1

Second draft Not for quotation

War without bloodshed? Public opinion and the war over Kosovo By Philip Everts* and Pierangelo Isernia#

Chapter 1

– Introduction. A new situation, a new kind of war

Chapter 3a

– Trends in Western Public Opinion

Chapter 3b

– Conditions of support of the use of military force

*Institute for International Studies, University of Leiden, the Netherlands. E-mail: [email protected] #Department of Political Science, University of Siena, Italy. E-mail: [email protected]

1

Chapter 1 Introduction. A new situation, a new kind of war Introduction

As a consequence of the end of the Cold War, the question of war and peace has fundamentally changed. First of all, it became apparent during the Nineties that the predominant change was one whereby the relative certainty and predictability of the Cold war setting was replaced by diffuse risks requiring a variety of actual or possible uses of the military forces. In addition, new issues emerged in international affairs not only in the areas of national defence and security, but also in the international economy and the global environment. Among the major changes was the (further) emergence of international conflicts that have their origins in domestic issues, arising particularly from the failure of existing states to prevent or to solve what became commonly known as ‘ethnic’ conflicts. The problems of ‘failed states’ and violations of human rights achieved a much higher place on the agenda of international concerns. In Western countries ‘humanitarian intervention’ rather than traditional interstate war became the primary form of the use of military force. As a corollary, the normative debate shifted from one in which the ethical acceptability of a system of nuclear deterrence and possibilities of overcoming it took a central place towards a debate on the pros and cons of intervention. In this debate ‘neo-interventions’ came to face ‘neopacifists’. The former emphasize the right and responsibility of the international community to seek redress for gross violations of human rights while the latter focus their arguments on the risks and potentially negative effects of using military force. This concern gave rise to the ‘zero-dead doctrine’. This doctrine is based on the theory that due to a number of changes a situation has arisen in which not only it has become technically possible to wage ‘war without bloodshed’, but in which this is also the only form of military force, which is still thought to be acceptable in democratic systems, except when direct threats to immediate national interests are involved. Humanitarian interventions are generally seen as lacking the quality of such an immediate interest, and thus the public is seen as hardly willing to run the risk of casualties. To some, this (alleged) unwillingness of the public to envisage the incidence of casualties represents an obstacle to a sensible and effective foreign policy and the pursuit of the national interest, which should therefore rather be left in the hands of trusted experts. Others, however, fear that if the concept of a ‘war without

2 bloodshed’ appears a real possibility, the public may no longer be willing to restrain the hands of their governments in decisions concerning the use of force (Ignatieff (2000)). This new situation forces us, therefore, to take a new look at the ancient debate concerning the nature, content and influence of public opinion and its implications for the democratic model concerning foreign policy in general and the use of military force in particular. Strengthening the empirical basis on which it is carried out may further the normative debate briefly referred to above. At least that is the basic normative rationale behind the research programme of which this paper forms a part. Our knowledge of the nature, content and impact of public opinion in this area of study may have increased considerably over the years, but it still leaves a lot to be desired. Fundamental gaps in our understanding remain.

Limitations of the debate; a new research programme

In empirical terms, much of the debate has remained limited in three important ways. Firstly, much of the evidence has been based on data from American public opinion and its relationship to US foreign policy. Less attention has been devoted to European publics. A lack of comparable data across nations and across time still poses severe limitations to our understanding of European public opinion and to our ability to extrapolate available findings across nations, although the gaps in our knowledge concerning public attitudes in Europe on foreign affairs, including the use of military force, are beginning to be filled. Secondly, research has not yet expanded much beyond the examination of the opinion-policy connection relevant to and in the context of the Cold War. As said, the problem of war and peace has fundamentally changed, however, as a consequence of the end of the Cold war and of the emergence of new problems. These changes in the international environment offer a unique opportunity to examine to what extent mass beliefs and attitudes have changed over time in connection with changes in the political landscape or have remained stable, and to specify the extent to which generalising conclusions reached based on analyses of the Cold war period still hold today and into the future.

3 This means that a much greater and systematic cross-national effort is needed to increase our understanding of the crucial relationships between public policy and public opinion in a new world of international politics. Such research should cover a wide spectrum of democratic regimes and a diverse set of issues. Thirdly, this debate has generally overlooked the crucial role that (expected or actual) casualties of war and casualty-related considerations have come to play in both the calculations of decision-makers and in the formation and perception of mass opinion regarding the international use of force. The research project mentioned above aims to contribute to fill these three particular gaps (Shapiro, Nacos and Isernia (eds) (2000); Everts and Isernia (eds) (2001); Everts (2002)). It is deliberately comparative, offering analyses across a wide set of countries (superpowers, medium and small size powers, countries facing the risk of or actual involvement in war as well as countries involved in peacekeeping operations). It aims to explore a variety of situations in which the international use of force is either decided upon or contemplated. It attempts to look at the role of public opinion at both the individual and aggregate level. The conflict over Kosovo of 1999, however horrible in humanitarian terms, offers a unique opportunity to study international attitudes, particularly on the use of military force, in a situation that differs considerably from earlier situations, either that of traditional interstate war, like the Gulf conflict of 1990-91 or traditional peacekeeping operations.

The conflict over Kosovo, 1999

During 1998 and early 1999 several efforts had been made to change Yugoslavia’s policy on Kosovo and international pressure in the form of threatened air strikes had been brought to bear on Milosevic on at least one occasion.1 Although the Security Council had not authorized the use of force yet, it had continued to ask for fulfilment of the demands of the international community to stop the violations of human rights in Kosovo. A last round of international negotiations, held at Rambouillet, failed to have the desired results, however. Critics have argued that this was the very purpose of the United States and its allies. Anyway, after the failure of Rambouillet NATO decided to carry out its threats to the government of Yugoslavia, under Milosevic, to punish it for

4 the persecution of the Kosovars, and to persuade it to change course and withdraw its military forces from Kosovo. The first of a series of air attacks started on March 24, 1999. The conflict lasted until June 10 of that year. One may argue – as the critics tend to see it – that the conflict over Kosovo in 1999 and the military actions undertaken by NATO and its allies in that context was merely the last typical war of the twentieth century. Others, however – including NATO itself - argue that it was an entirely new kind of war, indeed the first of the twenty-first century. In the former view it was only the most recent manifestation of the traditional struggle over power and influence in the Balkans, over spheres of interest and competing claims of identity and nationhood. In the latter definition it was the first example of a new and rather different kind of international use of force, being not aimed at protecting or furthering traditional national interests, but at protecting people and their fundamental rights. Protagonists of this humanitarian interpretation like Messrs Clinton and Blair would like to see it as paradigmatic and as a model for the future. Indeed, one of the reasons for NATO’s actions was to ensure that its threats would not be seen as hollow gestures and thus to deter others from taking such actions as taken by the government of Serbia against the Kosovars in the future.

The essential role of public opinion

The evolution of international involvement in the crisis and conflict over Kosovo has clearly shown, again, the complexity of the relationship between public opinion and decisions to use military force. More generally, it has emphasised the intricate ways in which contemporary democracies deal with the eternal tensions between the demands of peace and the demands of justice. As in all cases involving the use of military force – but perhaps even more than in other cases - the legitimacy of the action in the eyes of the public was at stake and the importance of public opinion was highlighted. It has often been argued that today the public is only willing to undertake and support risky military operations if national interests are directly involved (Everts 2002). Clearly, the conflict over Kosovo was no exception to the rule that, the public, stimulated by media reporting, is horrified, be it often selectively, by real or perceived injustices and violations of human rights,

5 and demands from their governments ‘to do something about it’. At the same time, there are enormous variations in what are seen as the best ways to deal with these problems, particularly when the question of the use of force arises. Governments may be mistaken in thinking that public opinion acts as a narrow constraint on the use of military force for purposes other than the protection of immediate national interests. But they can also be forgiven for thinking so, because a superficial reading of the data often suggests the existence of a strong reluctance among the public to countenance the possible consequences of warfare. Whatever the case, this perception can serve as a useful alibi to avoid taking responsibility for risky actions. It is not surprising that more than in any earlier recent conflict the question of the degree of public support for the various options and strategies available took a central place in the public debate and the calculations of decision-makers.2 In this connection, it is notable, for instance, that the Clinton administration since the very beginnings ruled out the use of ground forces in the case of Kosovo, allegedly for fear of jeopardizing public support. It is therefore interesting to assess to what extent the governments’ concerned did indeed or did not enjoy public support from the beginning to the end. It is here that public opinion research and the careful analysis and interpretation of available data can be both theoretically and practically relevant. Studying this case may help us to understand the relationship in this respect between governments and citizens, particularly the degree to which, in general, public opinion acts as a constraint, a driving force, or rather as a permissive consensus and as a factor which can easily be manipulated or even neglected without fear of retribution or punishment in questions of war and peace. As has been elaborated in more detail elsewhere, in many aspects the issue has not yet been settled and the jury is still out (Everts and Isernia (eds) (2001)), This applies both in general and in the specific case of Kosovo. Given the nature of the case and the role that public opinion allegedly played in the considerations of decision makers it is not surprising that a great many polls were made available during the war. Compared to previous wars – e.g. the Gulf war – on this occasion American pollsters were joined more by Western (and Eastern) European colleagues in taking the pulse of the public. From the pollsters’ viewpoint, it is probably not far from the truth to say that in the Kosovo war we saw the first manifestation of some sort of ‘world opinion’ (Rusciano 1998).3

6 Modelling the factors shaping the support of the use of military force and their impact

What in general determines the willingness of people to (continue to) support the international use of military force? It has gradually become evident from studies of public opinion in either specific historical cases or more experimental, laboratory-type surveys of public opinion that – particularly when looked at in isolation – (1) the actual or potential incidence of casualties (be they military or civilian) has a considerable, albeit not decisive, negative influence on the willingness of publics in democratic societies to enter into military conflict, or to continue to do so (See Everts (2002), ch. 9 for a survey of the literature and available evidence). This unwillingness is mitigated or mediated, however, by other factors. The willingness to run military risks and accept the consequences is influenced positively by (2) the degree of threat to what are seen as vital interests, material and other.

People and states that are not directly and immediately threatened may always ask,

moreover why they should be the ones to ‘mourir pour Danzig’ when it is not certain that others will share the burden, and when one may gamble that the collective good will be produced anyway (in this case: by the US); (3) the (real or perceived) effectiveness and success of the applied force and, finally (4) the quality of and/or confidence in the political leadership and the wisdom of government policies on these matters. If the case isn't made, support doesn't materialise. The three latter factors can mitigate (if present) or amplify (if absent) the impact of factor (1) (Everts and Isernia (eds.) (2001); Everts (2002)). Surveys and experiments in which the impact of fear of or opposition to casualties is measured in isolation from other factors tend to lead to an overestimation of this factor. The alleged ‘body bag syndrome’ serves as useful alibi to politicians and decision-makers and helps them to avoid having to take responsibility for what indeed often are awkward decisions in highly uncertain situations. When looked at in combination with other factors, especially when people are faced with trade-offs and alternative courses of actions, it becomes clear, however, that, despite conventional wisdom, public support for military action is not a knee-jerk reaction (Feaver and Gelpi (1999); Kull and Destler (1999); Kull and Ramsay (2001)).

