'''''.

'
8 / Foreign policy decision
making: rational,
psychological, and
neurological models
JANICE GROSS STEIN

Chapter contents
Introduction 130
Commonsemical understandings of rationali;y 131
Psychological models: the 'cognitive revolution' J32
Neuroscience, emotion, and computarian J39
Conclusion 143

e Reader's guide
This chapter looks at the contribution that ratianal models af choice can make to
foreign policy analysis. Ir then examines cognitiO'e models which identifythe bounda-
ries to rationality in decision making. The chap~er then looks at new research in neu-
roscience which reca~ts the role of ratianal mo::lelsand highlights the importance af
emotions. Jt condudes with suggestions for sy;tem design that (an improve foreign
policy decision making.

Introduction

In early 2007, the Munk School of Global Aflairs at the Univer~ty of Toronto assembled a group
of scholars to think broadly aboutthe meanlng of rationality in foreign policy and conflict reso.
lution. Among them were two recent Nobel Prize laureates and a distingulshed philosopher. We
asked our colleagues to wrestle with a practical problem, that of 'applied' rationality. They
followed in the footsteps of scholars and policy makers, who for thHty years have debated the
elements of a solution to the lsrael-Palestine conflict. The o'Jtlines of a 'ratianal' solution are
broadly known an both sides ofthe divide: tvvo independent states and a sharedJerusalem. Vet
the conflict is still raging and people are still being killed. What stand, in lhe way of a rational
solution ,to this enduring conflict? Why (an the governments on both sides no! move towards a
rationakcompromisethat is known tothe participants as well as outsiders? If analysts and politi-
callea~ers on both sides knowthe broad outlines of a rabonal 'iolution, whycan'tthey getthere?
 1'

This deceptively si'Tlple questíon masks a nos! orcomplex issues. What does it meanto 'know'
a rational soluti.on? What is a 'rational' solution in a real-world setting? Ra:ional by what crite-
ria? From whose pel'spective? If one ar more of the parties to the conflict rejects the solution,
are they necessarily irrationa!'? Are there good reasons why solutions thatscholars might [00-

sider ',ational' are not chosen by government leaders? Ali these questions pain! to even bigger
questions. How ratianal are politicalleaders when they make big and importan! foreign policy
declsions? How well do concepts of rationality, as developed by philosophers ar economists,
travei jota theworld of politicsand foreign policy making? The short answer: not ver'l\'vell at ali.

Commonsensical understandings of rationality

This chapter could get badly off track by investigating the m"ny meanings of rationality. The
focus ofthis ehapter is mueh narrower-how we understand f:Jreign policy dec!sion making. I
put forward two cO.l"lmonsensical madels of ratíonê.lity in decision making.ln thefirst, rational
decision making refers to the process that people should use to choose. In a ratl0nal deeision-
making process, people should be logical and orderly. Their preference, should be ranked, at
least inLJitively, in sueh a way that if I prefer A to B, ano 8 to C, then \ prefer A to C. lf I prefer
peace to all-out \Ivar,and 1prefer ali-out warto low-Ieve! insurgency, then lshould prefer peace
to insurgency. If I violate this requirement of 'transitive' preferenees-if I prefer C to A, or insur-
gency to peace-tren I am ruled out as a ratiana! decisian maker. We need to laok elsewhere-
and we will in this chapter-to explain choices that vialate thE axioms of rationality.
As part of the process af making informed cholces, ratianal decision makers should be good
at attending to new informatian that comes alang as they are making theír choice~; they need
to 'update' their estimates in response to new reliable information tha.t contalns significant
evidence. The attentive reader.may notíce all sorts af caveats here: 'reliable' informatian that
comes from a 'tru5tworthy' source, 'new' information ar information that the decision maker
did not previously have, and 'significant' ar diagnostic evidence that spea-ksto the likelihood af
some ofthe cansequences the policy maker is considering. VVhen President Bush was consider-
ing whether af not to go to war against !raq, he VIlastold that Saddam Hussein had sought to
buy ye!low cake uranium from Niger. This information was new to the President-he had not
heard It before-and itwas diagnostic: Itsignalled that Saddam was probably seekingto develop
unconventiona! Vleapons. What the information was not, however, was reliable ar trustworthy,
and at least ane Jfficial quickly identified it as unreliable; therefore it shauld have been dis-
counted orexcluded from any kind af consideration. lhe reliability of information is a threshold
barrier that any piece af evidence should eross an its way inta the decision-making process
However, àetermining the trustworthiness af any pieee of information is often veíY difficult to
do. ln:leed. 'raticnal' processes of information management are often swamped by the quick
intuitive processes and deep cognitive biases that political1eaders use to interpret evidence.
50 far, this picture of a rational deeision maker approximates eommon sense. People who
make tmportantchoices about foreign palicy need to be logical, discriminating while open to
1
new evidence, and 'coherent' and 'consistent' in responding to logical arguments. Rational
decision makers are tl105e who are open to arguments and eVldence, rree af senous hlinkers
as they weigh the evídence and think about the likely cansequences of aptiúns. The minima!
commonsens~(al r.equil.ements of rationality in íoreign pollcy decision making expect that
132 JANICf GRDS~ S,;"EiN

poliey makers (an learn frem history. and that they (an drav. some proposltions from the past
and app!y tnese ;Jropositions in an appropriate way to th~ future as they weigh the likely
consequences of the aptions they face (Jervis 1976: Vertzberger 1990: Tetlock and Breslauer
1991; Levy 1994; Tetlock 1998a).
The :;.€cond, more demanding, models Df ratianal choice expeet far more frem decision
makers. Borrowing heavily frem micro-economics, they expect decision makers to generate
subjedive probability estimates of the consequences of the aptions that they consider, to
update these estimates as they consider new evidence, and to maximize their subjective
expected utílity. Rational decision makers choose the aption that promises to give them what
is likelyof greatest value to them. In ather wards ratiana! choosers are reasanably goad esti-
mators of probability and efficient in the choices that they make.2
To put these requirements more formally, theories of ratiana! choice in fareign policy treat
both inltial preferences and expectations as given and exogenous. Models of ratianal choice
are powerful because lhey can idenlify the slrategy lhat le,ders should choose, given lheir
preferences and expectations. They take original preferences as given and specify the optima!
choice. In 50 far as formal models of rabanal choice discuss the process of .choosing. they
assume that people are 'instrumentally rationa!'. Given their existing preferences, people are
expected to engage in an appropriateend-means calculatian. =ormal models of ratianal chaice
do not claim to explain the beliefs and expectatians which le:J.d to choice, and therefore, in a
fundamental sense, leave out most of what is important in explaining foreign policy. Rational
decision makers resolve the conflicts they face in multi-attribute prob!ems by measuring along
a single attribute-that of subjective uti!ity-and simply trading off to find the best outcome.
Rational choice appears to do away with the conflict of choice by measuring along a single
dimenslon. They assume a common yardstick which makes complex measurements simple.
Howwell do these models mirrar processes ofch:)ice in foreign policy? Notwell atar!. There
is by now abundant evidence that foreign policy decisian makers, and people more generally,
rarely meet these standards. This evidence will surprise some readers because, as a species, we
are intuitive causal thinkers and like to think of ourselves as rational; in many ways, a discussion
ofthe deep Jimits to human rationality goes against the grain. The most important evidence of
the limits to rationality comes frem well-established work in psychology. specifically frem the
new, still tentative, research results in neuroscience which are challenging the most funda-
mentai tenets of the rational mode!' The work cf neuroscientists is impo:iant for the ana!ysis
of foreign policy because it is re-introducing conflict as a key ~eature in the choices made by
decision l1akers. What makes the work af psychology and neuroscience even more important
is that the two tend to converge, a facto r af real importance in analyses offoreign policy.

Psychological models: the 'cognitive revolution'

Forty years ago, psycholagists began a 'cognitive revolution' as they rejected simple behav-
iourist models and looked again at how people's thoughl processes shaped the choices they
made. They broughtthe 'mind' back inlo psychology. Althoughlhis was not its main purpose,
the cognitive revolution can be understood largely as a commentary on the limits of rational-
ity. Much af the work accepts rational choice as the default position and then demonstrates
its bounétaries. Research has now accumulated to show that people rarely conform to the
 _ 133

expectations of thE rationa! model (Kahneman et ar. 1982; Hogarth 2nd Goldstein 1996;
Dawes i998). The graunds for pessimism are twofold: the difficulty of rraking inferences (as
mode!s of ratianal choice antlclpate). and the li.-nitations of the human r:lind.
Cognitive psychology' has demonstrated irnpoítant differences bet1Jveen rhe expedéitions of
ratianal decision lTIDdels and the processes of attribution, estimation, and judgement people
frequently use. It explains these djfferencEs by the need ror simple rules of information process-
ing and judgement that are necessary to make sense of uncertain and complex environments.
Human beings have a preference for simpliciry. They are a!so averse to ambiguity and want con~
sistency instead. Further, they misunderstand fundamentalJy the essence of probability (Dawes
1998; Tetlock 2006), making them intuitiveiy poor estimators. Lastly,humans have risk preftles that
depart from models af rational choice; as a result, we are far more aver5e to 1055 than we are gain-
seeking. Together, tnese four attributes compromise the capacity for ratiünal choiee and affect
the decision-making abilities of leaders and officials who are responsible for foreign policy.

Simplicity
Politiealleaders making decisians about the woíld need to arder that world, making its com-
plexities some\tvhat simpler. To do 50, they unconsciously stfip the nuances, context, and sub-
tleties aut of the problems they face in arder to build sim pIe frames. When they 'ook to the
past to learn about the futu~e, political leadeís tend to draw sim pie one-to-one analogies
without qualifying conditions. In 1951, President George Bush cal ed Saddam Hussein
'another Hitler', w:th little attention to what was different either about the twa men or abaui.
lraq in 1990 an.d Germany in 1938. Yetfitting Saddam into an existingframe through use of
analogica.! reasoning-reasoning based on analogy-gave the Presidem: a readily accessible
seript about how:o respond to lraq's irvasion of Kuwait (Khong 1992)_
Drawing argun~ents from a complex historical past is even more challenging. V'Jhen NATO
leaders decided to take over responsibi!ity for militar; operatians in Afghanisté.n from the
USA in 2005, there was a relative\y straightforwc.rd discussion about the importance af pre-
venting 'terrorists' from regaining control of southern Afghanistan. Policy makers were partly
correet. al-Qaeda, now in its third generation, was certainly one ele:nent, but what made
Afghanistan so ehallenging to outside forces wa5 its own indigenous -nsurgency and strong
warlords who coexisted with al-Qaeda. Global terror, local insurgeney, local wadords, and a
local economy fuelled by narcotics interacted together to create cn environment of far
greater complexity than the simple one-dimensional construction of ~he ehallenge that pol-
icy makers used. Their simplified definition af the problem limited toe options that polky
makers then cor,sidel'ed-it both pushed other options off the tablE and blinoed decrsion
makers to some of the likely consequences of lhe optíon they chose to 'fight terror'. We ali
need to simplify a noisy. distracting, and eOll1p\ex env\ronment in arder ia see patterns, but
typically we overslfllplify bad!y_ Politiealleaders are no exception.

Cons;stency
Cognitive psy<:ho\ogists have produced robust evidence that people sirongly prefer
consistency, t:'al they are made uncomfortable by dissonant information, and that they con-
sequently der~'f OI" discount inconsistent inrormation to preserve their be!ieís. This drive for
.134 JAN:CE Gt<GS:S: $Ti::íN

consistency impairs the processes af estimation and judgement. The well-established ten-
dency to di5couni inconsistent information contributes significantly t:J the persistence Df
beliefs. Indeed, exposure to contradictory informatíon frequently results in the strengthening
olbeliels (Anderson et aI. 1980; Anderson 1983; Hlrt and Sherman 1985). People, it seems, are
hard-wired to be conservativ€.
The lengths to which policy makers wíil go to delend polícylorecasts gone wrong are quite
remarkable (Tetlock 1998b). For example, some ofthe strongest propone~tsol regimechange
in Iraq during 2003, when confronted with ihe steep violence four years later, insisted that
they hac been right but that Iraq's leaders simply had not risen to tre challenge. Others
insisted lhat they had been right to support the invasion, but that they had underestimated
the gross incompetence ofthe Bush administration. Poiiticiam on both sides afthe argument
did not revise thelr fundamental arguments when confronted with strongly discrepant evi-
dence, but shifted responsibility to the incompetence 01 others which, they insisted, they
could nol have been expected to foresee. Why they should not have been expected to see the
sectarian divisions among a brutalized population ar the hubris which infected planning for
the war remain5 unexplaíned.
Much af the work of cognitive psychology has been dane i, the laboratory with students;
experts have questioned how wellsuch results travei into the political world. That question has
been largely put to restby a remarkable study of political forecasts in different cultures, where
experts on loreign policy generally continued to delend their forecasts even though what they
expected did not happen (Tetlock 2006). Tetlock identilies se'en categories 01 belief.system
defences: challenging whether the local conditions required by the forecast were satisfied; in-
voking the unexpected occurrence of a snock; invoking a dose-cal! counterfactual-'I almast
got it right'; using an 'off-on timing' argument-'I'm ahead of my time, history will prove me
right'; declaring that internatianal politics 15 hopeless!y indeterminate and consequently un-
predictatle; deliantly asserting that they made the ',ight mistake' and would do it again; and
insisting lhat unlikely things sometimes happen (Tetlock 2006: 129).
Conservative foreign policy makers in the Bush administration implicitly argued that local
conditiors were not satisfied when they blamed lraq's leaders for failing to seize the apportu-
nity that they had been given. The problem was not with their lorecast (i.e. that the removal
01 a regime by force would open the door to democracy) but \Viththe would.be democrats.
A defence that 'alI things were not equal', that an unexpeded shock changed the direction, is
also a cOr1forting defence against error. NATO officials acknowledge that they did not expect
the lerocity of the Talíban insurgency in 2005, but insisted that their fundamental lorecast 01
accelerating reconstrudion and stability in Afghanistan would be proved correet over the
longer termo This defence combines the unexpected shock wlth an argument abouttimingto
rescue the forecasL
Some dthe most intriguing defences by experts claimed that they were almast right. Who
could ha\le expeded Saddam to have got rid Df his weapons programmes, experts insisted.
when they failed to uncover unconventional weapans programmes in post-invasion lraq.
8ut, they insisted, had the USAnot invaded, he undoubtedly would have forged ahead with
nuclear weapons as the international sanctions weakened. The'{would have been vindicated
with the passage of time, they insisted. had the USA not removed Saddam from power. Here
too, expertsdraw on 'close~call counterfactuals' and arguments about timingto preservetheir
beliel syst;ms. Closely related was the argument by those who got it wrong that it was better
 135

