompted the new rewlutionan wvernment to

t;lk control of the main sources of foreign earnings. And the frequent USC‘ of ‘trmporarq’ _ =

suspensions of dividend convertibility during difficult times in countries like Brazil underscore the

nerd for such anrtl>sis.

Sac-IO-Po/irrwl F11crors I /rrttTnui. To understand the politicA srtuation of the host cnuntr)- rtnd its

potential for inspirinp change one needs to begin with thr cohrsiventx of the soc~.tl structure. the

disparity brtxesn people’s briisfs 2nd aspirations on the one hand and the quality of Iradwship on

the other. the relative pwver of government and opposition groups and the strength 2nd traditions of

national institutions. Again. fig. 6~ illustrates six major headings that xill guide the analysis. It

should he evident that mwt of the information soupht under these various hruciings is highly

judgmental 2nd difficult to ev3Iu3te ntrjcctivcly. Sources intimrttci) familiar with loccll condititws a-e

tlssenticll to the analysis. Thrrefore, it i.s aivisrthle that awnal cspcrt opinion he c3htaintxi and that

their views be crossesamined hl; on-the-field assessments carried out tt? ;I large rstrnt by the firm’s

O=.Il local 5taff.

SOCYO- FoIJI~uJI FCJC.[II~.Y - E.VICY-IICJ~. Political in3tahility is often cxtcrnally inductA. .-\t hat. csternal

influcnccs can euxwhatc internal conditicws by playing on the fc;rrs or fru.\tr;Ition.s of the Ioc:~l

pnprtlrttictn. er hv lcndrn g moral. financial or itlwlogic;~l suppttrt to ~~ppcGtion groups. Five such

prt47lcrri ;xtxs ciea-b~ claw scrutzny as indicattxi in fig. 62. As xvith thu section ahwc. the knctwlcclgc

and data sow-cc5 rcquircd to complctc thi5 part of the an~~l~~is arc highly spccial~red. A Gmilar

conclusion is thins wxrtxntcd 2s to the utility of scckitig cxpcrt dclbiw supplcnw~tctf hv the ucu3 (31

thaw in the ficld.

local management. corporate culture and management philosophy. political responsiveness. and

financial policies. “’

It is obvious from the precedin, 0 list that the cost in time and money of carving out such detailed

analysis for each country and each operation throughout the world would be prohibitive for any

moderately large multinational corporation. Thus. any corporate system of forecasting political risks

ought to strive for a compromise betsveen the general and broadly based information available from

multi-country rating and evaluation services, and the provision of case-by-case specific inputs from

internal and contractual sources.

The macro or country risk component of the analysis is particularly suited to standardization and

the systematic manipulation of large time series data bases. No individual firm. except perhaps the

very large. can hope to duplicate the resources available to specialized agencies for collecting.

processing and analyzin g macroeconomic and socio-political data on a global basis. There are

significant economies of scale from operating across multiple countries, and specialized agencies can

amortize the cost of developing and constantly up-&tin g their forecasting systems over a large

customer base. Thcrc remains, however, a genuine requirement for an internal function in assessing

macro ri5ks. Corporate executives familiar with thu content of the models used by external analysts

[i.e., the internal logic. sources of data. assumptions and specifications] should perform a control

function which is free from any loyalty or commitment to the model itself or to its component parts.

Their task is to make sure that the cxtornally supplied assessments are suitahlc to the company’s

needs and acccptablc to those who have to act on the basis of the information provided. Furthcr-

more. specific project characteristics may limit the scope of the macro political cnvironmcnt which is

of concern. If a given project or affiliate can bc affected only by a narrow rungc of political events. it

WOLIICI facilitate the internal control funclion accordingly.

The second part of the analysis, that is, once the probability and time horizon of certain events

have been established, can only bc performed by those intimately familiar with the company’s

operations. How will certain events affect the profitability of, or the capacity to repatriate funds

from, a given project or subsidiary needs to be dctcrminsd on the basis of data only available to

management. Therefore. a second role for the function of political risk assessment within the

corporation consists of interpreting the results of the country risk forecasts in terms of the realities of

industr);. corporate, structural and managerial factors only known internally.

There is, however. an important qualitative difference between these two roles. Monitoring the

quality and accuracy of country risk forecasts provided by external services and adapting them to the

corporate reality should be a function performed and coordinated centrally. While operational

management can and should have an input to the process. the modification of the model’s

specifications and the interpretation of their biases can only be appropriately conducted after

considrxable experience with such a system over relatively long time periods. Given the value of

institutional memory in this process. it would be logical to centralize responsibility accordingly. The

second role, that of evaluating micro risks from a given set of country risk forecasts, has to be carried

out at the local level. Obviously. corporate involvement may be essential to assure impartiality and

comparability across countries and projects. but since local management will be called upon to act on

the results of the analysis, they must be party to its conclusions.