The quality of the data

7 It is by now common wisdom that responses to survey questionnaires can be heavily influenced by the nature and wording of the question. When we add the sheer lack of knowledge of and attention to political issues by many respondents to the oddity of the interview experience, the warning, issued by Mueller on several occasions (e.g. 1973 and 1994), that a single response percentage has not much meaning in itself, is well justified. To get an idea of the forest, one has to walk deep inside it, looking at the different trees, rather than stopping at only the first one met. In this context, the importance arises of comparisons to the understanding of public mood on any political issue. To deal with interpretive problems arising from what the public mood is on an issue, it is particularly useful to adopt a comparative approach (Mueller 1973, p. 10) and, more specifically, to compare identical questions over time and differently worded questions at the same time. In fact, turning a liability into an asset, the very sensitivity to question wording might be helpful in clarifying the conditions under which a certain stimulus produces its effects. Our book, beyond the traditional problems of comparing surveys over time and across different wordings, has to face a further obstacle, springing from the cross-national nature of our comparisons. We will examine polls conducted in several different countries. The crossnationally comparative slant adds further complexity and ambiguity to any interpretation of those results at the aggregate level. In fact, comparing aggregate public opinion across countries brings into effect a further source of difficulties: the comparison of different questions asked in different political and cultural contexts. There are two aspects to this problem. On the one hand, there is the problem of comparing identical questions asked in different languages and the equivalence among them (Verba 1971). Apart from this question of comparing similarly worded questions, there are four truly comparative surveys during (or immediately before) the Kosovo war: the surveys carried out by Angus Reid for CNN on April 22-25 1999 4 and for The Economist on March 25 to April 17,5 the Gallup Eastern European Poll in early March,6 before the bombing started, and the IPSOS poll for Libération on June 2 1999.7 On the other hand, there is the even more complicated problem of comparing differently worded questions asked in different languages on the same issues. The majority of the results presented here come from such data. In the latter case, the problem of comparing questions translated in different languages is added to that of comparing differently worded questions. Taking into account those constraints, the analysis that follows below is based on available polling data. Although an effort has been made to be as thorough and complete as

8 possible in collecting the data, it is not unlikely that other polls have been held that were overlooked. These were therefore not included in the analysis, although they might have thrown additional or perhaps even a very different light on the issues discussed in this book. As in other cases, much more polling has been done on the Kosovo war in the United States than in Europe. A survey of the available raw data is presented in the Appendix. As far as Europe is concerned, coverage is also, in relative terms, very uneven. This forces us to be selective with respect to the former and more speculative in the latter case. Also, to the extent that we shall compare and generalize across countries we shall have to bear in mind not only the possible disturbing effects of time or differences in question wording, but also of different national politics. It is no mystery that governments and political elites in the countries involved in the military operation reacted rather differently to the event (Weymouth and Henig 2001), and this is reflected in the domestic political debates in these countries as well as in their public opinions. As far as the impact of public opinion on policy-making is concerned, the distinction between salient and non-salient attitudes is vital. The willingness to act upon one’s convictions and participate in the political process is proportional to the degree of saliency and thus equally relevant as the content of opinions. Yet, saliency is an aspect which many, if not most opinion polls conveniently overlook. With these caveats in mind, let us proceed.

Outline of this book

etc

1

Chapter 3a Trends in Western Public Opinion Introduction

In this chapter we intend to explore, on a cross-national comparative basis, trends in support for the war operation and to speculate on the possible sources of variations in support for air strikes and other military or diplomatic strategies. First, there is the question of support for the various military strategies that were available to – or actually employed by – NATO to force the government of Yugoslavia to stop its policy of ethnic cleansing and abide by NATO’s demands. Available data focus primarily on the general support for the air strikes undertaken by NATO, but we are also able to pay some attention to support for alternative strategies, discussed at the time to deal with the problem: Should NATO, for instance, do more of the same or send ground forces, or rather return to negotiations? From the beginning, critics questioned whether the military actions would be effective (and hence justified) without sending ground forces into the conflict, or at least showing a willingness to do so. It was argued repeatedly in this connection, however, that the public, while possibly supportive of air strikes, would not accept this extension of the war for fear of casualties and, hence, that NATO governments were operating under severe constraints. Others criticized the bombing for quite different reasons. Arguments were voiced from the beginning, becoming more intense as the conflict went on and bombing failed to have the intended results, that bombing should be replaced, or at least accompanied, by further efforts to bring about a negotiated settlement. Available data on this question will be discussed as well. The chapter concludes with some considerations on the reasons for the remarkable stability of attitudes toward the war that were found.

Support of the NATO bombing actions

Levels of support

2 Having concluded that earlier efforts to arrive at a peaceful solution of the conflict had failed, At 19.00 GMT (8.00 pm in Belgrade) of March 24 1999, NATO launched its first salvo of air-strikes against Yugoslav military targets. Fearing adverse public reactions and hoping this would soon bring the desired results, it was decided to start air attacks aimed initially at military targets only. It was a risky strategy. ‘High tech warfare is governed by two constraints – avoiding civilian casualties and avoiding risks to pilots – that are in direct contradiction. To target effectively you have to fly low. If you fly low you lose pilots. Fly high and you get civilians ... Six weeks into the war the deployment of effective Apache helicopters had not yet been approved out of fear that the risks to crews were too high ... As the air campaign went into its second month, the alliance’s moral preferences were clear: preserving the lives of their all-volunteer service professionals was a higher priority than saving innocent foreign civilians,’ (Ignatieff, 2002: 62). Faced with the probability that in a television age, the fact the public would become aware of the damage done to Serbia, could entail heavy political costs: would the public at home continue to stand rising civilian casualties if the bombing was not having any discernible effect?’ How did the public react to this? If put in the form of a simple question concerning support for these actions, the public’s reaction to this decision in most of the countries involved was fairly differentiated, as shown in the three genuinely comparative surveys, held at different stages of the conflict (table 1). The first was carried out by Angus Reid group for The Economist between March 25 and April 17 in sixteen countries. A second, by ICM for The Guardian in May, in twelve countries and a third by IPSOS on June 2 in eight countries. These are the only surveys in which an identical question (translation problems aside) has been asked at the same time in more than one country. Fortunately, the three questions tap support for air strike in a quite similar fashion. They all refer to the NATO operation and asked whether the respondent supported or opposed this operation, rather than his/her own country’s participation in it – a slight difference that is not without consequence, as we will see soon. We will therefore start our analysis with this crossnational comparison of the level of support for NATO operation. Since we are not always able to be precise about the exact day each country’s survey was taken, some differences might well be

3 related to events occurring at the time the survey was carried out. However, as we will see later, it does not seem that the events during the war had much of an influence on the degree of support. All this said, starting the day after the bombing began, Angus Reid surveyed public opinion in a set of 16 countries whose reactions to the events were quite differentiated. Simply looking at the proportion of those who ‘support NATO’s decision to carry out air and missile attacks against Serbian military installations’ (table 1), we can distinguish at least three groups of countries. A first group is made of those countries showing a strong majority in support of the war. Support stood at the two-thirds level in the US (68 percent) and the UK (68 percent), and was also very strong, at the time, in Denmark (74 percent), Norway (64 percent), the Netherlands (68 percent)1 and Canada (64 percent). Not surprisingly, the most enthusiastic in watching their former enemies being pounded by NATO air strikes was the Croatian public, with an 82 percent endorsement.2 A second group of countries showed mixed feelings about the NATO action. Three of the older NATO members – Germany, France and Italy – found themselves in this middle group, together with Poles (54%),3 Hungarians (48%)4 and Fins (50%). Public opinion in these countries was divided, with slim majorities or large pluralities of the polled in support of the air strikes. A third group was made up of those countries in which the public was clearly opposed to the NATO operation. In the Czech republic, for instance, those against the NATO operation outnumbered those in support with a margin of 3 to 1. Russians, among which 94% of those polled opposed,5 were joined by Ukraine (89 percent) and Slovak (75 percent) citizens in nearly unanimously opposing the NATO air attack. [TABLE 1 ABOUT HERE] Table 1 - International public opinion and the conflict in Kosovo. Views on NATO actions

The ICM-The Guardian 12-countries survey of May 1999 shows a similar pattern, with a few new countries added to the picture. In Denmark (70%) and Luxembourg (61%) a majority supported ‘the military actions by NATO in Serbia’. Belgium (53%),6 Austria (41%), Finland (44%), Germany (52%) and Ireland (46%) were in the middle, with a slim majority or a plurality of the polled population supporting the NATO operation. In Spain (34%)7 and Italy (37%) no more than one third supported the NATO operation. Greece, with its historical ties to Orthodox Serbia, was clearly the odd man out in NATO. While the Greek government continued to go along with the NATO action, this policy was almost universally opposed at the mass level. The

4 two countries in which we have some changes – in opposite directions – are France and Great Britain. In France, the May polls records a majority of 68 percent in support of the war – an increase in support in respect to the Angus Reid poll – while in United Kingdom we have the opposite, support drops to 54%, still a majority but much less than at the beginning. The third comparative survey, by IPSOS for Libération, was carried out on the verge of the acceptance by Slobodan Milosevic of the peace plan proposed by the Russian, American and Finnish representatives and this can explain a slightly higher level of support for the NATO operation. However, there are still interesting differences among the polled countries. On a question very similar to the one asked by ICM the month before, high support for the NATO operation is found in France (62%) and the United Kingdom (67%); intermediate in Germany (54%) and Italy (51%), with a plurality or slight majority opposing it in Belgium (45% opposing) Portugal (51%) and Spain (49), and definitely strong opposition is found in Greece, with 97 percent not approving the NATO intervention. To this long list of countries, we can add data from Sweden, where on March 31, 39% were opposed to NATO air strikes and 34% in favour.8 Table 1 shows not only the differentiated response to the NATO operation among European countries and the United States, but also sheds some light on the relative stability of the publics in the different countries. With the exception of Belgium, and also discounting for the different question wording, public opinion did not change much during the war. A conclusion based on comparing a few points in time of differently worded questions does not, however, represent a firm conclusion. Therefore, we now move from a comparative analysis of level of support across different countries to a trend analysis of the evolution of support over time. Not surprisingly, the number of countries for which trend data are available is less. In fact, subscribing to an admittedly very liberal definition of trend data, as two time points in which the same (or similarly worded) question has been asked, we have enough trend data available for only a few countries: France, Germany, Italy, United Kingdom and the United States. What emerges from these trends? The general pattern is remarkably similar for all countries concerned, given the national differences in level of support among the five countries stressed previously: slow or no increase in support up to mid-April and a slight decline afterwards, probably due to growing disillusionment about the lack of success of the air strikes. In Germany and Italy, however, support for the war increases in late May, as the war approaches its end.