to have overestimat2d than underestimated his weapons programme. Better safe than sorry.
they said.
Tetlock confirms the same kind af seíf-serving bias in argumentatian among polítical
experts that cognitive psychologists have documented in the Ie.boratory. He a\50 finds a rela-
tionship between the size ofthe mistakes and the activation af defences. The more confident
experts were ii1 their original forecast, the more threatened they were \l1/len they 'A,ere faced
with disconfirming evidence, and the mDre motivated they were to use on€ or more af the
seven defences to preserve their beliefs. 'Defensive cognitions', Tetlock argues. 'are activated
when iorecasters most need them' (Tetlock 2006: 137). When political experts most needed
to revise theír judgements, they were least open to revision. This same partem was also very
much present among American decision makers during the Vietnam VV2.r.
The €'1idence we have reviewed thus far suggests that belief change i; very difficult. Yet at
times belief s'y'stemsar sehem as do change, at times dramatically. Theories of cognitive -eon-
sistency expect that the least central parts of a belief system, i.e. those \Viih the fewest inter-
dependent cognltions, Vvíl\change first. People will also make the smallest possible change:
they wili ehange their beliefs incrementa!!y, generate a large number of ex.ception5 and spe-
cia! cases, and make superficial alterations rathei than rethink their fundamental assump-
tions. Politi(al \eaders in the USA were generally resistant to changing theír beiiefs about the
Soviet Unio.., after Mikhail Gorbachev came to povv'er. In the face of coumer-evidence, three
years after he became General Secretary, senior policy makers in 'Washington were arguing
that Gorbachev's 'itrategywas to lull the West while the Soviet Union recovered, anel therefore
the USA should continue to ~e both sceptical and cautfous.
Cognitive psychologists suggest some conditions that iacilitate change (jervis 1976: 288-
318; Tetlock 2006:1. There 15evidence that greater change will occur when information arrives
in large batehes r3.ther than bit by bit. Cognitive models suggest that ,:vnen people are con-
fronted with oveíWhelmingly discrepant information, when there is nü other way to account
tor large arnount; oi contradictory data, belieis can change dramatically (jervis 1976: 288-
318). When central belieis do iinally change. they generally trigger iar-reaching changes in
related peripherê.1 beliefs.
Change in belief is also related to content.-Not ali belieis Of expecancies are equivalent:
they vary in strength and in the person's commitment to them. The most common distinction
is between long- and short~term beliefs. Bellefs that have relatively short¥run consequences
are less resistant to change, while bel,efs with longer-term consequences tend to persisto In
foreign policy, the 'inherent bad faith mode!' is an example Df a long-terrn belief that is im~
pervious to new information (Jervis 1976; Stuart and Starr 1982). Long-term belief iri anoth-
er's hostility is easy to confirm and difficult to disprave almast regcrdless of the objective
circumstanc..es. That kind of dynamic is obvious in the enduring con~lict between Israel and
Palestíne. (t ;5 di:Ticult to persuade lsrae\is, snort of a dramatic gesture that 15 irrevocab\e and
self-binding in tle way that PresidentSadat's visit to Israel in 1979 was, that Palestinian inten-
tions have changed_ It is irrpossible to persuade Pa!estinians, confíOn:ed daiiy by occupation.
that Israel-"síntentions CGuld change. Any disconfirming evidence is disccunted to preserve
the consístencyof !ong-term inherent bad-fai-::r, irnages on both sides_ln fOíeign poliey, con-
flict persist5 b~::ause images are 50 diffícult to change. Deeply raated cogniV/€ processes
systematicallYMork against ratianal expeetatíons af apprapriate, diagnostic updating. When
belieís and a'rgurnents do change, they generally change in uneven ways that reflect the
 136 jAr-dU: GP..DSS: STEIN

.•..... -.
-

somewhatarbitrary patterns in the inbrmation and basie processes of attribution. Nelther Df

these has very much to do with ratiana! expectations. However, there Is some evidence that
is more encouraging. It comes from the dose analysis af differences among foreign policy
experts in their willingness to entertain the p05sibiJity that they were wrong. Not ali experts
are resistant to change ali the time.
Drawing on a well-known distinction made by Isaiah Berlin, Tetlock c1assified foreign
policyexperts as 'Ioxes' or 'hedgehogs'. Hedgehogs know 'one big thlng' extremely well and
extend what they know jnto other domains af foreign policy analysis. Foxes, on the other
hand. know many 5mall things, are generally sceptkal af grand overarching schemes, stitch
together explanations with different threads of knowledge, and are sceptical of predietion
in worid politics (Berlin 1997; Kruglanski and Webster 1996: 263-8; Tetlock 2006: 73-5).
Applying lhis to the realm 01 loreign policy analysis, one could suggest that hedgehogs
tend to be deductive generalists in confronting evidence, while faxes are more likely to be
induet ve pragmatists, more Iikely to search for new informatian, and more open to new
information.
The ~vidence shows that the foxes do mueh better at short-term forecasting within their
broad domain of expertise than do hedgehogs. The worst ~erformers were hedgehogs who
made bng-term predictions, usually with cansiderable eonfidence. Hedgehogs are gener-
ally people with strong needs for structure and closure, who are most likely to discount and
dismiss inconsistent evidence when it contradicts their preconceptions. The more know!-
edge hedgehogs have, the better equipped they are to delmd against inconsistency. Foxes,
however, are seeptical of deductive approaches, more likely to qualífy analogies by looking
for disconfirming information, more open to competing arguments, more prone to synthe-
size arguments, more detaehed, and, not surprisingly, more likely to admit they were in
errar and move on. The hallmark cf the foxes is their more balanced style of thinking about
the world. Foxes have 'a style 01 thought that elevates no thought above criticism' (Tetlock,
2006: 88, 118). It is this capacity lar CrIticai sceptical thinking that enhances their capacity
to learn. When foxes are in positions af palltical responsibi'ity, foreign poliey is likely to be
more a::laptive over time. Hedgehogs are far more likely to drive policy in a consistent
direetian.

Poor estimators

People are not intuitive probability thinkers. They depart sYSIematicallylrom what objective
probability ealculations would dictate in the estimates they make. 'Human performance suf-
fers,' argues Tetlock, 'because we are, deep down, determini~tie thinkers with an aversion to
probabilistic strategies that accept the inevitability of erro r' (Tetlock 2006: 40). Foreign policy
experts are no exception. Where we can compare their estinates with those that would be
generated by objective calculations 01 probability, experts do surprisingl'f poorly. Highly edu-
eated specialists in foreign c.ffairs approached only 20% of the ideal across ali exercises
(Tetlock 2006: 77). This is so beca use they think causally rather than pay attention to the
frequencies with which events occur. Experts tend to overestimate the likelihood of war, for
example, pecause they ean easily imagine the causal pathways to war, a highly salient occur~
rence that they have likely studied. They pay less attention to the frequency of wars over an
extended period oftime (Tversky and Kahneman 1983; Koehler 1996: l-53: 293-315).'
 137

To rnake rnatters worse, pro:.able states of the world are very difficult ~o estimate because
we do ílot have rep2ated trials with large numbers in worJd polítics. Foreign policy analysts
generally do not livE in a world of risk, where the probability distributionsare known and they
only have to estimare the likelihoods. Perhaps in internationa! economics and finance and in
global health, analysts know the probability distributions of recessions. deflation, and epidem-
ies. Anal)fsts, even tl€ best infDrmed. do not know the probability af another attack by i11i1i-
tants against civilian infrastructure in the USA orthe UK. There have been toa few such attacks
to generate any reasonable estimate af likelihood. Here foreign policy analysts work in a m-uch
tougher environmEnt, one of uncertainty where they have no access to probability distribu-
tions. This world of uncertainty is one they particula..rly dislike, and it is uoder these conditions
that experts, just like other people, seek the certainty, the false certainty, of arder and control
In this world of Jncertainty, experts search for the relevant categories in which to anchor
their judgements. Cognitive psychology has identif;ed a number of heuristics and biases that
people use in environments of risk and uncertainty that can impair processes of judgement
(Nisbett and Ross 1980; Kahneman et aI. 1982; von Winterfeldt and Edw,:}rds 1986). Heuristics
refer to the rules OI indicators 'JVhichieaders use in arder to test the propositions embedded in
their Q\,'vn schemas. Heuristics help describe how dEcision makers actually process information,
using convenient short cuts ar rules ar thumb. Three of the best-documented heuristics are
availability, representativeness, and anchoring. The avaílabifiry heuristic refers to peopie's ten-
dencyto interpret present informatian in terms af what is mosteasily available in theircognitive
repertoire (Tversky and kah,'teman 1973). Representativenes5 refe;"s to people's proclivity to
exaggerate símilarities between one event and a prior class af events (Jervis 1986~483~S05).
Anchoring refers to the estimatian af a magnitude by picking a:1'available' initial value as a refer-
ence point and then making a comparison (Fiske and Taylor 1984: 250-6, 268-75). (See Box 7.1)
In all three cases, what is c.vailable in a cognitive repertaire has an undue inf!uence on how
an individual infl:..JencesthE likelihood af a future event What we know, and what we can
access, has a disproportionate impact on our forecasts. Anchoring, ôvailability, and repre-
sentativeness were ali at play for years in Britain, where policy makers were like!y to forecast
negative consequences from 'appeasement' because of the salience of Munich in British his-
tor'1'. British decision makers pai a correspondingly less attention to th~ unwanted escalation
that could come irom the threat of force. 'Hedgehogs' who are expert cn what happens 'when
we knuckle under to dictators' are especially likely to discDunt any negative consequences
from the threat of force.
Cognitive biases also lead to seriou5 errors in attribution which can confound policy mak-
ing. People exaggerate the like\ihood that other's actions are thé result of their own prior
behaviou. and overestimate the extent to which they are the target of ~hose actions; cogniti'/e
psychologists call this pattern of attr!';Jution the 'egocentric bias'. One af the most pervasi."e
bíases !s the fundamental attribution error, where people e:xagge'ate the importance of
dispositional over situational factors-in other words, explaining the disliked behaviour of
other,; as a result oftheir dísposition, 'Jvhile explaining their ovm behaviour based on the sítu-
ati orla I constraints that the)' face {Fiske and Taylor 1984: 72-99). When explaining behaviour
tl1at they like. people simply reverse the partem of inference. VVhen the government of North
Korea rnakes a.concessíon in the talks about its nuclear prograrnme, analysts in Washington
see tha\. concession as a function of the constraints Pyongyang faces, but explain their a'Nn
willingness tb ~articipate in the talks as evidence of their search rOr?, peaceful compromise.
138

BOX7.T Heumficprõcessing' .:' - . ~

Heuristics, those rules cr indicators that serve as cognitive 'short [Uts' ar 'rules ofthumb', often play into
decision making without us being aware af it. The following study by Kahneman and Tver.sky (1983)
rellealed the common use of heuristic processing. How would you answer this questionnaire?

The Linda problem

Linda is 31 yea~s old, single, outspoken, and very bright. She majored in philosophy_As a student,
sh2"was deeply concerned with issues af discrimination and sodal justice. and also participated in
antinuclear demonstrations_
Please rank the following statements by their probability, using 1 for the mos! probable and 8 for.the-
te2..St probable.

(a) Linda is a teacher in a prímary school.

(b) Linda works in a bookstore and takes Voga classes.

(c) linda is an active feminist

(d) linda is a psychiatric social worker.

(e) linda is a memb~r ofWomenAgainst Rape.

(f) Linda is a bank teller.

{g) linda is an insurance saf't'!sperson.

(h) linda is a bank teller and is an adivefeminist.

The results of the study revealed that8S% of respondents indicated that linda was less likel'l to be a bank .
telfer than both a bank teller and a feminist. Accordíng to the represenlativeness heuristic. s~bjects make
thér judgements according ~o the perceived similarity between the statement and the description af
linJa.

Kah.leman. D. and Tversky. A. (1983), .Extensio~al versus lntuition Reasoning: Conjunçtion FaUacy in Probability
Jud&ment'. PsychologicaJ Review, 90: 293~31. Reprinted with permission.

When North Korea (uts off the talks, it is doing 50, analysts argue, because of its determina~
tion to acquire nuclear weapons in violation af the rules. A.tough response to North Karea tS
dictçted by the constraints that Western powers face. The double standard in reasoning is
glarilgly obvious.
To make matters worse. foreign policy experts sufferec frem the dasstc hindstght bias, in
whích they systematicaily misremembered what they haó predicted. They c1aimeó that they
had assígned a higher probability to autcomes that did h3.ppen, and gave less credit to their
opponents for getting it ríght than they deserved. The strength of the 'hindsight bias' was strík-
ing because these experts knew that their original judger.1ents had been recarded and that
thes€ records were accessible. How (an we explain this kind af bias in the face af evidence
that these expe:ts knew existed? Tetlock argues as follows

A fuller explanation mus( trace hindsight bias to a deeper cause capable af producing genu-
ine self-deception: the largely unconscious cognitive pro::essing that is automatically acti-
vate.d. _that allows us to rapidly assimilatethe observed outcome loto our network af beliefs
ab3'ut what makes things happen. People manage to convince themselves, sometimeswithin
m~liseconós. that they 'knew it ali along' (Tetlock 2006: 139-40).
 139

Loss averslon
FOí"eign policy dec:sion makers, like peopie generally. are not neutra! about risk. Cogni-
tive psychology ha; generated robust evidence that loss is more painful than comparable
gain i5 pleasant, ano that people prefer an immediate smõ.ller gain rather than taking a
chance on a larger longer-term reward (Kahneman and Tversky 1979, 2000; Tversky and
Kahnernan 1992). People systernatícally overvalue lasses relative to comparable gains_
These proposítions about risk have held up ,acrass a wide variety of cultures and
situations.
The irnpact of los5 aversion on foreign policy decision making is considerable. Leaders
l:€nd to De risk averse when things are going well and relatively risk ac:eptant when things
are going badIY-iJl,Ihenthey face a crísis in which they are likely to 1052or have lost som€"-
thíng that matters to them. Leaders are al50 likely to take greater risk 10 prateet what they
already have-the 'endowrnent effect'-than to increase their gains. They are also likely to
take greater risk to reverse losse5, to recapture what they once held, than they would to make
n€w gains. And when decision makers suffer a significaot 1055,they arefar slowerto accom-
modatE' to these lessEs than they would be to incorporate gaíns. Fina!ly,leaders rEverse their
preferences and r1ake different choices when problems are reframed as losses ratheí thar.
gai ns_
These general flndlngs apply direetly to foreign policy ehoiees. President Sadat of Egypt, ror
example. never 'n:Jrmalized' for the loss of the Sinai to Israel in 1967. Even though Israel had
an obvious advantage in military capabilities, Sadat was undeterred anj, highly motivated to
reeapture the Sinai, he designed around Israel's military 5trengths and Jaunched a war in 1973
(5tein 1985, 1996). Under these kinds of eonditions, theories of rational deterrence whieh do
not systema{ieally build in aversion to 1055are likelyto mislead foreign poliey decision rnaker5
by generating unjue confidence.
This review of the need for simplicity and consistency, the impediments to probabilistie
thinking, and the predisposition [Q 1055aversion are often treated. as we saw €3.rlier in this
chapter, as deviations frem rabonal madels of information processing, estimation, and ehoice.
Rationa! choice remains the default and these 'deviations' are treated as limiting conditions.
I dis3gree. These'deviation5' are 50 pervasive and 50 systematic that it:s a mistake to con5ider
rational models Dfchoice as empirica'!y valid in foreign po!iey analysis.