‘I’ For romc of the pokics in question we Bdley (1977). Doz and Prahdad (1980). Gladwin and Walter (19X0). Shapiro

(IYRI ). Eltcman and Stonchlll (19X). Ghdar. Kohrin clnd Moran (lYti3). Ghadar and &foran (1944). and Encarnation and

Vachanr ( lYR5).

4. Conclusions

In assessing whether to set up. expand or contract operations in a given country one should

distinguish between two sets of issues. The first bus to do with the contribution the project or the

affiliate is likely to make to the corporation’s global strategy. including the returns from the venture

proper as well as any synergistic or competitive contributions to other units in the corporate system.

Such an assessment of ‘strategic attractiveness’ rsill be based on a number of factors such as the

degree of global competition prevalent in the industp- [as opposed to competition based on

fragmented national markets]. the size and importance of the market. the fit Lvith the company’s long

term strategic priorities. and so forth. This should result in a differentiated approach to global

opportunities. The more attractive a particular location and the more critical to the achievement of

corporate objectives. the more uillin g the company should be to undertake a high level of risk and to

commit the necessary resources. It follo\vs that the higher the priority accorded to a particular

subsidiary in the company’s global strategy. the greater the firm’s need for management integration

and control with respect to that subsidiarv.

These views ought to be tempered by -a second set of assessments concerning the quality of the

‘investment climate’ in the country in question. To the extent that the risk of political upheaval and

intervention is high. the expected returns may not mnterialize. or they may he significantly reduced.

irrespective of the market’s attractiveness. As the investment climate deteriorates. the foreign investor

will attempt to reduce its financial, technological 2nd human resource commitments [and exposure]

bvhile attempting to retain ;1 measure of market prtxncu as allowahlc under the circumstances. This

may call for unorthodox approaches to wvncrship and control which take into account the wed to

minimize exposure consistent with prcscrving a position in the mnrkct. In contrast. an exccllcnt

investment climate is no substitute for mxrkct potential. Joint ventures and other indcpcndcnt

arms-length tr;msxtions provide ;I useful vehicle to gain a foothold in such markets without unduly

committing scarce corporate rcsourccs. Where poor market prospects coexist Lvith ;I bad invcstmcnt

climate, it is clar that the firm Lvill tend to limit both its commitments and exposure by resorting to

independent market transactions, if at illI. Fig. 7 summarixs these choices.

It should have become clear by now that there exists such ~1 diversity of factors impinging on an

evaluation of the risk profile of ;L particular corporate project that the challcngc of constructing ;L

lliqh

rcrd i um

reeour~es and hiqh and hum‘,” I”“eRtme”ts: mdrket prencnce: aim for

to*cr.*“cr for commercial axept normal ccmmt:rcial minority position with

r~skr: wholly-owned rusks: ma jority-aned licenalnq *s a lonq-

.,ff,l,ates preferred. afflliatc preferred. term h“dqe.

mintal” hiqh resource Unrillinq to commit Little interest in

commitment and risk slqnlflcant rcBo”rces: market prcaence: pursue

tolerance sub jcct to prefer to .lct through only if possible

bettsr alecrnatlvc io~nt venture if without financial

I”d,ff,.re”t to markat I.,tt*e if nny rFR”IIrcc HO ,ntercst

opportunities: token commttmr”t dcnirnblc: except for “ccJnI”“al

f ,n..nc,al or hum.a” export aa*<*s aqP”tR / experts or limited commitment pornihle: pr.,fc.rrrd vchiclc for licensinq

indvpendvnt dintributor any mark-t activity: ‘lq~fWUC”t~.

or- joint venture. Llcenalnq poaeiblc.

single model that will faithfully and accurately represent their interaction and complexity is

monumental. Rough rankings of countries in terms of their relative political stability have limited use

as predictors of potential losses in specific situations. The facts that causality is not easily determined

in political phenomena. that up-to-date information is difficult to obtain. and that stability in itself is

not necessarily a good measure of risk. all contribute to the many doubts often expressed about

existing methods. Furthermore. the nature of the industry and the investor. and the timing and

characteristics of the project are critical variables that alter significantly the risk profile within the

same set of economic and political conditions.