5

Evolution of support over time We now pass to analyse these trend data in detail, starting with US, the country whose material is much more abundant and diversified than for the any of the other four countries for which trend data are available. In the United States (table 2) we found 21 questions aimed at assessing the public’s support for the NATO air strike against the Federal Republic of Yugoslavia, asked by 10 different survey organization in different periods of time. The longest series is a 9-points question by Gallup (question c) asking ‘As you may know, the military alliance of Western countries called NATO, has launched air and missile attacks against Serbian military targets in Yugoslavia. Do you favour or oppose the United States being a part of that military action?’. The second longest series is a 7-points one by CBS News (question f), demanding ‘Do you favour or oppose the United States and NATO conducting air strikes against Yugoslavia [Prior to 5-6 April ‘. . . against Serbia?’]?’. Then, there are two 6-points series both posed by ABC News (question j and q) and one 5-points series (question u) asked by Opinion Dynamics for Fox News. Of the remaining questions, 7 were asked only once, 6 twice, 2 were asked three times and one four times. [TABLE 2 ABOUT HERE] Table 2 – US public opinion and the conflict in Kosovo. Support for air strikes

All in all, four questions were asked before the war started and they constitute an interesting baseline against which to compare the war trends. Before the war started, American public opinion was divided on the desirability of using air force against Serbia. The first question available was asked on 9-12 October 1998, at the height of an exchange of threats between NATO and Milosevic. On that occasion, Gallup asked a sample of 1004 Americans: ‘Based on what you have read or heard, do you think the United States and its Western European allies should or should not conduct military air strikes against the Serbian forces in Kosovo?’ 42 percent were then in favour, 41 percent opposed and 17 percent had no opinion. The second question was asked five months later, when the situation was less tense and negotiators were on their way at Rambouillet. Attitudes toward the use of force were, however, remarkably similar to those of October 1998. The Gallup survey of February 19-21 1999 found 43% in favour of the

6 US being part of a military action against the Serbian forces in Kosovo, 45% opposed and 12% did not answer. A similar question asked one day before the attack, March 23, by ABC news shows 47% in favour of an air attack, 47% opposed and 6% uncertain. Apparently, a lack of explicit reference to NATO or the European allies in the question decreased the support, as a glance to an ABC News question asked on March 11-14, 1999 shows. To the question ‘The United States has said it may bomb Serbia unless Serbia agrees to a peace plan for Kosovo. If Serbia does not agree to the peace plan, should the United States bomb Serbia or not?’, only 26 percent were in favour of US being part of the action. This difference might be only in part due to expectation of peace embodied in the resumption of the Rambouillet talks on March 15, since the survey took place in a period – March 11-14 – in which the Serbian government was showing its unequivocal opposition to the terms of the peace agreement as agreed upon by the other parties at Rambouillet. The fact of the launching of air strikes of March 24, 1999 had no clear-cut effect on support for the NATO air operation. Contrary to the Gulf war (Mueller 1994, pp. 70-73), in the case of Kosovo we do not, therefore, find much evidence of a ‘rally around the flag’ effect among the American public (Brody and Shapiro 1989). This is a likely consequence of the divisions existing among American political elites on the appropriateness of the initiative and of the reluctance of the Clinton administration to be embroiled in the Serbian situation. No matter how the question is formulated, support for the war hovers around 50% in the first days of the war, and it does not increase much after that. The only exceptions are two questions, asked respectively by Princeton Survey Research Associates and Rasmussen Research that get a 60% support for air strikes. The explicit reference to the President in the Rasmussen question of March 24-28 might explain that greater support, even though the public’s rating of President Clinton’s handling of the situation in Kosovo at that time was lower, around 50% those ‘approving’ Clinton’s ‘handling of the situation’ (ABC News) and no more than 39% and 38% those rating President Clinton’s handling of the Kosovo crisis as ‘excellent’ or ‘good’ (Rasmussen on March 19 and 25 respectively). As to the PSR Associates question of March 24-28 (question g), the reason why it produces a higher level of support is probably the explicit reference in the text to both the peace agreement and the end of fighting. Looking at the four questions with enough time points to allow us examine the evolution of support for air strikes over time, we find, with some slight differences, a strikingly similar

7 pattern. Support for the US participation on air strikes against Serbia starts around 50 percent at the beginning of the war, moves slightly up to 60 percent by the half of April and then it begins to erode steadily and slowly or it returns to the level it had at the war’s beginning. As an example, in both the Gallup and CBS News series (question c and f respectively) support for air strikes moves from 50 percent on March 25 to 61% (59% for CBS) on April 13-14, it goes back to around 50 percent between late April and early May and it declines further to 49% in half or late May. A similar pattern is found in the two ABC trends (question j and q), at a slightly higher level of support. Looking at formulations of the different questions, we find an interesting pattern. Before the war broke out, as mentioned earlier, the explicit reference to NATO and European allies increased the support for the air operation. Once the war is going on, an explicit reference to NATO and the European allies seems to have an opposite effect on support: depressing it. A question phrased such as to show that the operation is run by the US (in one case with the help of its European allies, as in question j) apparently gets a higher level of support than a question in which the Kosovo war is presented as a NATO operation in which US is only one of the partners (questions f and i). This is further confirmed by looking at one question in which no explicit reference is made to the US participation and only NATO is mentioned (question h). Here we find the lowest percentage of support for the air strikes ever, with only 44% in favour and 40% who disapproves and 16% not answering. Apparently, after the beginning of the war, the idea of having the US forces to depend on NATO or the European allies had a different effect from before the war broke out. Once the war starts and the allies are side by side with the US, the American public seems to think that the US air forces should run the show (things would be different for ground troops, as we will see soon). Moving now to examine the four European countries for which trend data are available, we find similar patterns: no rally at the beginning, a stable or slightly increasing support between late March and the first half of April followed by a progressive erosion of support. In France, where the public was divided almost equally among those in support and those opposed to it at the beginning of the war, one witnesses an increase in support during the first half of April then followed by a progressive erosion of support. A similar, but less pronounced pattern is found in Germany, where support for both the NATO operation as such and for ‘the Bundeswehr also taking part in the air attacks of NATO’ [Die Woche] was at much the same level,9 around 60

8 percent, at the beginning of the war, to increase only slightly in the first half of April, and then declined through May, when it eventually turned slightly up again. In Great Britain, the set of questions covers a shorter period (until May 1st) and is therefore less easily comparable. However, we find also in the British case that support first increased and then slightly decreased, with a majority still in favour, however. In Italy, a trend similar to Germany is found. Support slightly increases during the first part of the conflict; it reaches its maximum around the half of April and then steadily declines, to increase again once the end of the war is in sight. [TABLE 3a-3d ABOUT HERE] Table 3a-3d – European public opinion and the conflict in Kosovo. Support for air strikes

As we have seen before, in the United States the public apparently differentiated between support for the NATO operation and support for the American participation in it. It seems that the European respondents made a similar difference between the actions undertaken in general and the participation by one’s own armed forces. In France, like in the US, support for French military participation was even higher than support for the NATO operation (with no mention of France). While in a series of surveys the percentages of those who ‘personally approve [or disapprove] of the military intervention by NATO in Yugoslavia’ were 57, 63, 72, 74 and 63 respectively; they were 59, 65, 73, 74 and 67 for those who ‘approve [or disapprove] the participation of France in this intervention by NATO in Yugoslavia’.10 Moreover, apparently, in France support for ‘the military intervention’ is higher than support for ‘air strikes’. Whereas support for NATO military intervention and, even more, French participation in it was always well beyond 50 percent, when asked whether they agree with ‘air strikes against Yugoslavia’ or ‘with the participation of France in the air strikes against Yugoslavia’, support dropped of 16 and 13 points respectively. Also here, however, support for French participation in air strikes is higher than support for NATO air strikes:11 40, 50, 55 and 50 were in agreement on the first question while 46, 58, 59 and 57 on the second one.12 In the Netherlands too, while in one poll only 47 per cent supported the dispatch of ground troops, 58 per cent agreed that if this should happen nevertheless the Netherlands should commit its troops.13 In Italy, before the air strikes 68 per cent preferred that the crisis should be solved through dialogue and only 27 per cent supported more drastic actions, but in the same poll 49 per cent felt that Italy should support NATO if it would decide to bomb Serbia.14 What we

9 observe, somewhat surprisingly, seems to be the opposite of what the ‘free-rider’ argument would lead us to expect.

Send ground troops?

Across most NATO countries a considerable gap existed in general, at least initially but probably throughout the conflict, between the generally high support for the bombing actions and the much smaller support for the alternative or complementary strategy of sending ground troops, should bombing turn out to be ineffective. On average, this gap was about ten per cent, but, as can be seen in Tables 4 and 5, that reports cross-national questions asked respectively by Angus Reid in two different surveys close enough to allow for a comparison and by ICM for The Guardian, considerable differences existed among the various countries. The occurrence and size of the gap is apparently dependent, among other things, on the phrasing of the questions concerned.

[TABLE 4-5 ABOUT HERE] Tables 4 and 5 – Support for air strikes and ground operations in several countries

Tables 4 and 5 suggest that in those countries where support for the air strikes was (relatively) low there was not much difference between the number of supporters of air strikes and those of ground troops, because these were ‘hard core’ supporters who would be prepared to accept almost any (military) means, whereas the opposite was the case in countries with high levels of support for the air strikes, where supporters included those who really had mixed feelings and certainly wished to do no more than bombing. Hence the gap. According to other polls, the gap was fifteen percentage points in Belgium,15 24 in Canada,16 in Denmark 19, the Netherlands 29,17 in the U.K. 18 (in other polls, however, 15 and 39 percentage points respectively). In Italy it was between eleven and seventeen percentage points according to various SWG surveys. In France it was only between five and eight points in various polls, and in Spain it was non-existent.18 In the United States, where the greatest number of questions were posed, the overall gap is around 10 percent points, but it can be higher or lower depending on the wording of the question.