Neuroscience, emotion, and computation

We are at the edge of an exciting revolution in our understanding or the numan brain and
how it \'\Iorks. t\ew imaging technology is allowing scientists for the first time to 'watch' the
brain as it think5, feels, remembers, and chooses; the pictures they are seeing are revolution-
iz"lngunderstandings of thought and decision. Even though the experimental results are stil[
far from definitiv€, two results stana out. First, many deeisions seem not to be the result of a
deliberativethought process, but precoJlseious neurological proceS5-c2s. The brain (an absorb
about eie'venJ:ni!\ion pieces of information a second, but can only process forty conscioL!s!y
The unconscio'.1s brain manages ~he resto Seeond, many decisions se-emto be the product of
strong ernori'onal responses. (See Figure 7 ..1.)
Jt..NlCE GROSS STEIN

Side view

Pmtenur
paáttal
corto:
Anterior U'n.mtopo{ar)
prefron:e.1 corta;

Insular cortex
(d~~pbetweenth~fromal
and tempora: lobes) Temporallobe

Midline view

Orbitofrontal Striatum
cortex Amygdala (deep tO the surface
(along the mediai surface ofthe ffildline)
ofthe temporallobe)
Notes: Lateral (side) and mediai (midli:le) views af the hUnlan cerebrum, identifying area'i critical1y
associate<! with decision making. Areas in bold have consistently been associated with emotional
processing, while areas in itali.::shave consistently been assaciated with higher-level cognitive
processes
Figure 7.1 The human eerebrum
Source: ':ohen,j. (200S), 'The Vulcanization ofthe Human Brain: A Neural PerspE'ctive on Interactions Bet"ween Cognition
and Emotion',]ournal cf Economic Pe.-spectives, 19: 9. Reproduced by perrllission of Jonathan D. Cahen, Department of
Psychology and Princeton Neuroscience lnstitute, Princetan Universit)o,_

Scientists now make a very strong, and startling, c1aim. There is growing consensus that
emotion is 'first', because it is automatic and fast, and that it plays a dominant role in shaping
behaviour. We know now that emotion operates in part telow the threshold af conscious
awareness (LeDoux 1996; Winkielman and Berridge 2004). Contraryto conventional wisdom,
we generally feel before we think and, what is even more surprising, we often ad before we
think. -here is widespread consensus that the brain implements 'automatic processes' which
are fas:er than consciaus deliberations with little ar no awareness ar feeling af effort (Bargh
et a/1996; Bargh and Chartrand 1999). Not surprisingly, the conscious brain then interprets
behavi.:lur that emerges from automatic, affective, processes as the Dutcome af cagnitive
deliberations (Camerer et aI. 2005: 26).
A useful way af thinking about emotion and cognition is to see affective processes as those
that address the ga-no-go questions, the questions that motivate approach-avoidance, while
cogniti'J.e processes are those that answertrue-false questions (Camerer etal. 2005: 18Zajonc
1980,1984,1998). Choice c!early invokes both kinds 01 processes. Establishing truth c1aims
about ~ates of the world is usually not enough for people to make a choice. What matters to
me, what I value, is an emotíonaJ as well as a cog:litive pro,-ess, and is importan: in what J

decide to do, whet1er I go, ar I don't go, whether I approach ar \ avoid. Whether ar not I am
treated fairly is an emolíonal as well as a cognitive judgement and, in this sense, emotion
carries utility.
How do neuroscien1ists ar',aiyse ematian and cognition? Some conceive af two separate
operating systems in the bíain: emotion and reason.£ 'Emotlon:; influence our decisions,' argues
Jonathan Cohen. 'They do 50 in justabout everywalk af our !ives, '"vhetherwe are aW3.reOí una-
ware of it, and whether we acknowledge it ar not' (Cahen 2005: 1). Errotions are automatic
processes associated v\'ith strong positive ar negative response. The brain, Cohen explains, has
different kinds of mechanislTIs: one, which includes emotional respanses, can respond auto-
maticaliy, quickly, and defjnitively but is relatively inflexible; cognition is less rapid and ha~
limited .::apaclty but is more flexible. There is a trade-off between the speed and specializatior
af emotions and the generality af reflection. In the circumstances of mocern Iife, these systems
may prescribe different responses and the outcome of this competition determ'nes choice
(Camerer et aI. 2005). This br3.nch of neuroeconomics has explicitly brought conflict bà.ck into
decisian making. Choice is a conf1ict berween emotion and c:>mputation. (See Box 7.2.)
Kahneman calls the firs!. emotion-based, system af decision making 'intuitive' a.,d 'as50cia~
tive', and the secold system of cognitive decision rnaking 'reasoned' and 'rule-governed'.~ The
fiíst system is preconscious, autornatic, fas!' effartless, associative, unreflective, and slow to
change. The second system is conscious, s!ow, effortful, reflective, rule-governed, and flexible.
The vast majority of decisions are made rhrough the first sysIem, which draw5 heavily an
emotions and in a c01l1petitíon between the two always tru:nps the rule-governed reasoned
system_ lt is extraordinarily difficult, Kahneman concludes, for the second system to educate
the first.
The well-knoVvn 'ultimatum game' highlights the camputatianal, cognitive, and emotional
elements at play in decision making (Figure 7.2). The game comes oul of econamics but has
relevance to international politics as well. One partner has access to a given resource-wealth,

taox 7.2 Scientists create 'ftu5t potton' ,

Scientists create 'trust potion'

A key hormone helps determine whether we will trust lovers, friends. or busines5 contacts, scientists
elaim. Exposure to an oxytocin 'potioo' led people to be more trusting, tests by Lniversity of Zurieh
rese",rehers founG. They repcirt in the jO'.Jmal Nature that the finding could help people with conditions
5uch as autism, lNhere relating to others ean be a problem. Sut aoe expert warnE<Í itcoutd be misused
by po!iticians who want to pe~suade more people to back them.
'Some may worry about the prospect that political operators will generousty spray ;:he crowd with
ox~rtDcinat raUies of their c2odidates' (Dr Antonio Damasio, University of lowa College of Medicine).
O:<ytocin is a molecule produced naturaHy in the hypothalamus area. of the brain which rEgulates a
va.riety of physiological processes, including emotion.

Storl' rrom BSC NE'NS:

h~ p:i /news.bbc(D.ukJgo/pr/frí -;'1ih I/ i1eailh/4 599 299 .s,m

~'ub1ished' 2ooSro6f02 6 BSC Mr,~v

142 jAN~('E GROSS STEIN

vast natural resources. high!y sophisticated military technology-and (an propose how the
reSQUfCeshould be split. If the other party accepts the proposal, then the resource is divided
as they have agreed. Ifthe split i:i rejected, neither receíves anythingand the game is over. and
eonfliel probably follows.
Rationa!ly, the second party spould acceptanythingthat is offered, because anythíng is c1early
better lhan nOlhing. And again, lhe firsl party, knowing lhal for lhe alher anything is better lhan
nolhing, shauld ralionally offer as Iittle as possible. The game ras been played aerass a wide
range ofsituations and cultures, with a remarkably consistent outC:Jm€. Contraryto what rational
models would expect. offers af less than 20% of the total are generally rejected out af hand. Why?
Perhaps th:Jse who rejected the offer were worried about their bargaining reputation forthe next
raund, as ratianal delerrenee lheory says lhey should. Bul players responded lhe same way even
when reputational effects were removed frem consideration, when they were told that they
would play lhe game only onee. When asked why lhey reJeeted an offer lhal would give them
somelhing, players responded that the offer was humilialing, insulling, patenlly unfair. They
responded quickly and intuitive!y to an offer which gave them something, but humiliated them
in the process.6 Their rejection was driven by a strong negative emotional response.

According to this experimental economics game, first studied by Guth et ai.

(1982), two players intera't anonymously and only ance, 50 reciprocation is not an
issue. The first pleyer proposes how to divide a surn úf money with the second player. The
second player can either accept ar reject the proposed diviSlon: if the second player accepts
the divisian, the first player gets his demand and the second playe! gets the rest; on the other
hand, if the second player rejects the divis.ion, neither player receives anything.

Illustratively this game can be demonstrated by the example belO\'f:

Player 1 propases
division af $10:
..•£9 for Player 1
* $1 for Player 2

Plarer -2 (nust
P~ayer i' Accepts accept ar reject
_.- the_propo_saI

P~yer 1 receives $9 Neither p~ayer .

Player_2 receives $1 receive$ ánytbing

Rationality would suggest lhat the second player accept whatever division is proposed,
because anything 1Sbetter than nothing. However, consistently, th~ second plarer rejects the
propcsed division when offers are less than twenty per cent af the total. Why? Emotional
reasoos: the second player rends to find such low offers 'humiliating' > 'insulting', and 'unfarr'.
thereby denying both players anything (Coheo 2005: 13-15).

Figure 7.2 The 'uttimatum game': how would Vou choose7

50urce: The ne 0tiation style known as the 'ultimatum game' was intraduced by experimental economists Güth. W,
9
Schmittberge~ R., and Schwarze. B. (1982), An EKperimental Analysls af Ultim~tum Ba(gaining',Jaum~1 af Economic Behavior
'fOrganizatio,1, 3: 367-88. t.-:: Elsevier 1982.
 143

Ernotion is an individuai embodied experience. How canthe study af emotion be extenaed

to lhe ana!ysis offor~ign policy and to decision makingwhich, €ven when decisions are made
by individual leaders, takes .place in a collective context on behalf of others? Emotions are
meaníngful in 2. social context (SaUí€tte 2006: 507-8)? It is only \'vith a shared sense af v•.hat
constitutes appropriate social behaviour, for example, that a person, a people, ar a govern-
mE:nt feels humiliated at ali. When the flag of one nation 15burned by ano~her, the humiliation
and anger that follovvs flov.' trom a shared understandingthat i:he burning of a flag is a delib-
erately insulting and hostile act. Physiological processes are layered by social knowledge
which shapes the a?propriateness af anger, fear, and happíness. It is in this sense that emo-
tiOIlS need to be conceived as not only an individual but a150 a social process (ROS5 2006).
Scholars in internationa! relations 2re nQ\,~Jbeginning to look carefully a~the neuroscience af
emotion, Research on emotion ;s having a significar.t impact on the analysis of a wide range of
global issues: the logic 01deterrence (Mereer 2005,2010), cooperation and the solution 01col-
leetlve action proolems (Rillinget ai 2002: Mercer 2005), nuclear prolileration (Hymans 2006):
the War 011Terror (3ennett 2002: Saurette 2006: Bleiker and HutchlSol1 2J08: Cra'Nlord 2009),
,€I.'enge, anger, and numiliation as motives forwar (::;ries 2004: Saurette 2006; Lbwenheim and
Heimann 2008), ard patterns of ethnic and civil conflict (Kaufman 2001), conflict resolution,
and pos~-conf!ict recollc\li.ation (Edkins 2003; Hutchison and Bleiker 20(8)""":all issues that are
central to foreign poliey decision makers in the twenty-firsr ce,'lury.
Psychologlcal models have long inlonned the study 01 deterrel1ce (jer,is et ai. 1985; Leoow
and Stein 1994), but building emotions into the explanation is shedding new light on old prob~
lems. The credibility of signa!s, an essentía! component in theories of deterrence, compellence,
and bMgaining, is rot only a property ofthe sender, as some formal models of signalling suggest,
but also a function of the beliefs af the reeeiver (Mercer 201 O). These beliefs are not oniy cognitive
but are emotional aswelL The emotiona! cues that signals evoke-fear, an5er-matter in 50 far as
these emotions thEn prompt beliefs and aetion in turno Researchdemonstrates that fear prompt::
uneer"i.ainty and ri::k-averse action, whlle anger prompts certainty and risk aeceptance. Threats
that evoke fear, unless they evoke 1055 avoidance, are likely to prompt hesitaney and a rlsk-averse
response; indeed, that is th12purpose af mos! deterrent threats. However, frightening threats are
less !ikely to be successful when they are designed to eompe! adve,sarialleaders to ae!.