Yet. no human being could possibly master this complexity for more than just a handful of

countries. Unaided by standardized quantitative tools. the political risk analyst would drown in a sea

of information. Judgment can best be applied when the range of variables to consider has been

reduced to a manageable set. Herein lies the challenge. Modelling political risk at the corporate level

must make use of good measures of quantifiable variables and systematic analysis that can reduce

large quantities of data. according to accepted causal models. to probabilistic estimates of possible

events in an efficient fashion. Secondly, it must call for many qualitative assessments of elusive

trends. such as levels of national aspirations and frustration, that can only be obtained through

intimate knowledge of the terrain. Thirdly, it must make all of this relevant to the particular project

at hand. And finally. it demands good judgement above all. to mix the many inputs in a coherent

manner so as to spot. as flolmcs. the dog that did not bark in the night. It follows that a systematic

approach such as is proposed here must be limited to those countries or areas of the world where

major investments or competitive positions are at stake.

A final question that may he asked is how best to incorporate this analysis into the strategic

planning process. The lack of acccptcd standards has resulted in significant discnchantmcnt and

hkcpticism with political risk analysis among many multinational corporations. Is political risk

forccasting one more short-lived corporate fad’? Most executives would readily agree with the

desirability of having such an input availahlc to thu planning function. but not many firms have

made the ncccssary invcstmcnts in terms of both staff and administrative systems to gcncratc the

information and incorporate it into the decision process. As cxistin g models are pcrfcctcd. one might

hope that the rcquircd commitment and organizational linkages will emerge.

Rcfcrenccs

.~rm~trong. J. Scot!. 197X. Long range forccasring (W~lcy. New York).

r\whcr. Wilham 1982. Pali[ical forccatrng: The missmg link. Journal of Forecasting I. 227-239.

Au~m. Jama E. ad David B. Yoffie. 19X4, Political forec;ls[ing as ;I managcmcnl owl. Journcll of Forccahting 3. 395-408.

Ba.Gry. G.R. and R. tlrulr Dekmqian 1985. hlNCs ad rhc Irarxm rcvolutwn: An empiric4 study. Management

Inwrn~~iond Rcvicw 25. 67-75.

Blank. Srrphcn, et al.. 19x0. Asses~ng the political environment: An cmcrymg funccwn in inrerrx~rional cornpanics. Report no.

79-l (Confcrcncc Board. New York).

Brdlcy. David G.. 1977. Mannging against expropriation. Harvd Dusine.rs Review. July-fug.. 75-83.

Bun”. D.W. amI M.M. Mustafaoglu. 1978. Forecasting politd rihk. Managcmcnt Szicnccr. Nov.. 1557-1567.

Burton. F.N. and Hisashi Inouc. 19X-l. E?cpropria[ions of foreign-owncd firms in dcvcloping countries: A cross-national

clndysls. Journal of World Trade Law. Scpr.-Oct.. 396-414.

Choucri. N&i and Thomas W. Robinson. cds.. 197X. Forcasting in inrcrnational rcl;ltions: Theory. methods. prohlcms.

prospects (Freeman. San Frwcisco. CA).

Dc 13 Torrc. Jose 19gl. Foreign invcsrmcnt and economic dcvclopmcnc: Conflict zmd negotiation. Journal of Internclrmncll

Busmcss Studies. Fall. 9-32.

Dc la Torrc. Jose and David H. Ncckar. 1987. Forecasting polttical risk. in: Spyros Makridakis and Stcvcn C. Whcclwnght.

cds.. The handhook of forecasting: A manager’s guide. 2nd cd. (Wiley, New York),

Dc St. Jorrc. J.. 1983. IRIS: A study of how to fail in huslncss. The Intcrnatwncll Hcrnlcl Trlhunc. April 20.

DOL Yves and C.K. Pruhalad, 1980. How &lSCs cope ~lth hobt government mte~ent~on. Haward Bu~mrss Re\-~w.

March-.Apnl, 119-157.

Elteman. Da\ld K. and .Arthur I. Stonehrll. 1YX2. Rextmg to pohtlcsl r&k. III: Slult~nation~l busmeas finawe. ch. 6

(Addison-Wale>. Readmg. .LU).

Encamation. Denms J. and Sushd Vxham. 19X5. Foreign ownership: When hosts change the rules. Harvard Business Rewew.

Sept.-Oct.. 152-160.

Fagre, ;Vath.m and Louts T. Wells. Jr.. 19X2. Barglunmg pouer of multlnatlonals and ho3.t governments. JoumJl of

Intemauonsl Business Studw. Fall. Y-23.

Frank. Isarah. 19X0. Foretgn enterpnse m developing countnes (Johns Hophins. Umversity Press. Baltimore. \tD).