10 The existence of a gap is one thing; the absolute level of support for sending ground troops is another. Looking at tables 4 and 5 one is struck (again) by the differences between the United Kingdom, France and the US on the one hand and Germany and Italy on the other hand. While a majority could be marshalled in support of sending ground troops if necessary in the former group of countries (more clearly in the British case, less so in France or the United States), only one third or even less in the Italian and German case were in favour of this. Considerable opposition, though less strong than in Italy, also existed in other countries for which no trend data are available: Belgium, Canada and Norway took a middle position in this respect. The situation in the Netherlands, again, differed somewhat. While early in April a gap existed as mentioned above of some 20 percent with 47 percent in support of ground forces, there was a clear majority (58 percent) that felt that the Netherlands armed forces were to participate if such a decision were to be taken.19 Taking these differences into account, one might think that the gap would have diminished over time, as it became evident that the bombing campaign did not have the immediate effects that were both expected and promised to the public by NATO, and thus in the form of decreasing support in general, or, inversely, in the form of mounting sympathy with the idea that ground troops would be necessary, both to shore up NATO’s overall loss of prestige and to effectively provide protection on the ground to the persecuted Kosovars. Neither of the two happened, at least in France, Germany, Italy or the United States, for which time series data are available. Starting again with the United States (table 6), depending on the way the question is formulated, one can point to a slight increase in support for the ground operation (question o), a slight decline (question k) or a pattern similar to that found for air strikes: a slight increase at the beginning and then a decline by April (question e). In all cases, however, there is no sign of escalating sentiments (Larson, 1996; Mueller, 1985) in favour of the use of force. Support for the ground operation is always lower than for the air strikes, it does not move up dramatically and it divides American public opinion, with only pluralities or slight majorities favouring it. The highest level of support is 57 percent at the beginning of April (by the way in the period in which also air strikes’ support is at its highest), a possible effect of the capture of three American soldiers patrolling the Serb-Macedonian border. [TABLE 6 ABOUT HERE]

11 Table 6 – US public opinion and the conflict in Kosovo. Support for ground operations

An interesting difference emerges comparing questions in which NATO and the European allies are mentioned with those in which only the US is mentioned. As we have seen in the previous section, once the war started, support for air strikes was higher when measured by questions in which the US are clearly mentioned as the leaders in the operation. On the contrary, support for a ground operation seems to work the opposite way: it is higher when NATO and the European allies are mentioned and lower when the question constructs the use of ground troops as a duel between the Serbs and the Americans (as in question d or j). In fact, on average no more than 28 percent answered yes to the question ‘Should the U.S. send ground troops into combat against the Serbian Army in Kosovo?’ [Opinion Dynamics/ Fox News] or to the question ‘Would you favour or oppose sending American ground troops into Kosovo?’ [Rasmussen]. On the contrary, to the question ‘Suppose the bombing does not stop Serbia’s military action in Kosovo. Would you support or oppose the United States and its European allies sending in ground troops to try to end the conflict in Kosovo?’ [ABC News], an average of 55 percent said yes. And to the question ‘If NATO (North Atlantic Treaty Organisation) decides to send in a multinational force to protect the people of Kosovo from the Serbs and the Yugoslav army, would you support or oppose sending US (United States) troops into Kosovo as part of this NATO force?’ [Louis Harris Associates], 59% said on average yes. Considering these data, one is left to wonder whether Americans deem air warfare something they are specialized in and better equipped to do alone, whereas ground warfare – entailing higher costs and risks – need the Western Europe Allies’s support. On the other hand, the last two questions contain in their banner a reference to other considerations – such as the ineffectiveness of the bombing and the protection of the Kosovo people – that might positively affect the level of support. Looking now at the European countries, the picture is closely similar to the American one, given the national differences in degree of support for the war. In Germany and Italy, where there was a consistent gap respectively of some 30 and 15 per cent between the level of general support for the NATO actions and the willingness to contribute German or Italian ground forces should this be necessary, there is no movement across the entire time period. In France and Great Britain, where the gap was much smaller (respectively from 5 to 8 percent and from 14 to 3 percent) between support for the NATO actions in general (and/or French or British participation

12 in it) and the willingness to eventually send ground troops, the trend is similar to that observed for air strikes. Support for the use of ground troops increases slightly in April and then declines steadily over time.20 [TABLE 7a-7d ABOUT HERE] Table 7a-7d – Public opinion and the conflict in Kosovo. Support for ground operations in various European countries

Bombing or other strategies?

Restricting oneself to support for the military action could hide a significant public demand for other strategies, including a greater NATO emphasis on diplomacy. Throughout the conflict several countries (e.g. France, Italy, Russia) attempted to find a diplomatic way out of the bombing strategy. The support that these initiatives might have found among the public are shown in several poll data from the first weeks of the conflict (Table 8).21 [TABLE 8 ABOUT HERE] Table 8 Which role should NATO play in the conflict around Kosovo/Yugoslavia? (in %)

When asked by Angus Reid ‘What role should NATO play in the conflict’, it was widely agreed that NATO had a role to play in the conflict, to begin with, since only 14 percent said NATO should reduce (4 percent) or end (10 percent) its involvement in the Kosovo conflict. Given a choice, however, NATO’s military actions were not generally accepted. A plurality of 39 percent of the polled citizens of NATO countries said ‘NATO should do more to seek a diplomatic solution, but should not send more troops’. This was the choice of at least a plurality in each of the eleven NATO-member countries polled – ranging from 59 percent in Italy and 51 percent in France to a narrow 30 percent in the US and 33 percent in the UK. On the whole, one in five (22 percent) NATO-country citizens, on the other hand, opted for sending more armed forces, but another one in five (21 percent) said that NATO should keep its involvement at current levels. Not surprisingly, the non-NATO members that more opposed to the war, namely Russia, Slovakia and Ukraine, were definitely in favour of a complete stop of the operation. In a similar vein, in April 3, 5 and 19 and in May 19, 1999 Rasmussen Research asked ‘what should America's involvement in the Kosovo dispute be? Should we participate with air

13 strikes only, with air strikes and ground troops, with diplomatic efforts only, or should we have no involvement?’. Approximately one fourth of the respondents mentioned ‘diplomatic efforts only’. Interestingly, a Princeton Survey Research Associates set of survey for Newsweek asked between the end of April and the early May, ‘Right now, what do you think should be the priority for NATO and the U.S. in Yugoslavia: continuing the air campaign at its current level, stepping up the air campaign, launching a ground war against the Serbs, or stepping up diplomatic efforts to reach a negotiated settlement?’. Between 52 and 62 percent of those polled answered to step up diplomatic efforts. That the public had truly mixed feelings became most evident in Italy. Apart from Greece, opposition to the actions was most pronounced in this country. 43 percent said that it should ‘refuse to participate in military actions’, only 36 percent were in favour. On the other hand, 52 percent agreed with the policy of the Italian government, described in one question as ‘to support the military operations but at the same time take autonomous peace initiatives’. Thus, when confronted with a number of alternative policy options, the most preferred (by 44 percent) option was ‘to take part in the military actions decided by NATO but at the same time promote peace initiatives’; 37 percent opted for ‘remain in NATO, not participate in military actions but promote peace initiatives, while only 8 percent opted for ‘take part in the military actions decided by NATO’ and an equal number wanted to leave NATO altogether.22 The strong and stable support for the policy of ‘return to diplomacy’ in Italy comes out very clearly in Table 9. [TABLE 9 ABOUT HERE] Table 9 - Bombing or return to diplomacy? Opinions in Italy

In another Italian poll, large majorities showed themselves in favour of a unilateral suspension of the bombing, especially if Milosevic were to indicate his willingness to withdraw Serbian troops from Kosovo if such were to happen (SWG, 24 May 1999). This preference for diplomatic initiatives is in line with the strong scepticism about the wisdom or desirability of an intervention by ground forces in this country shown above. Similar levels of preference for negotiations, however, can be found also among the US public, looking at a sparse, and quite diverse, set of questions asked on terms for a possible peace agreement between NATO and the Serbian government. At the beginning of the war, in a PSR Survey for Newsweek (N=752) 62 percent thought NATO and the US had offered Milosevic ‘enough in return for agreeing to

14 restore autonomy to Kosovo’ (and 15% thought ‘not enough’ and 23% did not know). And on April 8-9 to another PSR survey for Newsweek (N=751), 57 percent of the interviewed answered that NATO should stop air strikes only when Milosevic were to ‘agree to all current NATO demands’ (while 13% would have stopped the air strikes only when Milosevic began withdrawing forces and another 24% said ‘immediately’). However, on May 1-2, to a CBS News poll (N=1151), 67% thought ‘the United States and NATO should negotiate a compromise with Slobodan Milosevic in order to end the fighting in Yugoslavia’ (while 26% thought NATO should not do so and 7% did not know). In early May, different polls reveal that the majority of the American public (between 50, 52 and 58 percent depending on the question) was already willing to negotiate with the Serbs rather than continue the bombing. This was confirmed by French polls in which 56 (CSA 17-18 April) and 58 percent (CSA 6-7 May 1999) opted for the statement: ‘stop bombing and start immediate negotiations’. Understandably, more effort towards a diplomatic solution was also the preferred option across the six non-NATO countries as a whole (42 percent) – except that Croatians tended to favour the commitment of more forces (39 percent) and fully half (50 percent) of Russians urged NATO to end its involvement entirely.

One concluding puzzle

This comparative overview of the available polling data on the Kosovo war shows that, apart from cross-national differences at the absolute level, public opinion on all the central issues did not change very much during the entire conflict. It is remarkable and it needs further analysis how it was possible that support for the NATO actions remained stable and how NATO was able to sustain this support in spite of the widespread ‘collateral damage’, the fact that results were not forthcoming until the very end of the conflict and the widespread perception at both the political level and that of public opinion of the ineffectiveness of the entire air operation (Daalder and O’Hanlon, 2001; Everts and Isernia forthcoming). This conclusion is in line with what Mueller (1993) found for the Gulf war is that trend lines did not move within each country. But on the whole, the level of support was less than measured during the Gulf War of 1991, in which bloodshed on the part of the Western forces was equally relatively small (Newport, 1999). Those

15 NATO governments (almost all of them) that refused to countenance the use of ground forces for fear of a backlash from public opinion could find confirmation of their fears in the general gap between support for the air strikes and sending ground forces, particularly when the risks of such an operation were mentioned or suggested. It is understandable, therefore, that they hesitated to deviate from their initial course. Nevertheless support for the NATO actions not only slightly increased to a steady level in the first part of April but, even more puzzling, remained stable for more than two months in front of a bombing campaign whose results remained doubtful up to the very end and of a growing number of civilian casualties in Serbia. Even tough some data point to a slight decline toward May, it is even more remarkable that support for sending ground troops as we have seen it, for instance, in the case of Germany did not erode substantially over time, and in some countries even increased somewhat in the two months of the military campaign. There is reason to suggest that at least some among the public were sympathetic to the compelling logic for sending ground troops in view of the apparent failures of the bombings and the need to provide protection to the Kosovars. Incidentally, the Kosovo war points to the disturbing possibility that a ‘war without bloodshed’ might continue to enjoy public support even without actual or prospective success. In this sense, Ignatieff (2000, p.178) might be right in pointing to the fact that we should reconsider the ‘democratic peace hypothesis’: ‘If wars are cost-free and bloodless, what will restrain the citizens in giving their democratic consent? ...Democracies may well remain peace loving only so long as the risks of war remain real to their citizens. If wars become virtual -and without risk -democratic electorates may be more willing to fight especially if the cause is justified in the language of human rights and even democracy itself.’