Conclusion
We are at a hinge moment in the development of theof)l and evidenee on deeision making. fv1uch
that has been taken for granted is now being questioned as new a'Jenuesaf research and theory
apeno The work that this chapter has reviewed looks largeiy at individualdecislon making, and It
is a blg step from individual to eolleetive decision making. 'vVeneed to bevery careful in general-
Izing f,:Jn'l the ind.vidual to lhe co!lective; indeed many beiieve lhat il is inappropría:e to do 50.

ir is a well-established f2.llaey to give the properties af an individual to a colleetilJe entity Bu-

re2.L'craciesand s~atesdo not think Of feel-the individuais w:thin tnem do. A governlllent, as a
cDllective entity, does notcalculate and make ratiana! choices-poiicy makers do.lt is impossib.e
to agg-egate ind.ividual preferenc€s Vv'iIÍ1outsignificant dlstortioll. As olheI' ehapters in th:s vo!-
urnr- iook expli(itly at moóels af colleetive decislon making I:seeChapters Six and Fourte'2n), i~
rernaíns to asl<"hEre:How impo(tant are individuai models of choice in fJreign policy analysis?
144 }At...li(E GROSS $TE1N

P(esidential systems often give great scope to the choice of the president who seeks
advic= ano delegates authority, but reserves to him/herself the central decisions in fo(-
eign policy. Here, individual models of choice are very relevant. Even in parliament2ry
systems, prime ministers increasingly make decisions after consulting a handful of advis-
ers, and only after they have decided do they bring the issue to cabinet (Stein and Lang
2007). Here too, individual rnodels of decisior, making are relevant. li is in the framing of
decisions and in the implementation phase that more cDmplex models are nec€ssary. In
other words, what comes before and after choice often constricts the space for
decisions.
A second question grows out of the research this chapter has examined. There is strong ro-
bust evidence that most human choice is preconscious and strongly ar.d quickly influenced by
emotion. How relevant is this evidenceto public choice in foreign policy?We cannotyetanswer
this question with confidence, but the survey of foreign polícy analysts suggests ~hat the more
general patterns we have identified are present in foreign pJlicy decision making as wel1 (Tet-
lock 2006). Despite their expertise, foreign policy makers are no less biased than other people.
lfw~assume that foreign policy makers do not deviate significantly from these g~neral pat-
terns, then we face a third questiono Does this new evidence about cheice constrain what we
can do to improvethe quality of decision making in foreign policy? That is the power of mod-
eis of rational choice-they set norms and standards. The n€'N research should nct discourage
foreign policy analysts from efforts at improving decision making, but these efforts might
need a different focus.
The pioneering work of Damasio (1994), LeDoux (1996). and Panksepp (1998) joined 10-
gether emotion and reason. The revolutionary impact of Damasio's research was to make it
clear that peaple with impaired emotiona! capacity cannot make rational choices. Reason
presupposes emotion. At the same time, the conscious brain can conditian preconscious
choices after the fact. In other words, there is an opportunity to prime neural responses, to
educa~e our brahs for the next time::-Cohen argues that the brain has been 'vulcanized'; just
as rubber is treated with a substance to improve its strength, resiJiency, and usefulness, so
neurological systems leave roam for reason over time (Cohen 2005: 19). The capacity to
reasor and reflect protects people against impulsivity, over time. When these protections
are not adequate, social and.technological innovations can help to prateet people against
lhemselves. For example, people discounl the fulure badly and consequenlly do nol save
enaugh for their retirement. When the default was changed' so that people needed to opt
out of enforced savings which were deducted from their monthly pay. savrngs for retirement
grew dramatically. Neuroeconomists call this 'Iibertarian' ar 'asymmetric' paternalism; liber-
tarian because people still have choice. but paternalistic because default aptions are set to
produce individual and colleclive benefil (Camerer et aI. 2003). 11is nol hard to imagine
extending these kinds of aptions to foreign policies which deal with the environment,
health, and arms contra L
The new research emergi:1g from neuroscience does not eliminate the possibility of learn-
ing and change. Reflectian may come after choice, but it prepares decisian makers for the
nex! decision. The challenge is to understand far better how and when emotions are engaged,
when they improve decisions, and howemotions engage with refledion and reasoning (Och-
sner and,Gross 2005). Neuroscience research on choice cen:ainly does not preclude priming
the unc'tmscious mind through both repeti tive patterning and systems design, what some of
have ealled the 'free won't' (as distinct from 'free will'). In this context, research an learning and
 145

change, !argely neglected. becomes especially important in foreign policy analysis (Tetlock
and Breslaue, 1991; Levy 1994: Ste; n 1994).
R?tional models af decision making can be used inthree important ways. Fjrst, they are
useful as an aspiraton, or a ;lerm, but only with the fu!1 realization that fcreign poltcy decision
makers u€ unlikely everto meet that norm. They a~e no! descriptive theDries of choice (Elster
1936: 1). Second, the creative use af rational choice models has uncQ\'ered many counter-
lntuitive and non-obvious p3radoxes and traps that- can be very instructive to those who
make declsions. Knowing the traps ar games af Prjsoner's Dilemma and the risks ar Chicken
can serve as salutéry warnings for those who, for example, must make life-and-death de(Í-
sions about when and how to use mllitary force. Final!y', rational choice models can be de-
signed into the small cracks that neuroscience apens up and highlighlsas spaces that can be
used to correct some afthe worst biases in decislon making.
leaders who are aware ofthe dynamics of choice can build in compulsory delays and waít-
ing periods which leave time ror refleetion and analysis_ They can de5ign the equivalent of
autamatic check-offs in fore'gn paliey by putting in plaee systems of cheeklists before deci-
sions are finalized. They can seek ou! their critics and ex pose themsel'vEs, quite deliberately,
to counter-argurnents. They can demand worst-case scenarios and anticipate how they
\o'Jouldíespond were the.wor'sI to happen. They can insist on verificatiDn of the information
that they think is most important to -::h=irdecision. leaders can design the open spaces that
they need ror deliberation, ;"eflection, and analysis once they acknowledge their impulsive
ernotion-driven choice. It is precisely in these apen spaces that foreign paliey ana\ysis, work-
ing with the dynallics of choice, can make its most significallt contribLtion.

Ü Keypoints
"" Rational models have poor empirical validity. People rarely conform to tl€ expeetations of a
ratlona! model, which suggests there is something beyond rationality th;:.texplains human choices

c CognitivE psycholog}' reveals tha.t people prefer simplicity and consisten:y, making them poor
estimators; people are also more allerse to 10s$than they are gain-seeking.
5- Palitiealleaders uncor.sciously make use of analogical reasoning to simr:lify a eomplex
environment, at the fisk of oversimplifyingthe situatiaíl.
~", Foreign policy exper~5have been dassified as 'falies' ar 'hedgehogs'. Hedgehogs know 'one big
thing' and are poar at rmking long-tenn predietians; they are more Iike1tta drive paliey in a
consistent direetion E'venwhen they are wrong. Foxes, on the other hand, know 'many srnall things'
and are particularly strong at providing short-term foreeasts: they are more likely to drive paliey in
a more dynamic, adaptivf~ directicn.
e Many dfciSlons seem ,o h~ the resuit af strong ernational respo!1ses.Neuroscienee enabies
us to 'see' how the br.::.inreache-~c-ertain forms of behaviour. therefore 1l1aklngthe work of
neuroscientists an líllp0r"lilnt resOLlrcefor the al1al'fsis of foreign policy.

Ü Questíons
1_ Whenare rationai cnoiee models inadequate7
2. Wha~dces cognitive psyellology tell us about human behaviúur that '$ relevant to foreign polq'
analysis?
jANfCE GROSS STEIN

3. Do Vou knowing!y engasE' i" heuristic processing' Give examp!es of how heuristics affects your deci-
sion making_

4. Do Vou consider yourself to be a 'hedgehog' cr a 'fax'? What are the advantages of being one or the
other if yOl. are in a position of politicalleadership? (an Vou dentíty 'hedgehogs' and 'taxes' among
today's leaders?

5. Why are analysts offoreign policy looking to the work of neurü>cientists? W'lat does rhe :Jhysiology Df
the brain reveal about decision making?
6. How does human ematian affeetforeign policy decision makiflg? Giveexamples Df how emotion can
positively and negatively determine response in decision making.

7. How can we proceed in foreign policy analysis knowlng that éecision making is often dane at a pre-
conscious leveP

8. What can we do to improve the quality of decision making in foreign policy?

~ Further reading
Goldstein, E.B. (2007), Cognitive Psychology: Connecting Mind, Researeh and Eve()lday Experienee
(2nd edn) (Belmant, CA: Wadsworth).
This book introduces cognitive psychology and the theories of cognition, with a final chapter on
'reasoning and decision-making'.

HOOSOfl,N. &iid;Sincer;~992)Politkaf Psychofogy and Forei9n Po/icy (Boulder, co: Westview

Press).
His book discusses the irnpact of psycholagical processes on for~ign policy decision making.

McOermott. R. (2004), 'The Feeling of Rationality: The Meaning of Neuroscientifk Advances for
Political Science', Perspect;ves on Politics, 2: 691-706.
Ths article discusses the uses of neuroscience in understanding cecision mabng in politics.

Mintz, A. (ed.) (2003), lntegrating Cognitive and Rational Theories af Foreign Pa/iey Decision-
Making (Basingstoke: Palgrave Macmillan).
This edited volume draws hom both cognitive and r••.tionalist ideil.Sto examine how foreign policy
decisions are made, using case studies and experimental analysis

VEtiv, S.A. (2004), Explaining Fore;gn Policy: U.S. Decision.Making and the Persian GulfWar
(Baltimore, MD:Johns Hopkins University Press).
In this book a theoretical framework on foreign policy decision-making is presented and tested using
the Persian Gulf War of 1991 as a case study_

Zak, P J. (2004), 'Neuroeconomics', Philosophical Transactions af the Royal 50ciety af London

(Biological Sciences), 359: 1737-48.
Tms paper introduces the emerging transdisciplínary field known as 'neuroeconomics', which uses
neJroscientific measurement to examine decísion making. lhe role of emotions is exarnined in social
and strategic decísion-making settings

Virit the Online ResOLlrceCentre that accompanies this book for more information:
www.oxfordtextbooks.co.uk/orc/smithJoreigo/
e Implementation
and behaviour
ELlSABETTA BRIGHI AND CHRISTOPHER HILL

Chapter contents
Introduction
When actors meet their environm-ent-theoretical issues 1'13
Exerting influence 157

Tlle practical importaoce of context 1 S8

The instrllments cf foreígn policy

Condusion

o Reader's guide
This chaptel" looks at what is known as the 'implementatian phase' af foreign pclicy
making. i_E'.the period in which decisions are translated intoaction. Implementatlon
can lead to very big problems-and surprises-fordecision makers, whos€ intentions
arten get left behind in the comp!e}ities ar practice
In arder to throw light on why this should be 50 we first ar a!1look at the ~heoretical
problems involved in deciding where a foreigp paliey action ends and its environ-
m€flt begins, for the difference berween the 'inside' af state decision making. and the
'autside' woíld af ímernational relations is by no means as dear as has been tradition-
ally supposed. We theo go 00 to illustrate the variety of pr::lblerns which states en-
counter when trying to implement their breign policies, an.d the [Cinge af
instruments-díploma,ic. militar)', economic. and cultural-which are availa~le to
them. The chapter finishes by outlining the endless loops which connect-and blur
together-ends and means in foreígn policy. and by identify,ng the key les50ns which
practitlor.ers need to keep in mind.1

lntroduetion
The phase of implementation is that in which actors confronttheir environment and in which,
in turn, the environment canfronts them. In essence, this phase lmplies an interactil.r'e strate-
gic process which is very ofteo important when it comes to translating foreign policy obJ'2c-
tives into practice, and decíSive when itcomes to turning practice inta cesired outcomes. Thís
fírst half of this chapter will be dedicated to exam:ning some of the mast typical features and
diiemmas of thé phase of implementatlan. This phase req;Jires crossing the baundary be-
iween actors ;nd the outsde v%rld, íf outcomes are to be shaped on the basis af stated
.. 148 EUSAEETTA BRIGP,! .t<.f-lt) CHRISTOPHER l-HlL.

objectives. The second half af the chapter wHl look at the practical cheices QV€f means and
medes through which states conduct foreign policy. The e:<ercise af channelling intentions
jnto Dutcomes, via the use cf instruments, is complex and rarely a mere technicality; indeed,
it has the power to change foreign policy in the processo
By way cf introduetion to this set af issues, the chaptÊr will fjrstly look at theory and
presen: some general remarks on the issue af how to conceptualize foreign policy imple-
mentation as a form af strategic and dialectie interplay between a foreign policy actor and
its environment. As we shall see, the essential insight af whc.t will be presented here is that a
succes~ful implementation of any foreign policy depends not only on a clear definition of
objecti'/es and a sound choice of instruments (issues whicr will be dealt with in the second
half of the chapler), bul also, and ralher crucially, on the interplay belween lhe actors stral-
egy and the context surrounding it. Accordingly, it also depends on the actor's ability to ad~
just to unforeseen circumstances. The second section will move on to examine in more
delail whal we mean by conlext when dealing wilh forelgn policy; in doing 50, we will
present a pieture of the 'jnternatianal' seen frorr, the persrective of the actar. This wil~ in-
volve two steps: on the one hand, we will draw different pietures of the 'perimeter' of the
'internalional' (from lhe regional lo the global); on lhe olher, we willlook aI lhe many di-
mensions of which the 'international' is made ano explore their interconnectedness. Thirdly
and finally, we will shift the perspective to consider implementation from the paint of view
of both the actor and the context, and will foeus on some of the dilemmas and synergies
inherent in the process of connecting the 'damestic' and the 'international' while pursuing
one's foreign pollcy objectives.lmplementation thus emerges as a eomplex and fulfy politi-
cal activity; a 'boundary' process which connects actors to their environments via the pursuit
of foreign policy.