Gebclein. C.A.. C E. Pearson end ht. Sdbergh, 197X. .-\wzsslng polItical nsh of oil in>rstmrnt ventures. JoumJl of Petroleum

Technoloev Mav 725-730. _~ . Ghadar. Fanbarr and Thrctiorr H. Moran. eds.. 1YXJ. Pohtlcal nsk management: New dimrnswns (Georgetown Umverslty.

Washmgton. DC).

Ghadfar. Fanborz. Stephen J. Kobnn and Theodore H. &loran. eds.. 11)X?. \l.magng tntematlowl political nxh: Stratcgles Jnd

techniques (Georgetown Cniversity. Washtngton. DC).

Glllesp~e. John V. and Betty A. Nehvold. eds., 1971. ,Ltxro qu~lntitative analysis. Conflict. development and drmocratlz~tlon

(Sage. Beverly Hills. CA).

Gladwm. Thomas N. and Ingo H’altcr. 1980. Multlnarwnals under fire: Lessons In the management of conflict (Wiley. ;Vew

‘iork).

Gurr. Ted Rohert. 1971. Why men rebel (Princeton Unlvervty. Prlnccton. NJ).

Haendel. Dan. 1979. Foretgn Investments and the management of pol~t~c;d nhk (Westview. Boulder. CO).

~iaentlel. Dan. Gerald T. West and Rohcrt G. Meadow. lY75. Overwas investment and pol~tval risk. ~fonograph serws no ?I

(Foreign Policy Rcxxch In\tltutc. PhlMlelphla. PA).

Hawkins. Rohcrt G.. ;\;urm.m MIntA and Xltchael Provisscero. lY76. Govcrnmcnt t.lkcovurs of U.S. forclpn :ifflltatcs, Jk)urn.d

of Intcrnatwn;d Bustnew Stud~cs. Spring 3- IS.

ticucr. Richard J.. Jr.. cd.. 1Y7X. Qumtitativc approxhc\ t(~ pollt~c;~l intclhgcncc: The CIA cxpcrwxc (Wc>t\lcw. I$oulJcr.

CO).

f Iufbxwr. G.u-y C. and P. t1. Brlggs. lY75. fi.xproprl:ltlon Io~ws and 1.11 pohcy. tI:wv:lrJ Intcrn.~t~on.~l I.;lw J~wrnal, Summer.

s53-56-8.

Jodccc. Dawd A. IYSO. Sources of change 111 Third World rcg!mc\ for forclgn dlrcct tnvchtmcnt lY6X- 1976. Intcrn;~t~on.d

C)rg;mw;ttwn. Sprmg. 177-206.

Juhl. I’., IYXS. Economically ratwnal dcbiyn of dcvcloping countrle\’ c\propriatton policx.\ towxds forctgn ~nvutmcnt.

hl;magcmcnt Intcrn;rt~on;~l Rcvicw ?S. JS-SZ.

K;;lttenstc~n. Pctcr J , 197X. IIctwccn power and plenty: Forclgn economic pcrlicics of xlvanced indwtrial \I.IIL’\ (Univcrhity of

Wiwunin Prcsh. hixlwn. WI).

K:lm. W. Ch.m. lYX5. IIc dynamic harganing power pu\ltion of multtn;~troncll~: .Llan.lglng compctltion .md host gwcrnmenr

intcrvcntion in dcvcloplng countries. Working pJpcr. Dee. (Umvcr\zty of Michigan. Ann Arhor. X11)

Knud~~, tldrald. 1973, E_xpl;umng the national propensity of cxproprtatc: .An ecological approach. Juurn.rl of Intcrnahon.il

Businrss Stud~cs. Sprmg. 5 I-7 I.

Kobrin, Stcphrn 1.. lY7Y. PolItical risk: A rcvicw and reconsideration. Journal of Intcrncltional Buzinr.\s Studicb. Spring/

Summer, 67-N

Kobrin, Stephen J.. 19X0. Foreign rntcrprise and forced divatmcnt in LDCs. Intcrn3tion~l Org;inlzlrtion. U’lntcr. 65-Xx.

Kobrin, Stephen J.. 19X1. PoIitIc;LI assessment by mtrrnationdl firms: hlodcls or mcthodologte~. “. Journal of Polxy Modrl!ng

3. 2. 25l-270.

Kohrln. Stephen J.. 19X2. Managing political risk ascssmcnt: Strcltcglc rchponse to snvironmcntal change (Univrrslty of

California. Berkeley. CA).

Kobrm, Stephen J., 1YX-t. Fnproprlauon as an attempt to control foreign firms in LDCs: Trends from 1960 to 1979.