Even more puzzling, in spite of the fact that Serbian media were carefully exploiting any opportunity to report on any grisly incident and about ‘collateral damages’ they did not succeed in turning Western public opinion against the war. This stability that we find is more puzzling than the differences in absolute level of support that we have also found across countries. For an explanation of these differences, we should turn to traditional national foreign policy positions, as shaped by the domestic landscape

16 in each and every country. These differences, by the way, were also found in other cases, such as the Gulf war of 1991 and the response to the terrorist attack of September 11th, 2001 (Everts and Isernia, 2002).

1

Chapter 3b Conditions of support of the use of military force

Introduction

In this chapter we will analyse the actual or likely impact on the willingness to use force in the case of the conflict over Kosovo. It is widely acknowledged – and Milosevic might have given some thought to this as well – that the public’s willingness to use force decreases with the fear of casualties (the so-called body bag syndrome, on which see Larson, 1996 and Lorell and Kelley, 1985 and, of course, Mueller, 1979). This alleged unwillingness to face the risk of casualties is not a constant (Everts and Isernia (eds), 2001, Everts 2002). It is influenced, first of all, by the nature of the interests involved and (2) the legitimacy on which the use of force is grounded. Under this heading, we will also include attitudes to the content and origins of the conflict itself and the image of Milosevic, which are of course closely related to how it could end, and what kind of solution should be sought. A third (3) conditioning factor consists of the (perceived) effectiveness or success of the military force employed. Finally, the fourth (4) factor that plays a role is the quality of leadership and/or confidence in the wisdom of government policies on these matters. If the case isn't made, support doesn't materialize nor will there be much willingness to sustain casualties. We hypothesize that the more serious the interests that are at stake, the more legitimate the action is considered, the more effective and successful the action is seen and the stronger and persuasive the leadership the more the use of force will be supported and the less the fear of casualties will reduce support for the war. In examining these factors, of course, we do not intend to claim that what we found at the aggregate level might be valid at the individual level as well (that question will be explored later on). At this point we only want to emphasize the factors that might explain similarities and differences among the aggregate public opinion in different countries.

The nature of the interests involved and the legitimacy of the action

2

The saliency of the issue and the interests at stake Attitudes on the perceived interests involved and the legitimacy of taking action are closely related to questions as to what are (seen as) the origins of the conflict, who was responsible of and to blame for the need to use force and the motives of the parties. Unfortunately, hardly any comparative data on the perceived origins of the conflict and responsibility for it are available. Yet, the outlines of a more general picture can be sketched. As usual, we will start with the American case, on which a richer data set is available, and then add other countries, to qualify or differentiate the conclusions reached in the former case. From the analysis of the US data three pieces of evidence stand out: a relatively (in comparison with other international crises) low level of interest for and attention to the evolution of the Kosovo situation. There was no perceived sense of direct threat to national security and, consequently, no clear understanding of the reasons why NATO was bombing Serbia. Interest for and attention to Kosovo were not very high before the war started and, not surprisingly, although both increased after the war broke out, they never reached the level of interest and attention hit, for instance, during the Gulf war of 1990-91. To a Gallup question asked on the 9-12 of October 1998, 42% said they were following ‘very closely’ (11%) or ‘somewhat closely’ (31%) the situation in Kosovo. The week before the war started, when the negotiators were on their way at Rambouillet (March 19-21), an identical Gallup question found only 30% following closely (9%) or very closely (27%) the evolution of the situation in Kosovo. Understandably, the launching of air strikes on March 24, 1999 focused the American public’s attention for the Kosovo situation.1 On March 24, 71% declared to CBS news to be following closely (45%) or very closely (26%) the news about US and NATO air strikes and by March 28 this percentage was up to 78%. This increased level of attention never matched, however, that reached during the Gulf war, where, on the week after the breakout of hostilities, January 17-20, 1991, 97% said to follow very closely (70%) or closely (27%) the situation in Kuwait (Mueller, 1994: 214). The point of highest attention in the case of Kosovo was reached only in the first fortnight of April, when 84% declared to follow closely the situation in Kosovo, probably as a consequence of the capture (on March 31) of the three American soldiers patrolling the Serbian-Macedonian border. By April, 96 percent of those interviewed have heard ‘about

3 recent events in Kosovo’ (Louis Harris, April 8-13), while the week before the war 70% answered to a Harris poll they had seen or heard about them. The attention brought to bear on the Kosovo situation did not reflect a sense of urgency arising from a threat to American vital interests. The American public was divided on whether ‘America’s vital interests are at stake in the situation involving Kosovo’ (ABC news polls) or whether ‘The United States need to be involved in Kosovo in order to protect its own interests.’ (Gallup polls).2 Between 47% and 43% percent respectively thought that this was the case, with pluralities or majorities thinking the opposite. This, by no means, implies that American deemed the situation unimportant for the US. In fact, up to 83 percent thought that ‘what happens in Serbia and Kosovo’ was important to the US interests (different CBS news polls). But probably, moral or humanitarian considerations played in fact a more compelling role than purely security considerations. On average 62 percent thought that ‘The United States has a moral obligation to try to establish peace in Kosovo’ (Gallup from February to March, ABC News/The Washington Post in April), and this percentage increased between February (52%) and April (71 %). In May 12-16, 1999 69 percent thought that a very important reason for being there was ‘to prevent the killing of citizens in Kosovo’, 57 percent ‘to end starvation and a major refugee problem’ and 52 percent ‘to prevent a larger war’. On the other hand, approximately one third (between 30% and 36%) thought that ‘the United States and its allies bear some of the blame for the Kosovo refugee situation’ (ABC News/Washington Post, April 5-8, 1999). But to a question asking ‘What do you feel are the main reasons for the US involvement in the Kosovo conflict’ (ABC News), in which people was forced to chose only one option, only 28% answered ‘humanitarian purposes’, whereas 39% gave realpolitik reasons such as ‘to prevent conflict from spreading’ (23%), ‘to protect the credibility of the NATO alliance’ (16%) or ‘to protect the credibility of US foreign policy (8%). In any case, the refugee situation made no more than 58% in April 8 more likely to support the dispatch of ground troops (ABC News). The public may have supported the NATO action, purportedly aimed at stopping the persecution of the Albanian Kosovars, but it was – like NATO itself – uncertain about the political aims to be served by the use of force, or, in other words, about the future of Kosovo itself, particularly the status of the area. To one of the few truly comparative questions on war’s aims – ‘do you feel that one of the goals of NATO’s campaign should be to create an independent state of Kosovo, Or should NATO's goal be to simply keep Kosovo a separate governed province

4 of Yugoslavia?’ – asked by Angus Reid on April 22-24 in France, Germany, Italy, United Kingdom and the United States, a plurality (or almost a majority) felt that Kosovo should stay an autonomous province, but about a third agreed that it should become independent (The option to make Kosovo into a UN protectorate which elicited most support when a similar question was posed in the Netherlands3 was not included in this question, however.). A question in Great Britain squarely asked whether the respondent’s thought ‘NATO has clear objectives in Kosovo’: 38% said yes, 37% said no and 25% did not answer.4 [TABLE 10 ABOUT HERE] Table 10 The goals of the NATO operation

The public was much more unanimous, across Europe and in the United States, as to the view that president Milosevic should be forced out of his job and be tried as a war criminal (table 11). In one British poll disagreed, however that Britain and NATO should send assassins to kill Slobodan Milosevic.5 This percentage increased to 38% in late April.6 [TABLE 11 ABOUT HERE] Table 11 Attitudes toward Milosevic

In the US depending on the question one can marshal support for either the idea of making Milosevic to comply with NATO requests or rather removing him from power. In fact, 79 percent thought that the removal of Milosevic from power should be one of the war’s goals (Gallup, 7-9 May, 1999) and 86 percent thought he should be put on trial as war criminal (ABC News/ Washington Post, 5-6 April, 1999). In Europe, a few questions explored the responsibility of Milosevic himself in bringing about the bombing. The results show consistently a majority in favour of Milosevic’ responsibility. In France, at the beginning of the air strikes, a large majority (62 percent) felt that President Milosevic was primarily responsible for the bombing (25 percent felt that the Western countries were responsible).7 In another poll this was even more evident: 77 percent. No wonder then that only 15 percent agreed in that poll that Serbia was ‘a sovereign state which is the object of aggression by NATO’ and 70 percent rather agreed that ‘even if Serbia is sovereign, the strikes by NATO are justified to stop the massacre in Kosovo’.8 In Germany 68 percent of all respondents (72 percent in the West and 54 percent in the Eastern part of the country) thought

5 that Serbia carried the main responsibility for the conflict. 5 percent thought so for the UCK and 7 percent put the blame on NATO (14 percent in the East), still another 5 percent blamed ‘all equally’.9 Among the British 69 percent believed the allegations of ethnic cleansing and felt that NATO had the right to intervene. Only 23 percent thought that the Kosovo crisis was an internal matter for Yugoslavia.10 This sentiment was echoed in another poll in which 69 percent disagreed (21 percent agreed) with the statement ‘the problems of Kosovo are none of our business and Britain should not interfere’.11 More in line with the American polls is one survey from Italy in which 62 percent expressed agreement with the argument that the decision to bomb Serbia was necessary to bring the repression in Kosovo to an end.12 In the same Italian poll the humanitarian argument ‘to prevent further massacres in Kosovo’ was also the most frequently chosen among a list of possible reasons to bomb Serbia. This is not to deny that this view was not also mixed with apprehension about possible ulterior motives of the United States, i.e. to reaffirm their own superiority in the world (54 percent agreement) or even to divert attention from the internal problems of President Clinton (25 percent agreement). It is interesting to compare the Italians’ perceptions with those of the Americans themselves. Here, on March 25, 1999 (Rasmussen Research) 35 percent thought that the President Clinton ‘authorized military action at this time to distract public attention from the stories concerning the Chinese theft of nuclear secrets or other troubling issues’ and in April 7-8 (Opinion Dynamics/Fox News) 49 percent thought Clinton was diverging attention from his domestic scandals (namely the Lewinsky affair and the Chinese campaign contribution and spying). In view of all of this there is reason to doubt the strength of the impact of the factor of perceived interest and the clarity of purposes on the willingness to fight.