When actors meet their environment-theoretical issues

The issue of how social and political actors pursue courses of actian and, through aetions,
succeed in attainingtheir objectives is a conundrum of intere5t notjust to foreign policy ana-
Iysts, but to ali social scientists. How is it that sometimes even the best laid plans do not suc-
ceed in achieving one's goals? And conversely, what does it take to turn situations to one's
own advantage? These puzzles confront foreign policy makers daily in their efforts to project
lheir counlry's inleresls and goals abroad, and go lOlhe heartofthe 'problem' of implementa-
lion in foreign policy. Hardly a lechnicality, implemenlatlon is in faet a fully polilical aclivily,
not leas1in the sense of reflecting a c1ash Df wills between different actors, ar between aetors
and their often intractable environment.
Despitethe rather inchoate Iiterature which has developed 3.round it, the best plé.ce to start
for considering the question of foreign policy behaviour and implementation fram a theo-
retical poJint of view remains the so-called 'ageney-structure debate' (inter afia Wendt 1987;
Hollis and Smilh 1991; Carlsnaes 1992; Wlght 2006). Atits mosl basic levei, lhe debate con-
cerns the vexing question of whether action can be explaine:J from the 'inside' af actors, ar
from the 'outside'. 15it possible to find the roats af actions in the actor's preferences, interests,
and mean{ngs,
, . or is it rather the externai context, constraints, and patterns which steer aetors
in certain directions and not in others?
 149

As some of the most COl1lpelling literature in foreign po\icy anaiysis has nO\l1/ made ciear,
foreign policy (an be fully considered as a form of adian {Carlsnaes 1989); lndeed, foreign
poiicy is an important site af política! agency in contemporary v,;orld polítics (Hill 2(03)- In
this sense, the agency-structure debate does have something to say aboJt foreign policy.
and in fact it has interesting things to say primarily about the phase (ir behaviour and
implementation
Consider the following counterfactual. If an actor {for instance a state) existed in perfect
isolation (o:'",alternatiliely, if it \I\'ere all-powerful). it v•.•.
culd have surely no prJblem in translat-
ing intentions, motivations, anc desires into objectives. lndeed, objectives and oulcomes
would be practically the same: the process af implementation \I'..fQuldbe quite 5mooth, either
becau5e it would be accomplished in a vacuum, ar because the aetor would be ful\y in contrai
of the eontext, able te manipulate it at will.
World polities, hawever, hardly resembles this picture_ The international scene is made up af
aclors, states, and non-states, each with their own set af interes-:s, objeetives, and pri'Jrities-
not necessarily in cOlfliet, but very often distinct fro-m one another. Therefore, for all but
the most powenul acors, a degree of resistanee is Qound to be encountered in the pr:xEss af
'having one's own way' in the system with the intent to produce desired outcomes. Further,
even the most povllerful actoís might not be in the position to manipulate the environment
around them fully. beca use of ei(her failures of judgenent ar dísadvantageous asymmei.ries ir.
other important dimensions besides that af power (e.g. informa:::íon or legitimaey).
How can one come to an e,egant formalization or the set of issues and processes with
which aetors on the international scene are eonfronted when trying to implemenl their ob-
jectlves, thus producing foreigl poliey behaviour? The argument advanced here is tl1at in
orderto eoneeptualiz.e behaviour and implementation, foreign poliey analysis needs to adopt
a strategic-relational approach (for the original statement of the model, 5ee Hay (1995,
2002), for a full appiication to 'oreign policy, see Brighi (2005)). The intuition aI lhe heart of
the strategic-relational model-a type of systems approach-is that foreign policy behaviour
is produeed via a dialeetic interplay betw'een the aetor's own strategy on the one hand and
context on the other hand. The approaeh is ealled strategie in that aetors are conceptuaíized
as oriented towards the attainment of stated goals. Furthermore, in the process ofelaborating
appropriate eourses of aetion, actors inevitably have to take ioto aceount the strategies of ali
other players. It is also relational beca use it assumes that aetors and their behaviour only
become intelligible when analysed in relat!on to t:,eir pro per contexto In tum, the context
only becomes truly 'real' when looked at from the perspective of the indh'idual aetor in ques-
tion; therefore it always exists in relation to sameth;ng. ar some other aetar.
The strategic-relational mode! was first introduced in politícal science.n arder to reject the
vievlI that (polítical) action could be reduced to either external constraints ar internai preíer~
ences. lf it is reasonable to assume that both elements are in play most of the timé, what be-
comes interesting is to invEstigate how constraims ano preferenees interact, s:Jmetimes
clashing and sametimes producing virtuous synergies
If one applies th:s approach to foreign policy, eertain aspe:ls af implementation become
intelligible. Firstly, ~he strategic-relational approach tells us that neitherstrategy nor eontext
taken in isolation can explain the sueeESSar failure of a ce~tain foreigr, policy to deiiver an
íntended outcame. An exclusive foeu5 on the domestic politieal prceess cannm Explain
those ir.stances in which OULComesdel/iate from intentions {which 15 the rule ra.ther thar.
1 SO EUSABETTA !3RIGHI AND CHRISTOPHER k!tt

\
Aetor
'/
ldeas
lê
~

-----'Polirica!
h" I,
foreignPoliev
, \. ~process BehaviollT
Context "---~

Figure 8.1 The strategic-relational approach to foreign policy (Brighi 2005).

the exception). Conversely, an exclusíve focus on context places toa much emphasis on
the constraints and opportunities shaping actian, and cannat convey any real sense -Df
intentionality.
Aschematie illustration ofthe model is provided in Figure 8.1.
In thinking about this model as applied to foreign paliey implementation and behaviour,
three considerations seem particularly relevaot. Firstly, with regard to context we should
avoid the fatalism usually associated with the term 'structure' in much International Rela-
tiom (IR)literature. Context is oot a monolithic, impenetrable, entity which pre-exists actors,
and against which actors stand virtually powerless. Rather, context is here mainly intended
as Oi.her actors, no more and no less than the set of relations whicn they entertain and the
patterns they have generated. Even the material enviro:lment, which is an importaot and
arguably 'objeclive' parI 01 the contexl, becomes lully ,lleaningful only lhrough lhe rela-
tiom that actors establish with one another. The coexistence of different actors, their inter-
action and complex aggregation of interests, is what makes 'the international' an unev€n
terrain for foreign policy. The likelihood of achieving an objective is dependent 00 how
strategically placed the actor is on this terrain: in other words, given it5 position in relation
to .lhe context, some actions will be more successful than others. Moreover, and even more
funoamentally, because of its inherently relatior1al nature, context means different things to
different actors, depending not only on where they are placed, but also 00 how they ioter-
pret the features of the terrain surrounding them. The c(cles of isolationism and interven-
tionism in US foreign policy, for instance, are to be understood not 50 much as resulting
from an objective change in the country's position in the world but, perhaps more impor-o
tantly, as a result of different interpretations of the same position, with its balance of con-
straints and opportunities.
Secondly and relatedly, there is a constant interplay between actors and context, and it is
throJgh lhis inlerplay lhat behaviour is produced. This in lurn does nol play itsell oul at lhe
material levei only, but ís mediated bythe role of ideas anc discourses. Thus it is important not
just tO take inta account the way the context responds te the actors' behaviour, but also the
way such responses are filtered through perceptions, paradigms, and narratives, eventuallyto
be internalized in the political processo
Thirdly and finally,just as there ís a constant interplay, 50 is there a constantfeedback frem
the ador to the context and vice versa. Gnce produced through an interactive process, for-
eign policy behaviour leeds back into lhe context (Figure 8.1f,), restrudurinE; the environ-
ment ,ór leaving it unchanged, and into the actor itself (Figure 8.1fJ by making adaptation
possible. Think for instance of the various repercussions af US foreign policy actions in the
 CHAPTf:R i:.. lMí-'lE"/jENTATJON AND 8t~AV!QUR 151

The OperationaJ En\'ironment

L\."ternal Internal

1
Decision Makers

The.Psy:hological Environment

Tmage-:-Perce pti o ns--C ulture

1
Formu.lation and Decisíon

Action

Implementation
,-". '~i ~ __ ~

Figure 8.2 The plac= of implementation in the foreign-poilcy-making processoThE arrOW$represent the
flow of decision making and the main lines of feedback_

Source: Adapted from Brecher (1974: 7i.

Middle East. ThES€ have not only changed the context at the regional, if not ir,ternational,
leveI, but have impacted on the US itself, sometí'11es causing a domestic reacIlon against in-
volvement and at other times an upsurge of nationa!is;T:.
Figure 8.2, building on the work af Michae! B;-echer among other systems theorists who
have worked on the subject, outlines the proce,:,ses of action, reaction, and feedback which
characterize the foreign pclicy making process, creating endless loops of policy and imple-
mentation ra{her than the :lear progression of stages, formulation-choice-decision-action,
vvhich a rati~nalist approach might be thought to presuppose. (See Box 8.1.)
152 EUSABETTA aRIG~! ANT~ CHRISTOPHER HH.L

BoJi 8.1 Syste:mstl1eóry -~ ,. -'

Systems theory is the approach. deriving [rom natural :;cience but G.5sociated in polítical science mostly
with me work af David Easton, which sees mos! phenomena as interrelated through processes af input,"
outPl.lt, and feedback fram the environment. The system is characterized primarily by a process af
homeostasis. ordynamic equilibrium through interaction af the various forces involved. This cao be as
true of international polítics, for example via the balance af power, and foreign policy, via the instind for
political survival, as it is af natural features such as body temperature ar dimate.

Since both strategy and context are important in foreigr, policy. we now take a dose r look
at each in turn, starting frem the latter. 'The international' is the natural eontext of foreign
palie)'. and yet there is more than one sense in whieh this habitat, natural as it may be, pro-
vides a rather eomplex and challenging environment for states to operate in. The section that
follows wilJ look at why this is the case. We then move to the side of strategy, 50 as to examine
how the 'domestic' affect5 foreign policy in its implementation.

Ideas of the 'ínternational': a view 'from somewhere'

As illustrated in the previous 5ection, the 'international' means different things to djfferent
aetors, depending not only on where they are placed, but ai 'o on how they (actively) interpret
the canstraints and apportunities offered by contexto Fromthe perspective Df a single foreign
policy actor, then, the 'international' appears a rather varied landscape, with features that ean
be on y manipulated in part.
The ecologlcal metaphor is probably a good place lo sla-t lo consider differenl piclures of
lhe 'inlernalional' (se e lhe pioneering 5prout and 5proul 1965). In facl, when discussing lhe
reach of a country's foreign policy, it is customary to turn to geographieaVgeometrical meta-
phars sueh as 'cirdes' or 'spheres'. Winston Churchill's image of the 'three cirdes' af British
posl-war foreign policy is probably lhe best known case in poinl, bul one only has to lhink
about how diffuse the expression 'spnere af influence' is :0 understand how this mode of
language is ingrained in the exercise of representing the 'outside', or 'abroad' (Dodds and
Alkinson 2000).
From the perspective of a single foreign poJiC'y'actor, the 'international' has at least two di-
mensions: horizontal and vertical. Horizontally, the international unfolds on th-:! continuum
frem proximity to distance, fram 'near' to 'far', frem regional to global. Vertically, the interna-
tionai is stratified into a number of functionallayers: political, social, economie, military, nor-
m"atlve, and 50 on_ Without any doubt, when called to formulate in~erests and implement
objectives, the greatest challenge forforeign poliey makers is to harmonize the two dimen-
sions and keep a certain degree of internai consisteney within each.
In horizontal terms, the implementation of fareign policy objectives starts from the envi-
ronment dosest to the actor, which means the neíghbouring states usually grouped in a re-
gion. Regional environments are specific to where aetors are plaeed, and depend on how
eonce1trated ar widely spread their interests are. For most continental European countries,
for instan'ce, the regional environment coincides with the borders of Europe. However, as the
case o'::~urope well testifies. the regional borders of an actor's foreign policy are farfrem fixed.
 153

They are constant!y susceptib!e to being renegotiated foliowmg h-lstoric21, political, ar simply
ideoiogical developnents. \Ni;:ness the fate of Eastern Europe, hardiy a foreign paliey prlority
for most European states before 1989 (with a fevoI nOlable exceptions such as Germany.
France, ar Italy) and now cOf\:iidered to be part of a single European region (Wallace 1990)
Consider also how g€ographica! proximity does not by definition ensure ínclusion in the r€-
gion or ínterest t? foreign poiicy-geography must aiways be read in conjunction with politics.
Thus, for instance, despite erupting at the heart Df the cOIli.!nent geographically, the Balkan
wars were at first dealt wltn as if at the periphelY ot Europe (Simms 2001)_ Conver~ely, it was
the political and ideological band of cümmunism which connected geographically distant
5tates (as with Cuba's ar North Korea's reiations with the forrner Soviet Unian) in a relatively
homogeneous enviíOnment.
If ali states have 2. region of priority for their foreign policy, Dnly a fevv can really aspire to a
have a genuinely global frame of reference. That the .intemalianal.isllore and more fre-
quently equated with the 'global' testifies to the success af the globalizatian paradigm, but
does nat de facto imply lhe possibility for ali actars to exercise a wide-ranging fore;gn policy.
Yet, given the technologlcal revolution and the power ot ideas, foreign palicy as a política I
activit'y' has the potential to be g!obal in both its causes and it5 effects--and has no!, contrary
to expectations, be::orne obsoiete_ in fact, there i5 a sense i,1 whích, in conditions af globaliza-
tion, aI! politic5 hQ5 beco me loreign palie}' in Orle way or the other.
,ó,ndyeI, not mal'y' acwrs can elaborate, let alone atford, a truly global foreign policy. The
USA has most notably laid clairn to this status, pan:icularly since the-enc of the Cold War, re-
inforced in the wake cf the 2001 attacks. However, despite its overwhelming m'lítary and
econamic power, America's vision af an effective global reach has been only partly fulfilled,
suffering a number af important setbacks.
Interestingly enough, it is precisely in the phase af impiernentation that the USA1sforeign
policy designs have mOSI frequent!y failed. lf we laok back at the strategic-relatianal model
presenteá earlier in the chapter, this is nat at ali surpnsíng. A failure to íake into accaunt the
strategic and interactive nature af foreign palici means a high likelihood of problems at the
implementation pt-ase. As analyst5 have noted, many af the difficulties encauntered bythe USA
in its foreign palicy (let alone llilitary) projection are due to 3. poor appreciation ofthe crucial
relation betvveen ends and means on the one hand, and between foreign policy actions and
context, on the otr-.er_This alc-ne wou!d explain much or the rrustration encountered, without
consideringthe additiona! faiiure to take into account the mediation ofideas and tt'-eir impact
on such an interplay. At anotl1er ievel, it simply means that even the USA cannat dominate
everythíng in a complex wodd, a fact that it does not a\wavs acknowled~e.
For ir.iddle and 5mall state3, the [global' remains but an aspiratíon, or a rhetorical commit-
ment. The case of Britain is instructive here: despitetne Ne'N Labour pledge of a foreign policy
Jnformed by global normati\'2 commitments, the difflcu!ties af implementing such a grand
design have been countless- over the last decade, in economic, militar~/, and political term5
(Dunne and Wheeler 1999)_ tv',ore generally, as some Df the :iterature h3S made clear, one af
the paradoxical effects of globalization has been that ar reinfarcif1g the regional dimension,
pushing middle t(l small states especially to strergthen thE: regional scope af their foreign
poltey (on the .rise of reglOnallsm, see Hurrell (1995)).
Thus, the pE}~imeter af lhe 'international' varies great!y depending on the actor cansidered,
and especial~! on its position in the context, on the resources at its disposal, and or; the
-154 ElISABETTA 6RiGH! ANO CHRISTOPHER Hlll