Intcrnatwnal Studies Quarterly 2X. 329-33X.

Kraycnbuchl. Thomas E.. 19X5. Country risk: Awzssment and monitonny (Woodhead-Faulker. Cxnbridgc).

Lail, %nJaya and kid SmXten. 1977. Foreign investment, trans-nationals. and developing countrws (MacMillan. London).

Lcorclw. Donald J.. 19x4. Bargaining power. ownership. and profit;lbility of transnational corporations in dcvcloping countries.

Journal of Intcrnatlonal Business Studw. Spring/Summer. 27-43.

Xlarols. Bcmard. IYYI. Comment les cntrcpmes franc:li.\es gcrcnt Ic rquc politique. Revue Francase de Gestmn. May-Aug.

j-9,

Macarcnhas. Briancc and Ole Christian Swd. 19X5. Country-risk assc>smcnt systems in banks: Pattcms and pcrformancr.

Journal of Intcmatwnal Business Studies. Sprmg. 19-35.

Miqucl. Rafael. 197X. The case for ESP studies. Unpuhlishcd manuscript. June

\~quel. Rafael. 1950. Soms comments on ESP analysts rind its uses by management. Unpuhhshrd manuscript. Feb.

;Va_e. P3ncras. 1979/1941. Countr) nzk (Euromoney. London).

Penrose. Edith. 1976. Ounrrshlp and control: .LIult~natmnal firms m less developed countries. In: G.I(. Hcllrinrr. cd.. A world

dtkided - The less developed countnes m the intemxioncll economy (Cambridge l_‘m\rrslty. Cambridge) 147-174.

Poyntcr. Thomas .A.. 1982. Gorcmmrnt intrncnrlon in less developed countries: The experience of multinational compames.

Journal of International Busmess Studies. Spnng/Summcr. 4-Z?.

Raublt_xhek. Ruth S.. 1983. Scenarios and strategy formulanon. Research paper. June (Haward Business School. Cambridge.

%1.4).

Reuhcr. Grant L.. 1973. Private foreign Investment in development (Clarcndon. OKford).

Robock. Strfan H.. 1971. PolItical risk: Identification and asessment. Columbia Journal of World Buslncss. Julk--Aug. 6-‘0.

Robock. Stefan H. and lCennsth Simmonds. 1983. Assessing politwxl nsk and natwn;ll controls. International business and

multinatwnal enterprises. ch. 15 (I~ln. New York).

Root. Franklyn R.. 1968. U.S. hwness abroad and political risk. &fSU Business Topics. Winter. 73-80.

Rummel. R.J. and David R. Hrenan. 197X. H ow multinarwnclls unalvzr polltic risk. Harvard Bus~ncss Rswrw. Jan.-F+.

67-76.

Shapiro. AIM C.. 1981. %fanagng pollttcal r&k: A policy approach. Columbia Journal of World BusinrcT. Fall. 6.7-70.

Simon. Jeffrey D. lYR2. Polttxal nsk awssment: Past trends and future prospects. Columbia Journal of World Business. FAI.

62-71.

Slmtw. Jeffrey D.. 1984 A theorettc,~l pcrspectlve on poltitvzcli risk. Journal of Internatwn.d Busintxs Studw. Winter. 123-143.

Stmon. Jeffrey D.. 1985. Polltlccrl nsk forecasting. Futures. Apnl. 133-147.

Stohauph. Rohcrt B.. Jr.. 1969. How tu an.lly-le foreign Investment cllmares. 1Iarwrd Buvness Rewcw. Sep.-Oct.. I(H)-IOX.

Trulrt. J. Fredcnck. 1970. E.xpropn,~tuw of forcgn Invcctmcnt: Summary of the post World War II cxpencnce of Amcncan

/Iiogrc~pi~~: Jo& dc I;I TOKRE is I’rofcnsor of Intornationxl Business Strategy at the Gracluatc School of Management. University of California, Los Angeles. Sincc 1974 hc w;ls affiliated with the European Institute of f3usincss Administration [INSEAD] as ;I Professor of International Business and as Chairman of the Strategy and Environment Arca [ 19Sl-851. Prior to that time hc has held various faculty positions at Gwrgia State University, Harvard University. the U.S. Drpartmrnt of Commerce and at several institutions in Latin America. His main research interests are foreign trade and investment and international corporate strategy.

David NECKAR is Managing Director of M.J. Marchant Underwriting Ltd., ;1 trading Lloyds insurunce syndicate. He is ;I graduate of Oxford University and of INSEAD. the European Business School. He has been unkwriting political risks insurance since 1978.