The source of legitimacy Another factor usually considered important in military missions out of defence of national security is the source legitimating the action. Is the military tool used on behalf of NATO, the United Nations or rather the European Union (for the European countries, of course)? Or is it rather a unilateral initiative of the country undertaking the action? In the Kosovo case, contrary to other previous operations, a formal approval by the United Nations was lacking. Not many questions are available to assess the role of legitimacy in propelling or reducing support for the

6 use of force in Kosovo. PIPA on May 13-17 squarely addressed this issue, asking ‘As you may know, the UN Security Council has not made a specific resolution to authorize NATO’s use of military force in Kosovo. I am going to read you three statements. Please, tell me which is closest to your own view.’ The first statement – ‘It does not concern me that the UN has not approved NATO military action’ – got the support of 19 percent; the second statement – ‘It concerns me that the UN has not approved NATO military action, but I think NATO’s operation should continue’ – was approved by 48 percent and the third – ‘It concerns me that the UN has not approved NATO military action, and I think NATO should stop its operation’ – was approved by 30 percent. In Denmark, to a similar questions asking if the respondent was in agreement with the following statement: ‘NATO should be able to decide to bomb even without approval of the United Nations’, 55 percent disagreed, and 25 percent agreed, with 9 percent uncertain and 11 did not know (Politiken, March 31, 1999, N=?). The Dutch were divided almost equally on the question whether is was ‘problematic or not that there has been no official permission of the United Nations for the NATO action’ (with only ten percent saying that this was ‘very problematic’ and 16 percent that this was ‘not problematic at all’ (NIPO/SMK, 3-5 April 1999, N=980).

The effectiveness of the action

One obvious shortcoming of these and other related poll questions is, of course, that they refer to hypothetical cases or abstract situations. They should therefore be complemented by what we know about (perceived) effectiveness and its impact on support of the action. To this we turn now. In spite of the fact that by mid-April, when the survey, whose results we now present, was made, Serbia had not yet met the demands of ‘the international community’ (represented by the countries co-operating in the framework of the G8) and the end of the war was not in sight (indeed, worse was yet to come), the public in the various European countries was, more or less, fairly satisfied with the results of the bombing campaign until then, as is shown in Table 12. [TABLE 12 ABOUT HERE] Table 12 Has NATO done a good job? (in %)

7 Other data contradict this, however. The Hungarians, while strongly supportive of the air strikes in general, were, being closest to the conflict, most sceptical. A plurality felt in April that the action had been a failure, and on April 8, 84 percent (78 percent one week later) felt that the region was now worse off than before.13 The French were only slightly less sceptical. While two weeks after the beginning (9/10 April) only 29 percent in France felt that the action would fail, this had increased to 52 percent by 19 April and, again on 21/22 April 1999.14 In another poll, it was found that already by early April 59 percent felt that the effect of the NATO action had been ‘rather negative’.15 And in still another poll, on 11 April, 64 percent felt that the military intervention had been a failure rather than a success (21 percent). In Germany similar doubts were expressed. 61 percent thought by the end of February 1999 that the chances of success for an international force to restore peace in Kosovo were ‘rather bad’.16 Data from Germany also allow us to trace the evolution over time of the belief that NATO's actions would force Milosevic to yield. It is remarkable that throughout the conflict Germans on balance thought that the NATO intervention would not force Milosevic into compliance. Yet, probably at least partly out of alliance loyalty, they still continued to support the actions and German participation in the air strikes (see Figure 3).17 [FIGURE 3 ABOUT HERE] Figure 3 Support for the air strikes, German participation in sending ground forces and expected success in Germany

Other data also suggest that one could well be in favour of the air strikes and feel too that they should continue, and yet not believe that they would be effective. Thus, to give one example from France, in one poll, held 11 April 64 percent felt that the military action had been a failure. Yet, in the same survey 56 percent felt that it was necessary to continue the action until the Serbs gave in to NATO’s demands. 42 percent thought, however, that it would be better to stop bombing and start immediate negotiations. By 15 May, support for continuing bombing had decreased to 49 percent and support for negotiations increased to the same level.18 As noted already, in Germany a similar effect occurred. While in one poll (29-30 March 1999) 50 percent felt that the war NATO was waging against Serbia was justified, 64 percent in the same poll said that NATO should ‘now’ introduce a pause in the bombing in order to achieve a negotiated outcome.19

8 One other interesting assessment, showing the existence of considerably mixed feelings, was made in the Netherlands (June 1999). While 22 percent (+57 percent ‘strongly’) agreed that ‘without the bombing Milosevic would never have agreed to the diplomatic settlement’, 40 percent (+12 percent ‘strongly’) also agreed that ‘the diplomatic efforts by Russia have been equally important as the bombing in reaching agreement’. Moreover, 40 percent (+20 percent ‘strongly’) also felt that ‘the people of Yugoslavia have been hurt too strongly in their daily lives’.20 After the cessation of the bombing and the agreement of Serbia to the deployment of KFOR there was much more optimism. In Germany 65 percent agreed that ‘a political solution of the Kosovo conflict’ would now be possible.21 A set of questions asked by different survey institutes in the United States offer a clue to what might be going on in the public’s mind. The American public felt that the conflict was going well ‘for the United States and NATO’ (from 50 to 58 percent saying so), and that the US military involvement in the war ‘stand a good change of bringing peace’ (from 50 to 56 percent), but that the NATO intervention had made the situation worse for the region (64% in March, 49% in May). However, in the future, once the war ended, the situation would be better (53% thinking so in March). The expectations on how long the war is going to last are a crucial factor affecting the perceived effectiveness of the action. Only scarce data are available on this point, but they seem to point to an expectation longer than that it actually was. CBS News asked three times ‘Do you think the current conflict in Kosovo will be over in just a few weeks, or do you think it is more likely to continue for several months, or do you think it’s likely to continue for a year or longer?’. Between early April and early May a slight majority (between 49% and 51%) thought it would have lasted for months and more than a third expected it to last for a year or longer. Only 8% (and they were 13% on April 5-6) thought on May 1-2 that the war would have lasted, as it actually did, i.e. a few weeks. A similar question was asked by MORI (for Mail on Sunday) in late March and late April. At the beginning on the conflict (March 26-27), 27% expected the war to last no more than a month, 11% from one to three months, 33% more than three months (with 26% more than 6 months) and 29% did not know. On April 30 (MORI for Mail on Sunday), the percentages were respectively 6%, 15% and 48% (with 33% more than 3 months), and 30 did not answer.

9 Looking at the data, messy as they are, one gets a picture of a public that was ambivalent in front of the developments in the Kosovo crisis. A real dilemma apparently faced the population in those countries surveyed. On the one hand, expectation of success were low because air strikes were not seen as effective and ground intervention was seen as needed to fix the problem. On the other hand, there was the perception that such an operation would have implied a much greater effort, both in terms of human and financial resources; an effort the public was generally not ready to sustain. Finally, in a flip of mind typically reminiscent of cognitive dissonance, the gloomy perception of the situation right now left room to a much better one in the future. Let us look at these three aspects in turn. There is enough evidence in several countries to show that a majority of the public thought that the air strikes were ineffective or, worse, ‘more likely to strengthen his [Milosevic’] resolve to continue fight’. On March 25, 1999 36% thought that the air strikes now underway ‘are likely to make Milosevic to sign a peace agreement’, whereas 45% thought that the air strikes were ‘more likely to strengthen his [Milosevic’] resolve to continue fighting’ (Los Angeles Times, N=544), and 19 did not know. To a similar question asked in April, 53% thought that the air strikes made ‘less likely’ to sign the peace agreement.22 In part, it was the war itself to determine such a change. This is nicely captured by a NIPO survey in the Netherlands on April 6, 1999, in which it was asked ‘Since two weeks NATO carries out air attacks against Yugoslav territory. When you think back of two weeks ago, did you then agree or disagree with bombing of Yugoslavia?’. 64% agreed with the bombing and 25% disagreed, with 11% uncertain. Then the following question was asked: ‘How do you feel about the air strikes against Yugoslavia today. Do you agree or disagree with bombing of Yugoslavia?’. This time 57% agreed and 34% disagreed, with 9% uncertain.23 We can only speculate on the reasons why support for the war never increased and sometimes, as in the Netherlands, decreased. In the US, a majority thought that ground troops would be required,24 and this worried them very much, since they were thinking that the war was not going well at all at that time.25 Table 13 too shows how it was quite clear, at least to the American public, that the air strikes, no matter how long they might last, were not enough to bring Milosevic to a peace agreement. Apart from the fact that a reference to Milosevic apparently increased the public uncertainty on the role of both air strikes and ground troops, a stable majority thought since the

10 very beginning that air strikes were not enough to bring Milosevic’ Serbia to the negotiating table. [TABLE 13 ABOUT HERE] Table 13 Air strikes are enough? (in %)

In a similar vein, asked if they believe or not ‘that NATO will eventually have to use ground troops’, 70% of the polled British said yes to a Gallup survey on March 24-28, 1999 for the Daily Telegraph. More divided the Dutch, with only 9% believing that ‘the war in Yugoslavia can be won without sending ground forces’, while 43% think it cannot be won without a ground operation and 48% do not know.26 Again, to a slightly different questions asking whether ‘a ground war is necessary in Kosovo’ to stop the war, 92% of the Turkish said yes and only 8% no.27 Finally, as we showed in the first section, support for such a ground operation was lacking and unstable, due to a wide set of considerations. It is remarkable that, caught in a real dilemma between what is acceptable – air strikes – and what is useful – ground force, public opinion nevertheless remained stable for the entire duration of the war. In fact, there is no sign of fatigue or decline in support. Apart from the national differences in the absolute levels of support due to domestic factors, support (or opposition) remained quite stable for more than three months. This experience shows, therefore, how prudent and calm public opinion can be when there are uncertain choices around unpalatable alternatives. In part, this can be explained by a certain optimism among the public on the long-term evolution of the situation. Table 14 shows an interesting set of results of a couple of questions asked in three different periods to samples of the American population. These data show that the public was making a distinction between ‘now’ and ‘the future’ in assessing both the impact of the war in the region and its success. The assessment is very sober when one is brought to look at the present situation rather than at the future one. And this is a reflection of genuine uncertainty, as evidenced by a slightly different question, asked by ABC News on April 8, in which, asked to tell whether ‘the military action (in Kosovo) so far has been a success or a failure for the United States and its allies, or is it too early to tell?’, 68% answered that it was too soon to tell and only 16% and 14% thought respectively it was a success or a failure. [TABLE 14 ABOUT HERE]

11 Table 14 NATO actions as a success or a failure depending on the time frame (in %)

The quality and persuasiveness of political leadership

As will be shown below, there were remarkable differences between Europe and the United States in the degree to which their respective leaders were trusted and people were satisfied with their decisions. It makes sense, therefore, to analyse the available data separately and after another.