strategic value af these resources. Sut there is also a second dimension along which actors
measure the 'international', and that is the vertical axis af functional differentiation. Thus, the
'international' results not just trem it5 horizontal extension but trem its stratificatlOn in differ-
ent layers. the most important being polítical, economic, military, normative. and cultural. Two
qu~lifications must accompany such a tharaderization, however. Firstly, the hierarchy among
layers is by no means fixed; indeed. the traditional distineticn between 'high poiitics' and 'Iow
politics', which c1aimed a primacy lor political and military issues (Hoflmann 1966), is increas-
ingly problematic in a world in which issues such as culture have became (ar rather, have re.
turned to be) the terrain of greatest contestatian. At the very least. what counts as 'high' or 'low'
pclitics changes from actor to ador, and is inevitably subjeet to political. let alone idiosyn-
cratic, considerations. Secondly, while analytically separable. these layers are at least margin-
ally interlinked, partly because any given loreign policy has efleàs at many díflerent leveis,
and p.:.rtly because layer5 overlap in importan! ways, empirically as well as conceptually.
The polilical layer 01 the 'internatíonal' ís lormed by the complex web 01 interrelations
which bind actors together. Diplomacy is one, the traditicnal but criticai expression of the
existence of such a web, which consists of far more than just international institutions ar 're-
gimes' Moreover, in conditions of globalization, the political dimension Df the 'international'
acquires, ai least potentially, lurther depth in three directíons (Held and Archibugi 1995).
Firstly, the damestic politics of states, especially large ones, becomes a factor in this interde-
pendence, affecting other a:tor5 through their foreign policies and sometimes also their own
domestic polilics. Secondly, the progressive lormation 01 a 'global public sphere' means that
political interdependence gradually comes to feature processes of normative adjudicatian
and contestatian, until recently exclusive to life inside state;. Thirdly. a variety of actors. stat€
and non-state, participate in the interdependence which makes up the politicallayer 01 the
'international'. However, this does not happen on a conditian af parity, as states still express
their agency thrcugh channels which are far more institutionalized, accountable, and varied
than thase at the disposal of non-state actors.
lhe po1itical dimension af the 'international' has important areas of overlap with the social
and normative layer. Diplomacy is in fact one 01 the key institutions 01 whatthe English School
oflnternational Relations calls the'society' of states. or'international scciety' (Buli and Watson
1982). Norms are another important component, in both their more codified version (inter-
nationallaw) and their inlormal variety (customs). More generally, this is the levei at which
ethical concerns play themselves out. lhe eXlent to which :hese have come to afleà loreign
policy is nowhere more apparent than in the wave of 'humanitarian interventions' which was
initiated in the early 1990s (Wheeler 1997). This practice provides evidence that loreign pol-
icy must now confront an environment which has element) cf society-and a society com-
posed 01 individuais and social movements, as well as state; (Linklater 1998).
Theeconomic layer is. if possible. even more pluralistic in its inclusion of a variety of aetors
oldíflerent nature. Not surprisingly, here the superiority states enjoy is lar less marked than at
the politicallevel, partly because of the less hierarchical nature of economic transadions and
partly because ecanomic interdependence has afien thrived irrespective cf the internationaf
political systems in place (Strange 1988). However, economic issues are constantly suscepti-
ble to being highly politicized: witness the case 01 natural resources, and how Ihis issue has
become ~"matter af greatest concem for the foreign policy of states. especially emerging
powers S>Jchas China (Alden 2007).
 155

At yet another levei, foreign policy mus! take into accoum: the exístence of panerns of miii-
tary a!ignments, batr. cooperative and adversarial. Accmding to some treories of interna~
tional rela-rions, mos;: notably neorealism. this is the layer "A;hich is ul1imate\y the mos!
significant in foreign poliey terms. While this may be::rue in the sense that military affair::. carry
with them the greatest threat, that of physl(al annihilation (Aron 1966), security problems are
usually multifaceted and often derivative of political, economic. or cultura\ conflicts (Omand
2010). Still, the asymnetrical distribution of military power, and:n partícula of nuclea.rweap-
on5, is a major fact for foreign poliey mak€rs in ali states.
Finally, there is an important cultural dimension to the 'internationar which decision mak-
ers cannot afford to leave out in thelf effcrt to ímplement foreign paliey objectives. It is nat
just that cultural factors such as religion have come back, supposed\y to ignite fundarnental-
ism and terrorism, but that these ractors today play an increasíngly important role in a.1Iinter-
national relations (Petito and Hatzopoulous 2004). This is due on the one nand to the forced
contiguity among different cultures brought about by globalization, and Dn the other to the
decline ofthat modEm paradigm which marginalized ali forms of culture (religíon in primis),
confiningthem to lhe realrn of the private. Foreign polJcy now finds itself dealing with these
issues as priorities, and with the complications produced by their entangiement with a\l the
remaining dimensions_
To sum up, the context of foreign poliey means different things to dífferent actors. accord-
ing to who and where they are_The 'international' is a kaieidoscoP1C forrration v,lhlch devel-
ops both horizonta Iy, extending from local to regional to global, and vertieally, layeríng
polítical, economic, miiitary, normative, and cultural dimensions. Despite its varied complex-
ion, context is often perceived as a who!e, as a 'system', by forelgn policy makers (Hill 2003:
164). And yet, intere;tingly enough, the greatest challenge for them 1sprecisely how to ensure
that ali these dímensions do not contradiet each other. There is, in fact, a natural centrifugaI
tendency that threatens eonsís-;:encyand coherence in foreign policy. Complexity breeds spe-
eialízation, if not fragmentation. Thus, it is very frequent ror economic foreign policy to devi-
ate from that officia,ly played out at the política1 level; thls in turn is ofter in tensior with the
pnnciples governing the normative dimension of the 'internatíonal', and 50 an. The exercise
af making these diferent logics work in synergy in the pursuit of coordínated objectíves is
eertainly one of the most daunting ehallenge5 for ali foreign policy make's.

Balancing 'inside' and 'outside': implementingforeign policy

!f implementation i:; about reaching out into the environment to transform one's objectives
into outcomes, one should not think of this process as exclusively directed to, let alone fram,
the outside. On the =ontrary, the implementation of goals in foreign policy involves an impor-
tant damestic ar 'internai' component. More specifically, it in'JOlves an act of balancing, and
indeed a process of interplay between 'vvhat goes on ínside the actor and its prajection to-
wards the outside. A,S the strategic-relational rnodel presentEd above illustrates, aI! af these
dialectical processEs take plaee in the political process and are mediated by the impact of
ideas and díscourses.
There dre at !east tiNO general ways in which thE 'domestic' is implicated in fore:gn palicy
in,plementatiol{l., aside from the veryfundamentai role Df decidmg whích objectives to pursue
in the first pla~e. Te 5tart vvith, implementation presupposes not onl)/ the capacity to pursue
156 ~L1SABETTA BRIG!11 A,1\iO CHRISTOPHER H!tl

goals with effective means, but more generally the ability ofgovernments to extract and mo-
bilize resources frem their audiences, both material and immaterial, and channel them Into
the pursuit of given objectives (Mastanduno el aI. 1989). The most classic example of mobili-
zatian happens. af (ourse, when states go to war. In the kinds of'total wars' experienced in the
twentieth century, entire societies were involved in sustaining the war effort (natiam go to
war, as the expression has it), with theireconomy and culture transformed bythe will to attain
war aims. But more prosaically. either simply thro~ghthe colledion af taxes ar through more
specific actions, societies take a dired ar indirect part in realizing foreign policy aims. Sec-
ondly, but relatedly, at least in demacratic societies, the 'domestic' enters the picture of imple-
mentalion in the form af the consensus needed to susta n the foreign policy projection
necessary to attaln objectives (Lambom 1991). When a modicum of cansensus is missing,
foreigr policy is undermined from below; as a result, implementation is potentially much
weaker, ar can be even at risk. In fact. jf consenSLJSbreaks d:Jwn entirely, a crisis can erupt to
threaten not only the foreign policy in action, bu: the survival of the government itself.
In general, therefore, implementation always develops on twa leveis, 'damestic' and 'inter-
national', which are in constant interaction. This is what the political scientist Robert Putnam
had in mind when he imagined foreign policy as a 'two-Ievel game' (Putnam 1988: Evans et
aI. 1993). Using Ihis metaphor. Pulnam focused on the issue of how democratic foreign pol-
icy tends to be internationally and domestically constraine::J in the specific context af multi-
lateral economic bargaining. As the Iiterature in Foreign Pol;cy Analysis (FPA) has made e1ear.
this imuition can be applied to foreign policy lato sen5U, and indeed encapsulates an essen-
tial feature of the process of implementation. In the wards of Wolfram Hanrieder. the first
foreign policy analyst to examine this issue in detail, implementatian hinges on a 'compati-
bilily-consensus' balance and unfolds wilhin a 'double constraint' (Hanrieder 1971). In order
to be successful in achieving their objectives, actors need to pursue a foreign palicy that is
compatible with the context ando at the same lime, supported by a reasonable degree of
agreement inside the state. Thus implementatian calls for an attention to both fronts, do-
mestic and international, and foreign policy makers need to make them work in tandem as
much as possible.
Butthe exercise of balancingthe damestic and the international does not exhaust the ways
in wh ch these ambits can be connected in the phase of foreign policy implementation. In
facto s:Jmetimes it is the dynamic inteíplay, ar synergy, between them which is of most inter-
est. This happens, for instance, whenever the attainment of a foreign policy goal has domestic
implications, or vice versa. In fact, sometimes domest;c objectives are achieved via particular
forei9,1 policies, whereas foreign policy objeetives are pursued via domestic policies. When this
happens, the synergistic (or dialectic) nature of foreign policy manifests ilself most clearly,
and the process of interplay between actor. context, and foreign policy at the heart of the
strategic-relational model seen abave comes full circle.
As ;or the first possibility. the choice of many countries to join the EU provides a goad case
in po;nt. Naturally. entry into the EU is portrayed primarily as a foreign policy issue; negotia-
tions. after ali, take place at the levei of the foreign policy apparatus. And yet, Ihere is a sense
in w~ich historically the entry into the EU (or the European Community before Maastricht)
was pursued by policy makers primarily for domestic purposes. Think of lhe pursuit of en-
largement by Spain or Greece duringthe 1970s and 1980s.and howthis was funclional to the
overriding domestic objeetive af democratic consolidatian. The same logic applies today to
 157

some af the applicar.ts Trom former (Tovias and Ugur 20G4). "Further, there is
Eastern Europe
no doubt that part Df the controversy surrounding the accessian ofTurke;' to the EU-in both
parties-o~íginates rrom cancem over the set of domestic objectives which accessicn is sup-
posed to facilitate, !T,ost notab!y poiitical refOírT'L.
Examples of the cpposite case are also frequent, and indeed very relevant to the current
predicaments of some democratic states. Today's conditíons of globalization, and especially
multiculturalism, mean Ihat minorities become the focus of cancem as the result ar foreign
policy entanglemenrs whicn then b!ow back into one's own society (Hi112007). For instance,
the high degree of suspícion, and accompanying restrictive policies, attaching to Muslim citi-
zens in the USA and Britain aher the terrorist attaeks of 2001 and 2005, respeetiveJy, are the
indirect eonsequence of foreign paliey engagements in the Middle Eas'(, and reverberate in
turn on relations w!:h major Muslim states such as Saudí Arabia and Pakistan_ The~efore, as
argued in this sectiol, a degree of interplay between the domestic and thE international in the
process Df foreign policy implementation is inevitable, and indeed nece5sary for its success.
This is true in at least two ways: firstly, do:nestic participation features in ~he implementation
phase in terms of either the need fOi consensus aí the specific resources ta be rnobilized;
secandly, It is thraugh foreign paliey that "ihe 'domestic' can become the c,annel by which the
'internati:mal' is pursued, and vice versa.

Exerting influence
Ali foreign policy, bi defínition, is about the outside world. While the issJes or the definition
of 'outside' and thE boundaries between 'us' and 'Lhem' are highly (onrested, they are dis-
cussed eisewhere in this boak. The purpose of this seetion is to examine the practical prob-
lems which occur vvhen foreign-policy-making processes collide with ~he v\i'orld for whích
they are intended. Intention, however, is itself a vaíiable phenomenon il1 this contexto Some
foreign policy is initiated at home, whether by a new government, a strollg-rninded Jeader, ar
pressures from belmlJ, such as those represented by natianalisrn. But many other foreign
po!icy positions are reactions to events beyond borders, and thus either to the initiatives af
others ar to chaim of events which have spiraOed beyond any single actor's control. Either
way, a poJícy can be rationa! ar not, and compatible ar not, with other aspects of the govern~
ment's programme. 'let whatever its interna! logíc it still has to face up to the probiem or im-
plementation, i.e. the putting of a po:icy into practice through engagement with other
independent actars, vvhich are often physiea11y beyond reaeh.
lmplementation has severa1 meanings in FPft., two ofwhich are focused 011 here: on the ane
hand is the issue cf the channels through which foreign policy aims aretranslated into prac-
tice, involving the often complex relatiol1ship between ends and means, on the other are the
difficulties whicn states have in operating in \fvhat is Iiterally a 'fareign. anel quite often a highly
intractabJe world, a:'ld how they adapt theír beha'criour on the basis ofthe interaction with,
and feedback tram, that outside world. Tnose \!vho work more on the poltcy-re1ated si de of
FPA have always w:ítten aboutthe challenges rept.esented by a particulôr instrument. partícu-
larly diplomacy.~nd militaryforce. Detailed research has also been conducted oro propaganda
and the use oteconomic sanctions, while ke~1figures like A!exander George trave ur,derscored
the links which e>:ist betvo/een instruments, as in his influential work on .coercive diplomaey',
158 EllSA£SET1A BRIGHI ANO CHRISTOPHER HJll

which has now spawned lhe subfie:d of 'defence diplomacy' (George 1994). More direclly,
however, theoretical work is also vital to an understanding cf implementatian, whether relat-
ing to the bureaucratic dimension, ar to the underlying problems af planning and rationality.
Graham Allison has been lhe most influenlial figure on bclh counlS, providing a bridge, as he
does, intothe work cf economists and administrativetheoristslike Herbert Simon and Charles
Lindolom (Allison and Zelikow 1999: see also Chaplers Si>:and Seven).
The above lWo meanings of implemenlalion will be explored by looking atthe following in
turno

• T1e variety of relationships that exist with the outside world because cf the many
differenl kinds af .stales conducling foreign palicy and lhe varying challenges lhal lheir
e:<ternal adivities involve. For instance, the implementatian cf British policy towards
I':ew Zealand is a ralher more slraightforward malter lhan ils conducl tawards Belarus.
Despile lhe much grealer geagraphical dislances beM'een lhe parties la lhe firsl
relalionship, lhe degree af 'fareignness' (i.e. palilical and cultural dislance) is far less lhan
in the secando

• The foreign poHcy instruments available to decision makers as they contemplate the
best way to translate their intentions into actions which have a chance of success in
the international environment. The main instruments :all into four categories: political,
military, econamic, and cultural/idealogical. Vel any a,alysis of lhem saon encounlers
camplex problems, in the first instance of the choices aver the use of which instrument
for what purpos€, and in the second instance over therelationsh.p between the
inslrumenls lhemselves and lhe underlying capabililies which make them passible.