Data from Europe To the extent that poll data on this topic are available there is ample evidence that people in the European countries concerned were generally both confident in their leaders wisdom and satisfied with their decisions concerning Kosovo. In Great Britain, for instance, two thirds of those interviewed were satisfied with the way Prime Minister Tony Blair was handling the conflict over Kosovo.28 Likewise, majorities agreed with the government’s policies on Kosovo and continued to do so throughout the crisis.29 In Italy, majorities also supported the government’s policy on Kosovo, although with some more and increasing hesitation.30 In countries like Croatia31 and Greece,32 majorities were in agreement with their government’s policies too, Nevertheless, at the same time, majorities also felt in Great Britain that the government had not provided enough information prior to the bombing actions to explain what the reasons and possible consequences of NATO’s actions in Kosovo were.33 We should also remember that many in Europe were profoundly pessimistic that the bombing actions would be effective in reaching NATO’s goals, in stopping the flow of refugees and restoring human rights.

The situation in the United States

Compared to Europe, there was (much) more scepticism and lack of trust in the government’s policies and President Clinton’s leadership throughout the conflict over Kosovo in the United States. True, when asked whether they approved or disapproved of the way the president’ was handling the situation in Yugoslavia, the level of approval did not differ that much from Europe

12 with percentages of approval varying between 44 and 60 percent throughout the crisis and with no clear pattern of in- or de-crease over time. Asked in a more precise fashion to rate his handling from ‘excellent’ to ‘poor’ did not produce a majority rating him either ‘excellent’ or ‘good’ in any of the Rasmussen polls held from 19 March to 12 June 1999.34 At the same time only pluralities or very slight majorities agreed in the period from March to May that the President has explained the situation and the reasons for the NATO actions well enough.35 Again, only pluralities felt in various polls that he had done ‘a good or excellent job’ in doing so.36 Gradually, the number of those who were not satisfied with his explanation of why the air strikes were being carried out increased (from 14 March to 5-6 May), from 25 to 47 percent.37 Even more discomforting for the administration was the outcome of another series of polls which showed that (slight and diminishing) majorities thought that the Clinton administration had a ‘clear and well-thought-out policy on the Kosovo situation.38 On 5-6 April 58 percent thought that President Clinton was ‘just reacting to events as they occur’ and only 28 percent were convinced that he had ‘a clear policy and long-range plan’.39 On 17-19 April these figures were 71 and 22 percent respectively.40 In another series, percentages increasing from 43 to 56 percent felt that in the area of handling the war in Kosovo the President was ‘in over his head’ with only 42 to 35 percent stating, on the other hand that he was ‘doing a good job’ It was only after the war was over that the majority became positive for him.41 Yet , in another poll 29 percent trusted him ‘a lot’ and 28 percent ‘ a good amount’ ‘to make the right decisions about the military confrontation in Serbia’,42 By and large, with respect to government leadership the American public was strongly divided and only by a small majority leaned towards the position that the President and his administration were doing a good job and could be trusted on Kosovo and had a clear cut plan in their minds. We pointed out already that the administration had ruled out the use of ground forces from the beginning rather than leaving this possibility open to add to the Serbian uncertainty. Under these circumstances it is understandable that the American public too showed considerable hesitations with respect to both the purposes of the American and NATO involvement and the need to run the risk of military and civilian and military casualties in this connection. The lack of convincing leadership did little to reduce the impact of this fear on the willingness to fight.

13

Support for the casualty hypothesis?

The last, and most often mentioned, factor influencing, negatively, the degree of support of military action is the fear, or actual incidence, of casualties (the so-called ‘casualty hypothesis’. In the case of Kosovo, given the absence of truly vital national interests, the relatively low saliency of the issue and the perceived lack of success (at least in the short run) discussed above, one might expect a comparatively strong impact of this factor in the case of Kosovo. The previously observed gap between support for NATO’s mission in general, almost in the abstract sense, and that for sending ground troops of one’s own country suggests indeed a considerable impact of this ‘body bag syndrome’. Supporters of the casualty hypothesis can interpret this gap as confirmation of what they had thought all along. Its existence also suggests that the governments concerned were right in their caution about speaking openly about and preparing for the (apparently much more risky) use of ground troops, assuming that this could present them with costly electoral liabilities. Yet, there is ample reason for caution in interpreting the data. Let us first look at the available data in some more detail and also inspect the evolution of the willingness to sustain casualties in function of other aspects of the war. For the sake of clarity, we will discuss first the data on the European countries and then those for the United States.

Attitudes on casualties in Europe In spite of all the public debate on this topic, not many relevant questions have been asked in Europe, and those that have do not provide much possibility of cross-national comparisons, nor can the outcomes be generalized with some confidence. One interesting outcome can be noted for the Netherlands. In October 1998 respondents were asked about a possible intervention in Kosovo. This found not only widespread support, but also this was not reduced if the prospect of casualties were raised.43 Arguably, such an outcome should be taken with some scepticism since we were dealing with a hypothetical situation. However, when the military actions had begun, the question was repeated in early April, and then the same outcome was obtained. 68 percent supported the NATO actions and the same number felt that the Netherlands should continue to participate in these actions, even if casualties were to be incurred

14 among the NATO forces. Only 14 percent felt it should not, and 18 percent hesitated or would not reply. This figure was also obtained after the termination of the military actions in June.44 This was in stark contrast to the situation in the UK, where the level of general support was equally high, but many also feared the prospect of casualties. True, according to one poll in this country at the beginning of the actions, a majority was prepared to send ground troops as well as accept casualties (51 percent).45 Another survey at the same time, however, suggested quite the opposite when 56 percent (and 57 percent one week later) agreed that the protection of the Albanian Kosovars was ‘not worth a single British life’, and only 19 percent agreed with casualties in the order of 1-1000. Many, still, could or would not answer such a question (34 percent at the end of March and 24 in early April).46 Unfortunately, we do not have later figures for the UK. In this respect Italy took a middle position. Toward the end of May the following question was put: ‘Would you favour or oppose a hypothetical NATO intervention by ground forces that would succeed in ending the repression and allow the return of the Kosovars if this would entail the death of ten Italian soldiers?’ While a slim majority of 51 percent said ‘no’, still 41 percent answered ‘yes’.47 Figure 2 illustrates the limited impact of the fear of casualties in the Italian case. It not only shows the relatively low but stable level of support for sending ground forces in general in Italy, but also the relatively small impact of the fear of what this could imply for the risks involved for one’s own soldiers. Indeed, the level of support for ground intervention was hardly affected by either the prospect of participation of Italian troops or the fear of casualties due to such participation. [FIGURE 2 ABOUT HERE] Fig. 2 - Support in Italy for sending ground forces, Italian participation in such an operation and the incidence of casualties

Attitudes on casualties in the United States More figures are available for the US. Most of them by themselves suggest both concern and considerable reluctance to incur risks. Clearly, Americans were worried about possible casualties in the war, and these worries increased as the war was prolonged and an immediate end was not in sight (Table 15). [TABLE 15 ABOUT HERE] Table 15 Worries about casualties in USA (in %)

15

In another poll, expectations of losses went up in the following way: 68 (24 March), 78 (30 March), 84 (1 April) and 85 percent 4 (5/6 April).48 Faced with the following question: ‘In general, do you think the United States military should be able to win a war against a country like Yugoslavia without any American soldiers being killed or wounded, or do you expect that to win a war some American soldiers will be killed or wounded?’ 16 percent expected no casualties and 81 percent expected that soldiers would get killed. Asked, ‘As of now, in the conflict in Kosovo, how many American soldiers would you expect to lose their lives—a lot, some, or hardly any? 13 percent replied ‘a lot’, 45 percent said ‘some’ and ‘hardly any’ was referred by 35 percent.49 Clearly too, although various polls gave varying outcomes, many Americans, at least before and at the beginning of the conflict doubted whether it would be worth suffering casualties. To begin with two polls that suggested that majorities would accept casualties just as the war started: 53 percent disagreed with the statement: ‘It's not worth risking American lives to bring peace in Kosovo’.50 A clear majority was also found when the question was formulated in the following, appealing way: ‘Do you think it's worth risking the lives of American soldiers in order to demonstrate that Serbia should not get away with killing and forcing people from their homes, or not?’51 Fifty-four percent felt it would be worth it and 38 percent said ‘no, not worth it’. A similar effect occurred when reference was made to ‘a limited military action’: ‘[is] … this goal (withdrawal of Yugoslav forces) worth having a few American casualties in a limited military action?’. ‘Yes’ said 50 percent and ‘No’ 42 percent.52 When prodded on how many casualties would be acceptable, the degree of acceptance dropped off rapidly, however. Only 14 percent said ‘many casualties’, ‘only a few’ said 36, and 42 stuck to the original ‘not worth any casualties’. In the same survey about equal numbers were and were not prepared to allow a son of military age to get into a combat situation (46 percent: ‘yes’, 49 percent ‘no’). Another poll suggests that the wording of the question may not be so relevant here after all. It showed an opposite outcome, in spite of the reference to ‘if it could bring peace’. Only 7 percent now said ‘many’, 26 said ‘a few’ and 56 percent said ‘none’.53 It is not clear whether fears increased or diminished over time. The information is contradictory, as shown in table 16, although it indicates how tolerance increased from just before to just after the beginning of the air strikes in March 1999. But, whatever the case, this table also shows rather clearly that by and large there was a rather critical opinion climate with

16 respect to the acceptability of casualties. One observer, looking at the available evidence noted at the beginning of April: ‘There is very little appetite for casualties […] Support that exists now would evaporate if the spectre of a quagmire were to be evoked by Kosovo’.54 But others said that things might also turn out differently. One is tempted to agree with the comments of two other observers, who added, looking at the data: ‘The numbers are suggestive but not predictive. Public response to actual casualties in actual ground war remains to be seen’.55 [TABLE 16 ABOUT HERE] Table 16 Would it be worth to suffer casualties? (in %)

How the fear of casualties (which by itself diminished somewhat over time), and the number of them influences support for sending ground troops (at least in opinion surveys) is illustrated nicely in table 17. [TABLE 17 ABOUT HERE] Table 17 Casualties and support for sending ground troops (in %)

The obvious dilemmas and possible trade-offs were explored in somewhat more detail in a set of other questions, dealing with the situation that arose when the bombing failed to realize the announced objectives. Respondents were asked: ‘Given what you know right now, which would you prefer the United States and NATO to do in the conflict in Kosovo: (1) send ground troops OR (2) continue the air strikes but don’t send ground troops, OR (3) withdraw its forces before becoming more deeply involved?’ If the answer was ‘send ground troops’ another question followed: ‘What if sending ground troops meant there would be a lot of American or NATO casualties? Then, what would you prefer 1) send in ground troops, OR 2) continue the air strikes but don’t send ground troops, or 3) withdraw forces before becoming more deeply involved? Taken together, the outcomes shown in table 18 were obtained.56 [TABLE 18 ABOUT HERE] Table 18 Support for sending ground troops (in %)