• The theoretical issues raised by any discussion of the ends-means relationship in foreign
policy. In lhe canlext af implemenlalian, lhis means lhe issues af ralionalily, slippage,
a1d complexity.lt also means some particular reflections on one of the central concepts
in ali IR, namely pawer. The key issue here is lhe dislin:lian belWeen power as a means,
aíld as a contexto

The practical importance of context

FPA is a camparalive field af sludy which generales abservalions of varying degrees of gener-
ality. Samelimes its insights will need lo be heavily qualified lhrough lhe particularity of pe-
riod and circumstances, while others will amount to propositions of wide applicability. In
lerms of implementalion, il is unarguable thal lhe fallow-thraugh phase of decisian making
(which in lhis case is more praperlytermed adion, ar agency) always has the capacity lO raise
newproblems and to derail the original intentians. If this now seems an unremarkable state-
ment, it was not always the case. Even today, rationalists often do not make allowances for the
faet that choices and trade-offs are not the only determinant af outcames-chaices are not
self-executing. For their part, in their enthusiasm. politidans very aften neglect to factor in
either the 'foul-up' factor ar the inconvenient unwillingr.ess af outsiders to canfarm to the
roles expected af them. It is enaugh to mention the gap between intentions and outcomes on
lhe pa[Í of lhe proponents of lhe Iraq war la make lhe painl.
 CHAPTEP. 8 IMPltt~fiENTATlON :",i-.f!.l fiEH.L.VIOUR 159

Great powers, smal! powers

However, this kind of high-Ievel generalization is only a start. To understand implementation
more fullywe also need fine-grained work::>nthe basisof dislinctions bet.vveenlhe kinds cf ac-
tors producing foreign policies, and between the kinds of relationship in 'Nhich they are en-
gaged. On the first count. for example, it might be thought lhal great powers (to say nothing of
the world's only superpower) would have far rewer problems in implementing their externai
policies than small ar.d/or weak states. Sul this is not necessarily the case. It depends, :rucially,
on what aims are being sought. Jespite itsstatusas a middle power, and the considerable array
af means at its dispmal, Britain 's tom between aspir2tion and (in)capacity.lt has, for instance,
failed to fulfil many of the foreign policy aims dictatee by its 'new' global agenda. This 's only in
part due to the faet that 'lhe global' has been useó primariiy as a rhelorical strategy. as men-
tioned above. More i1terestingly, itwas thevery nature af some foreign policyaims, most nota-
blythose of a normative kind, which was difficult to match with the means usedto pursue them.
A small countrywhich overreaches itself, in terms of seekingto changethe whole character
of lhe international Iystem (as Fidel Castro's Cuba has oecasionally tried lo do. and as Hugo
Chavez beca me fixated on in Venezuela) riskseven greater complications if not outright fail-
ure. On the other hand. this is not to say that they wi!! achieve nothing. If they have already
discounted the risks and the unlikelihood 01 aehieving the stated goals. they rnay slill fulfil
lesser, and probably unstated. goals of a satisfying kind. Thus for four decades Muammar
Gaddafi in lib\'a managed to defy predictions of his demise and, despite his undoubledly er-
ratic behaviour and the hostilit\' ofthe USA. exerled a disproportionate degree ofinfluence in
the Maghreb and su:)-Saharan Africa. By extension, a small country whicr. remains modest in
its goals may have telatively few problems of implenentation because it will be toa cautious
to attempt anything which anlagonizes the more powerful, or exposes it to other kinds of
potentially destruetive blow-baek.
Conversely, the USA often encounters 5erious proalems of inplementation with itsforeign
policies, precisely because it has global i1terests ano is active on almast every froni. As the
'hegemon'(leader of the free world', it has a forward stance on so man'/ issues that almost
inevitably it runs in:o difficulties in some of them. Overstretch is a term which refers to a
structural condition over a historieal period (Kennedy 1988). It refers to the lendency of great
powers to take on imperial commitments which they cannot sustain, financially ar militarily.
In particularcircumHances, it may take the form of a foreign policy which is undertaken with-
aut the available reSQurcesto follow it through, even jf in principie the state in question
should have no pro'Jlem in doing so. The USA discovered this truth in Semalia in 1993, from
where it withdrew ,!ter only a few easualties, concluding (possibly with undue haste) that
there was not the domestic support for a long engagement. The Soviet Union suffered the
same fate in Afghanistan. after much longer and mu:h greater losses,during the years follow-
ing their tnvasion in December 1979.
Thus, the foreign poliey designs of great powers have moS! frequemly failed in relation
to implementation. This result, though puzzling, can be illuminated th'ough the strategie-
relation model presented in the first part of the chapter. Two of its insighrsmust be kept par-
ticularly in mind..,Firstly,given :he constant interplay between strategy and context, 5uccessful
implementatior" requires a certain degree of f!exib~lityto 2ccommodate on-going feedback
processes.A fO'Teigílpolicy which 15 projected to lhe outside without much understanding af
160 ElISABETTA 9iR~GH! ANO CHRISTOPHER HIí..L

such in:erplay is likely to backfire, as recent American foreign policy has vividly demonstrated.
Secondly, a successful implementation depends al50 on the crucial retation between endsand
means. No matt~rhow
,;6':" __ ~.
powerful ar big a state is. the pursuit Df foreign policy aims is contin-
gent on the ever-important choice of the appropriate means.

Multilateralism and the complexity of adion

The military dimension is not the only cne in which problems of implementation arise. Tony
Blair apparently succeeded in getting the G8 to commit ta a policy of debt cancellation in
Africa during the Gleneagles Summit of July 200S, only to find that many of his partners sim-
ply failed to live up to their promises. This is one example among many which demonstrales
that, almost by definition, any foreign policy action depends on others for its full implemen-
tation.lfbeing pursued bilaterally or multilaterally, it will require the cooperation of partne".
Sut even then, and certainly in ali unilateral aetions, it depends on how the majority Df actors
affected, whether hostil e, supportive, or just indifferent, respond to the action. Ifthey choose
to tak€ an interest in the subject, for whatever reason (and the indifferent may decide to take
a stance just to give themselves leverage on something else), these actors are likely to cr~t~
friction, add costs, or at least complicate the implementation of the policy. Eveo if they are
neutral on the substance, theirtechnical assistance may still Je needed, as with thecontrover-
sial (and therefore secret) rendition flights of US aircraft in ferrying presumed terrorists to and
from tleir detention centres.
Despite the controversy of recent years between the USAand its allies as to whether unilat-
eralism, multilateralism, ar 'effeetive multilateralism' (the compromise position) is the pre-
ferred approach to international relations, the reality is that most implementation entails
some or orh'êr form of multilaleralism. Occasionally states indulge themselves in pure my-
opic solipsism, as with the wild calls of Iranian President Ahmadinejad for Israel to be wiped
off the map. Even then, their real aim is usually to rally support in a particular quarter ar to
provoke reactions in another. But for the most part, states take for granted the faet that sue-
cess in foreign policy will require mobilizing support, neutralizing hostility, shaping the bal-
ance af influenee, and (increasingly) winning ihe rhetorical wars which charaeterize the
modern multilayered international system. Often this worktakes plaee within formal interna-
tional organizations, whether universal through the UN system, or partial, in theform of net-
works of allies, regional partners, or the 'Iike-minded'. But just as much is ad hoc, cutting
across institutional boundaries and not restrieted by formal rules ar agreements. Even in it5
moment of maximum self-assertion, when it disregarded ~heArticle 5 offer of help from its
NATC allies immediately after 9/11, lhe USA was collaborating pragmatically with a wide
range of countries involved in the hunt for al-Qaeda members. The 'War on Terror' could not
be other than a colleetive affair, even if it also divided the 'Norld crudely into 'those who are
for us, and those against us'. In other words, the deployme"t of the immense national power
which the USA has at its disposal is in itself no guarantee of effective implementation. The very
use ofthe famous 'axis of evil' image was an attempt to mobilize the international community
on one side by 'othering', or seapegoating, a small number af seemingly irresponsible states.
There(9re implementingforeign policy usually requires the simultaneoys useo~variousleveis
and tecbniques af international cooperation: bilateral, multilateral, and transgovernmental-
that is, links betv/een parts of one state's machinery and parts of another, as with lhe privileged
 161

li:1ks beh'oIeen rhe F,ench and German ministries oi Defence which produced the joint bri-
gade in the late 19805 Not ali of these will be visibie to the public; indeed many, perhaps
most, wil! operate at the level af what used to be ealled 'secret diplomacy', Secrecy is an ave r-
rated quality these cays. as reiative!y few activities (equire the absolute darkness associated
\",'ith the preparation of, say. a 5urprise attac"k. Even in those cases, it is impossible to malntain
absolute surprise. There were plenty of indicatiolls Df Hitler's impendingattack on the USSR
inJu"e 1941, of Israei's 00 Egypt in 1967, ar of Argentlna's on the Falkland Islands in 1982 for
those who wanted tJ listen, ar were capable of reading the signs correctly. In the contempo-
rary med:a-driven envrronment, it is especially difficult lO keep somethirg secret for long, as
the Reagan administration discovered in 1986-1987, when jts attempts to use money from
arms deals with lran to fund iIlegal military campaigns in Central America (the Iran-Contra
affair) were exposec with serious consequences for its policies on both rronts.
Ali t~is is to say that most foreign palicy implementation involves a tangled web af connec-
tians with other stales, ar at least parts af othel' states, which is both necessary and a serious
camplication aí agency, as it may compromise the a~piration tcwards a single fatiana! strategy
as well as the control af outcomes. 1fforeign poliey inevitably means subcontracting out vari-
ous parts ofthe enceavour to different parts af the state bureaucracy and to outside entities,
then those subcontractors have the capacity to refract, di5tort, and even subvert tre policy's
original intentions. fhis is the ~trategic-relational approach 10 :)factice-looking inside as well
as outside the state. One might adapt Truman's fam:)us remark abaut General Eisenhower, as
the latter prepared;:o take over the Presidency: 'He'll sit here, and he'll sal, "Do this! Do that!"
And nothing will happen. Poar Ike-it won't be a bltlike the Army' (cited in Neustadt 1960: 9)
lnside any polítical machinery, and e'/en more <::'0 in the complex world of international
relations, 'arders' may easily be issued, Dut that is only the beginning of the process of at-
tempting to achiev::, one's goals. Moreover, thís Is true for ali kinds of states, whatever theír
slze ar level of sophstícation

The instruments of foreign policy

When it comes to choosing the instruments with w-1ich to act, the differences between states
do come into play. The wide variation in state capacities is a key determinant ofwhat can even
be attempted in the outside world. The larger states will possess the full portfolio af patentia!
instruments, from lhe hardest af hard power to the most subtle and irdirect cultural influ-
ences. They will alsD have the capacity to act well beyond their own loca!ity, perhaps globally.
They simply have more people, more contacts, and mo.e money than the others. lt is, indeed,
a plausible definiticn af a superpower that it can expect to determine cutcomes in any geo-
graphical arena anj 'viz any available instrument. At the other end of the spectrum, it will be
a major challenge :Oí a micro-state just to preserve its auwnomy. Its foreígn policy will have
no further ambitíon than to assist in the achievement of basic domestic policy goaís through
diplomacy alone. Betv.'een thes€ two extremes, mast stales survive on the basis of a I\mited
and patchy range of instruments, possessing armed services of highly variable size, quality,
and scape, embassies In som2 parts of the ~'voíldbur 110tali, the ability to exert econornic in-
fluence accord61g to ievels of development and/or the lottery of geographical position, and
probably very límíted cultural outreach.
 EL1SABEITA 8:R:iGH; ANDo CHRtsTOPHER HH.l
162

Any uf1derstanding af how states approach the problem af deciding on the best means af
impleme1ting their foreign policy must remember twO dieta: first!y. instruments are them-
selves dependent on underlying capabilities, which are in tufn a function af the resources at
the dispcsal af the society in question; secondly, decision makers do no! choose instruments
as the surgeon selects the scalpel-rather, the nature af the available ínstruments tends to
shape their policy choices in the first place. These points are expanded in what follows.
Resources refers to what the French school referred to as the 'basic forces' af foreígn policy
(Renouvin and Duroselle 1968; Merle 1987), i.e. a country's sum total of (dis)advantages de-
rived frem climate, position, geography, population size. education. tradition. and leveI of
development. T"",e things are not unehangeable-the Law 01 the Sea Treaty expanded ter-
ritorial waters from 12 to 200 miles in the 19705, and the Channel Tunnel was linal1y built in
1994-but for the most part they change slowly. This is on the assumption that territorial ex-
pansion is not genera\ly acceptable. Where it does occur, as with lsrael's conques!s of 1967,
the parameters of both security and access to raw materiaIs (in this case water) can change
dramatically. Resaurces are a criticai faetar in determining a state's cho:ces in foreign policy,
although there is no simple correspondence between the possession of an asset and the abil-
ity to exert influence, as with Nigeria's wasting of its oil revenues ar lndonesia's failure to
translate its status as the world's fourth most populous country (c. 238 million) into an equiva-
lent pol:tical ranking. Conversely. states with no apparent resource advantages, such as Singa-
pore and Switzerland, have managed to achieve both securit{ and prosperity. Thus resources
have to be managed effectively. What really makes possible the pursuit of ao effective foreign
policy are capabilities. which in turn determine the range of possible instruments at the dis-
posai oi decision makers.
Capabilities are resources made operational but not yet translated Into the specific instrLi-
ments (Iike propaganda or the use of force) which may be ,pplied in praetical polities (See
Figure 8.3.). Accordingly, they are the elements which an in!elligent government will always
seek to improve, to give itself a better chance of implementing an effective foreígn policy, but
which will be seen more as a long-term investment than as provlding an immediate pay~off.
Into this category fali such factors as the strength of the national currency, the size and profi-
ciency of its armed forces, and the skills of its people-this last was the reason why Prime
Minister Blair continually stressed the importance of education to the UK's position in the
world, both economic and política!.
On the other hand, sueh eapabilities are of importance in themselves and to the wel1-being
of any society; their role in underpinr:ing foreign policy is incidental except in cases where
leaders see the latter as providing their primary goals. This was evidently true for Hitler and
Mussalini, who provided full employment as a means to pursuing their country's interna.
tional greatness, rather than the reverse. This means that fcreign policy and its implementa-
tion is to a large extent at the mercy of factors beyond its control and of \ong-term
development5. \t is for this and other reasons that the second dictum referred to above ap-
plies: that decision makers cannot choose on an abstract rational basis the instrument which
woulc best serve their immediate purpose. They are limited not only by the size and wealth
of their cquntry (i.e. by basic resources), but also by the decisions of their predecessors in of-
fiee te develop (or not) a particular eapability wheh would have made possible the prelerred
instrum't'nt. Anc that in turn will have depended on the prior(ty given to externaI policy.
Franc~ in the 18805 was determined to reverse the humiliation it had suffered atthe hands of
 [HAP"T!:R $ !MPlEMENTAnON AND êEHAVIOUR 163

Capabilities

Ar01ed [ndusHi~1& Streng(o oí AS!icultural

GOl' Quality of
forces Technolng;: Currcncy P.-oductivit}'
Repmation!Prestige \ Civil Servire
\ ~ Skilh

,I \\
\ ,,/ Vigorous

- N \
Civil50ciety
,

./
Det(;"rn:nce Diplomar}' Economic Cultural
lInd Militar}' Uip]0macy
Stirks UIlG
!ntervention
C •.•rrots

Instr~ments J
Figure 8.3 -inks betvveen the principal capabillti-es and instruments af foreign policy.

Germany in 1870-1871, and focused on Educationai reform and population growth (but not
with great success) as the means of doing so. Israel has made foreign and defence policy the
overridíng priority tnroughout its existence, although it has only been able to do so through
unwavering US support. But for many states foreigr: pollcy is rather like an expensive insur-
ance policy whose dues seem disproportionate to the risks they face. They often neglect the
relevant capabilities ar divert them In other directions: especially if they 2re able to free-ride
on more activist alliES.They may also misunderstand the link between capabilities and instru-
ments, assuming more choice when it comes to implementing a foreign policy than they in
fact possessoTo put it at its simplest, if they have allowed weapons procur::ment to run down
ar have closed embê.ssies for financiai reasons, they will have r:luch less leverage avai!able to
them when the need arises.ln such circumstances misperceptions are COllmon, and mal' be
fatal.
The actual instruments of foreign policy, that is to sal' the forms of pressure and influence
availableto decision makers, represent an ascending scale af seriousness in termo; ofthe com-
mitment of resources, the impact on third parties, and the according degree ar [isk In use (see
Figure 8.4). This sca1e is akin to the spedrum from 50ft to hard power now familiar in the
discussion of international polítics (Nye 2004; Parmar and (ox 2010). If a píoblem occurs
which requires a foreign polie,! response.~t !.MOUld take a particularly irrationàlleader (it can
happen) to go straight for the high-risk option (inter€stingly now referrEd to as the 'nuclear
option' in everyday speech)_ The pragmatic initial response is to discuss lhe issu€ with ot~er
relevant states, l.e. to empioy diplomacy. If that is unproductive, there may be some attempt
to incentivize comp;iance by various forms of posiLve ar nega tive sanctions. not ali of them
economic. Appeals to an adversary's cwn domestic opinion, through public diplomacy or
employing civil 'Society in direct culi.:Jral ~inkages, may have so.rt:lf:. chanc!? of weakening hjsl
her political ba'"se_
.~-
164. EU$.AB£TTA ~RlGHI AND CHRtSTDPHER H!ll

Military Action

I (puni tive; invasive; occupation)

I

Political Intervention
I (propaganda; subversion; interference)

Negative Sanctions
(boycotts; embargoes; laser sanctions;
restrictions 00 cultural c:ontaets)

Positive Sancrions
(aid; trade agreements; public diplbmacy)

Diplomacy
(disc .ussio
-
01negotiatio
-- -
n)

Figure 8.4 The ascending s.::aleof foreign paliey instruments_

Failure at this levei then leaves the initiating state with a serious choice. Does it go on to
escalate the dispute by exerting punitive measures (assuming it has that opportunity), which
will almas! certainly raise the levEI af tension between the two parties to the paint where it
might as easily spiral ou! of control as produce compliance, ar does it decide to fut lesses and
back off, with the possible consequences of international humiliation and domestic criticism?
The same choice, but of an even more serious kind, awaits furtherdown the road if and when
sanetions turn out to have been ineffective. This was the dilemma faced by the USAand Brit-
ain in 1998 as they attempted to enforce the no-fly zones in southern Iraq on Saddam Hus-
sein, and to press him to renounce the suspected programmes of biological and chemical
weapons productlon. lhe economic sanetians which had been b place since 1991 seemed
not to be working. and indeed were attracting ever more criticism on the grounds of their
damaging ímpact on lraqi civilians. Yet to abandon them without any alternative course of
action would have been to hand a díplomatie victory to Saddam, and perhaps to encourage
him to develop further 'weapons of mass destruction' (WMDs). This reasoning produced Op-
eration Desert Fax. namely the major air attaeks on southern lraq launched by the USA and
UK in'December 1998. In time, catalysed by 9/11, it led to the aim of regime change and the
full-soale invasion of Iraq (Kampfner 2004).
 165

The ladder of escalation in the use offoreign paliey ínstruments is a tendelcy ratner than an
absolute rule. It conforms to a ratiena! idea! type which may anty be hooO\..red in the breach.
Powerful st3.tes are able to use dífferent instruments simultaneous'IY ar in rotatian. They are
certainly ab(e to benefit from the law of antícipated feactions by keeping the mere possibility
-of escalation in the mitds Df their weaker aCversaries. who may decide that.prudence lSprefer-
able to any kind or risk Lesser powers ha'l€ fewer options, and not just in retation to hard
power. Their embassies may be restricted to a rev"! major capita~s plus the UN netwo:-k, their
economic weakness wi\l ft.!\e aut any use cf sanCliom, and their ability to project themselves
abroad culturally wil! be very limíted. This does not mean that they are totaHy hamstrung; if
prepared to take risks, like Castro or Kim Jong il, they may have surprising degrees of succe55,
evei1 over long periods, by being prepared to defy ali theír opponents' imtruments, short of
regime chê..ngeitself. Ir, conversely, they do not ,tJishto attract hmtility, they may still have some
capacity to imp\ement effective policies 50 long as they show creativity and do not become
over-ambitious. Juliu5 Nyerere of Tanzania fitfully displayed these characteristlcs in the 1960s
and 19705, giving his (ountrf an influence on P.Trícanpolitics that it has not had since his depar-
ture (Nzomo 1999: 134-6). More recently. PreSlden; Ivlorales 01 Bolívia ha' raísed his similariy
weak country's profíle with a shrewd mixture of diplomatic activism and dignified restraint.

Power, and the ends-means relationship in foreign policy

The eoncept ar power is a com;non thread not only in the stol)' of implementation but in the
analysis o; foreign policy more broadly. AI! action implies the exercise of p'JWeíto a greater ar
lesser extent, both as a means and as a contexto In the former sense pov ..•.
er, and thus foreign
policy, is an inherently relational activity in that it on\y exists in relatior to some object or
some other party (Baldwin 1985). In the latter sense, as context, power i,npinges on foreign
paliey through lt5 unavoidabjlity; if decislon makers behave as if the power of others, or their
own lackof it, is not relevant, they will soon suffer some unpleasant shock:;. Conversely, ifthey
become overconfidçnt about their power position, ar interpret it toa narrowly, they risk the
usual result of hubris-a hostilecoatition and probable faílure (Nye 2008). C.omplaceney is also
a routine danger. The European states engaged :n the Barcelona Process after 1995 assumed
that while they might not be able to solve the Israel-Palestine dispute. theír diplomatic and
eeonomic engagements with the eountries of the southern shore cf the lv1editerrar:ean were
sufficient to promo;:e stability. This proved, in 2011, to be a fundamental miscalculation.
In a theoretical sense, power is often defined as getting A to do what they might not other-
wise do, ar even eonsider doing. Yet in order to understand the way in vvhich power both
works in the imple:-nentation process and can be drained away duringthis crucial phase, it
needs to be disaggregated and contextualized. In the inevitable shorthand talk in IR of 'great
powers', 'power poiltics', and ~he like, FP,A.provides a useful eorrective by employing middle-
range theory to explore the different leveis and processes beneath the 5uriace of events. As
we saw in Part I, tl€ FPA perspeetive aliows us to unpack the interplay of structures with
agency, :lno in 2. rruch more concrete I,\'ay than the usual díscussion of that reiationship. lt
tends to be sceptical of single-factor e,:_p\anation:,. whether at the levei of the m-:Jtivations,
actions ar enects o~foreign po\icy. Equally, ít shows-and most c1ear\y in the particular context
of implementê-~Ion-how ends and means exist ir- a perpetuai loop of interconnectedness.
with the latter often determining the former
166 EUSABEfTA BRIGHI ANO CHRISTOPH£;R HiLL

The 7'"3tional model, which stresses setting cne's goals in line with available power and then
choosing the mos! appropriate instrument to achieve them. rarely conforms to actual foreign
policy ;:Jractice.ln confronting an unexpected problem, decision makers aften turn to thefirst
potentialsolution to hand. bearing in mind the need to build a (oalitio:l af supportwithin the
goverr.ment and (at times) to carry domestic public opinion with them. They may then be
suckec imo an unforeseeable tunnel cf events wilich throw up yet further chaices aver ends
and means. This was evident in the Ba!kans during the 19905, as the Western states grappled
with the complexconsequences af the disintegration cf Vugoslavia. finally takingon commit-
ments to de facto prateáo;ates in three countries (Bosnia, Macedonia, Kosovo), and (in the
case of the EU) accepting a major enlargement of membership acrass the whole region. The
many complex instruments thus deployed were ostensibly as a means towards the ends cf
stabiliration and pacification, but the longer they stay in place the more difficult it is to distin-
guish the two.lndeed, most foreign policy implementation is bestjudged not via a snapshot
in the moment, but over the long term in relation to changing goals and the flexible use af a
range af means.
Certainly, teaders need to be clear and refleáive about their goals and about the ends-
meam relationship. but in foreign pol!Cy they should nat be under any ilIusion that the latter
(an be held steady, orthat any given means can be relied Lpon to deliver r~sults. The imple-
mentation phase af policy making always involves some loss af momentum through transac-
tion costs,political friction, and disillusion. Because decisions are neverself-execut~ng(except
in thecase of Saddam Hussein, who is said to have literally executed one death sentence on a
ministerial colieague), leaders rely on subcontracting to bLreaucratic agents, some of whom
may take the opportunity to slow down or undermine the policy, or even to run their own
policies in competition, under the cover of agreement. More likely, they will just be guilty of
inefficiency, but this can still endanger the original policy. As John Kennedy famously said
during the Cuban Missile Crisis of 1962 (after one of his SJY planes had slrayed over Sov;et
territory, strictly against his orders), 'There's always some son-of-a-bitch who doesn't get the
word' (Allison and Zelikow 1999: 241)'

Conclusion
The key points which emerge fram the analysis of the confrontation between foreign policy
and the world in which it has to operate ali qualify rationalist notions of power and the ends-
means relationship. They stress the importance of understanding the interplay between con-
texl and policy, between struáures and actors. And the'( highlight the huge potential for
slippage between intentions and outcomes, between acticns and consequences. lndeed, for-
eign policy decisions should be seen primarily as heightened moments of commitment in a
perpetuai process of adion, reaetion, and further adion at many dif.erent leveis and involv-
ing a range of different aetors, inside and outside the state, ali of which need to be taken into
accoJnt. In short, they are best understood through the strategic-relational mode!' The most
important thing for praetitioners to remember is that the point of decision in foreign policy is
usual1y only the start of a long process of immersion in a f1uld ano unpredictable externai
environment. lhe onset of implementation denotes not the end of politics, but simply a new
,
phase of i1.
 167

o Keypoints
0& FOieign policy is not self-executing: the ímp\ementation phase is criticai to success.

<?; The rneans of foreign paliey can distart and even transform its original el1ds.
••. The implernentation af foreign paliey net:ds to be highly flel(jble-it is se!f-defeating to rely on one

instrument alone, ar one strG.tegy for too long.

l} Th!: international envirOnmEnI is fluid an::l difficult to manage_ Foreign polícy makers should be
alert to lhe consta0! feedback it provídes and adapt to its changing circumstances. however clear
their initia! obje::tives.
•. lrnpiementation takes place in severa I ciifferent arenas simultaneously-the local. lhe states system,
lhe global/transnational, and even lhe domestic (of both lhe acting and the receivlng state).
~ Implementatioli can be a purely techníca\ executive matter. For the most parto however. it is as
political-and therefore as ethícal-a dirnension as an'{ other aspect of foreign policy.

~ Furlher reading
(ohen, R. (1991), Negotiatrng Across Cultvres: Communication Obstacles in /nternational
Diplomacy (Washington, DC: US institute of P~ace).
A richly íníorllled cnaly:sis of culturaliy derived misperceptions in foreign policy.

George, A. and Sinons, W.E. (eds), with contributions by D.K. Hall (1994, 2nd edn), The Umits af
Caerc.ive Diplamacy (Boulder, CO: Westview).
The best dlscus:síon of how force a"d diplornacy are oftEn combined, if not always to good effect.

Hili, C. (2Qo3), The Changing Politics af Foreign Policy (Basingstoke: Palgrave Macmillan).
A wide-ranging dilcussion of the conduet af foreign polir.y, with much attention given to the problems
of acting in an intractable env\ronment.

jervis, R. (201 O), Why Intelfigence Fo.ils:Lessons(ram the Iranian Revolution and the Iraq War
(Ithaca, NY: Co~nell University Press).
p.,sophislicated di;cussion of cne of the key issues in foreign policy i;nplementation from one of the

world's leading analysts.

Nye,j.5. (2004), 50ft power: the Means to 5uccess ;n World Politics (New York: Pub1ic Affairs 800ks)_
lhe f.JlIest statement of the argument that even major powers need the fuli range of foreign policy

instrumenb.
Wolfers, A. {1962}, Discord and Callo.boro.tion: Essays on International Palitics 1:8altimore, MD:
johns Hopkins University Press).
;'. classíc. contain ng several essentia\ essays, on the natJre of goals anel the ends-means problem_

Visil the On/ine Resource Centre thar Clccornpanies this book íor rnore ill,formation:
www.oxfordtertbooks.co.uk/ore/smith_foreign/