This was followed by two questions, which read: ‘…. Is it better to restrict the bombing/restrict ground troops even if it means a longer war or is it better to increase the

17 bombing/increase ground troops even if it means risking American and NATO casualties in order to bring a quick end to the war?’ The answers are shown in table 19. [TABLE 19 ABOUT HERE] Table 19 Support for sending ground troops under different conditions (in %)

Restricting the war was the most preferred option by both those who preferred bombing and ground troops, but more of the same for a shorter period was by far a preferred option by those in support of air strikes than by those willing to risk the deployment of ground troops. It is also possible to be more specific, and consequently much less pessimistic about the willingness to run risks. This is suggested by the outcomes of another, much more detailed survey on the war in Kosovo, conducted in May 1999, which, incidentally, also illustrates the usefulness of using a variety of questions including some that are deliberately one-sided.57 First, some introductory questions were asked, including one in which an affirmative majority reaction appeared to the statement: ‘The longer the NATO operations in Yugoslavia continue, the more likely it is that American soldiers will be killed. We should get out now before Americans come home in body bags’. This is what supporters of the ‘body bag hypothesis’ would expect to find, and 52 percent indeed found this argument convincing (42 percent did not). Asked directly in this survey, 59 percent opposed sending ground troops ‘at this point'. The survey quoted above, however, also made a deliberate effort to connect feelings about casualties to possible positive outcomes of the conflict and – which is rather exceptional in polling – to present respondents with alternative options in deciding how to react if casualties were to actually occur (which is what the ‘body bag hypothesis’ in the strict sense is about). One fourth of the sample each was confronted with one of two scenarios in which respectively (a) ten or (b) fifty Americans were killed. Respondents could react by choosing one of four reactions, including not only the option of withdrawal, but also that of doing the reverse, hitting back with all disposable means. The data in Table 20 show that only a minority of some 20 percent manifested the gut reaction of the body bag syndrome and that equal numbers would want to do the very opposite. It is remarkable that the number of casualties by itself does not seem to play a role here. [TABLE 20 ABOUT HERE] Table 20 Reactions to casualties in Kosovo (in %)

18 Each of the two other quarter samples was confronted with one of two scenarios including the occurrence of casualties in a ground war but with a positive outcome of the actions. They were asked to state whether they would consider such actions to be right. The results are given in Table 21. It seems evident that casualties, and the number of them, do matter but their impact is strongly mitigated by the purposefulness of the action. [TABLE 21 ABOUT HERE] Table 21 Reactions to casualties and success in Kosovo (in %)

In the PIPA poll quoted above some scenarios including a reference to the effectiveness of the military action were also submitted to the respondents. The first sketched a situation in which a few American pilots were killed due to flying at low altitudes, but in which the bombing would also result in largely ending the ethnic cleansing: 51 percent (against 44) would accept this and continue the action. The other half of the sample was confronted with the situation that five pilots would die, but president Milosevic would be persuaded to stop the ethnic cleansing and withdraw his troops: in that case 67 percent would think it right that NATO had taken such a dangerous action. Yet others heard the scenario, in which 50 pilots were killed but Milosevic was persuaded to stop the ethnic cleansing and accept the demands of NATO. This decreased support somewhat, but 52 percent would still consider this form of risky bombing to be a right decision. Although, as mentioned above, 59 percent rejected an intervention with ground troops, a similar number also accepted the argument: ‘Whether or not NATO ultimately intervenes with ground troops, it needs to have that option in case bombing fails to make Milosevic stop the ethnic cleansing. A ground invasion is the only direct means of resting control of Kosovo from his forces.’ After other arguments against sending ground troops had been mentioned, respondents were now asked again whether they were on balance for or against sending ground troops. 53 percent now agreed and 42 percent was still against. 48 percent was prepared to begin to do so ‘now’, a considerable increase over the original ‘raw’ figures. This seems to show, as was suggested already above, that the impact of the fear of casualties on action support is indeed mitigated by the real or perceived effectiveness of the action.

Some general conclusions

19 The available survey data concerning the war over Kosovo of 1999 illustrate anew a number of findings from earlier conflicts. First of all, many people see the problems concerned as being, in principle, ‘far away’. This means that it takes time before attitudes are crystallised and settled, but when they are, they evolve according to understandable and recognizable patterns. In this process various factors are at play: some initial willingness to ‘rally ’round the flag’, whatever the government decides (a weak impact in this case), habituation to the use of force, crisis, fatigue, sense of revenge, sensitivity to, and fear of, the risks involved, in combination with positive or negative expectations about the outcome and success of the action.58 The available data also show, once more, the effect of variations in the way questions are phrased. If one restricts oneself to one or two questions it seems true that ‘with opinion polls one can prove anything’, but if differently phrased questions are juxtaposed a much more detailed, differentiated and clearer picture emerges.59 As far as the content of attitudes is concerned, it is remarkable how support for the NATO actions grew after the initial hesitations, and how NATO was able to sustain this support in spite of the widespread ‘collateral damage’ and the fact that results were not forthcoming until the very end of the conflict. But on the whole, the level of support was less than measured during the Gulf War of 1991, in which bloodshed on the part of the Western forces was equally relatively small.60 Those NATO governments (almost all of them) that refused to countenance the use of ground forces for fear of a backlash from public opinion could find confirmation of their fears in the general gap between support for the air strikes and sending ground forces, particularly when the risks of such an operation were mentioned or suggested. It is understandable, therefore, that they hesitated to deviate from their initial course. Ignatieff questions what, if there was public consent to go to war at the beginning, actually meant. The public, he argues, certainly did not sign up to 78 days of bombing, expecting like their leaders, a short conflict. The decline of support of the bombing suggests that with hindsight many in the public would have said no, and he concludes. ‘The President secured the public’s consent by withholding the fact, attested by military commanders, that ethnic cleansing could only be deterred or stopped by ground troops ... It is not clear whether consent would have been given had all the facts been out before the public in a democratic fashion’.61 Some available data show clearly, however, that once confronted with casualties the public’s reaction would not be automatically to ask for withdrawal and run. In fact, the opposite might have

20 occurred. The fact that general support initially grew, and stayed at the same level in spite of strong scepticism as to the likely effects and outcomes of the actions combined with widespread disbelief in success at the same time also indicated the robustness of that support. The fact that support of the air campaign over Kosovo remained at some 50 per cent in spite of misjudgements of Serbian resolve is indeed remarkable. It is true that a majority in the United States opposed sending ground troops but that is not strange given the fact that Clinton had ruled this out himself. ‘More telling is that four out of ten Americans nonetheless supported sending troops.’ 62 The data also suggest a potential for growth in support as well as potential support for more risky operations provided that there would have been convincing leadership in the countries concerned and in the NATO alliance as a whole. Kohut may have been right, however, when he argued that given the circumstances and the existing hesitations, a (much) larger degree of support for sending ground troops would have been required in political terms than could have been mustered even by decisive leadership.63 Although, as was shown above, the data at first sight show considerable support for the casualty hypothesis in its simple form, there is additional reason for caution in this respect. First of all, we should remind ourselves that, as so often, we were dealing with a largely hypothetical possibility, which in actual fact had been ruled out deliberately in advance by many governments concerned. How people would actually react if and when a concrete decision to intervene with ground forces would be taken, or after such a decision, remained to be seen. Precisely because in humanitarian cases, compared to those involving direct security threats or the wish for revenge, several and contradictory emotions play a role, it is difficult to anticipate how support would develop if sacrifices were actually at stake and called for. Moreover, public opinion is not really an independent variable that could not change under the impact of the same events and developments to which political leaders are also exposed. Nor is it insensitive to the policies of or, more precisely, the leadership displayed by their respective governments. Both support and the willingness to run the risk of casualties could be significantly influenced by such policies, and support could hence increase or decrease accordingly. The message of an absence of public support, as reported in the media, can - to some extent and within obvious limits - function as a self-defeating prophecy.

21 The public is certainly prepared to follow its leaders, at least initially. Thus, while support for the NATO actions tended to be somewhat hesitant in the beginning, it increased gradually to a steady level in the beginning of April. It was only when the success of the bombing campaign remained doubtful and the growing number of civilian casualties in Serbia became evident that doubts began to grow regarding the wisdom of the whole campaign. Thus, this was not a time to expect growing support for sending ground forces. In this light it is even more remarkable that support for sending ground troops as we have seen it, for instance, in the case of Germany did nót erode substantially over time, and in some countries even increased somewhat in the two months of the military campaign. There is reason to suggest that at least some among the public were sympathetic to the compelling logic for sending ground troops in view of the apparent failures of the bombings and the need to provide protection to the Kosovars.64 The data also show that there were not only similarities among NATO countries like the initial hesitations and the growing support in the early phases of the conflict – with respect to which international news media undoubtedly played a homogenizing role. There were also substantial differences. There was a group of countries where support was clearly more pronounced than elsewhere. This included the major protagonists, like the United States and the United Kingdom, but also countries like Denmark, the Netherlands and France. Support in Germany lagged behind, but within Germany there was hardly any evidence of opposition along, polarized, party-political divisions.65 There was relatively much more scepticism in the Eastern compared to the Western Bundeslaender. Scepticism was also much in evidence in Italy as well as in the more ‘peripheral’ countries Portugal and Spain, not to mention the three new NATO member-states. It was evident that this was exactly where the faultlines in the initially rather coherent NATO front would appear if the actions would continue without visible positive effects. Such cleavages indeed already began to appear by mid-May, and the fear of them certainly had a constraining effect on the freedom of action of the alliance, although one would underestimate the effects of the perceived constraints of public opinion in the United States to argue that Clinton’s hands were tied more by public opinion in European countries than in his own country. Apart from the obvious reason that public opinion was willing to follow their leaders in supporting or even joining in NATO’s air strikes because of a basic feeling of solidarity with other countries in the Atlantic alliance, the most important parameters of the evolution of support for the military actions in their various forms seem to be 1) the perceived legitimacy of the actions and the

22 interests involved, 2) the perceived effects of the actions, i.e. perceptions of success or failure, 3) the sustainability of the idea of a ‘clean war’, a war without bloodshed, first of all on one’s own side, but perhaps also with respect to civilian casualties on the other side. The public wanted to see a war that would be the equivalent of what President Clinton in another context, called ‘smoking pot without inhaling’.66 A fourth factor, finally, could be called ‘fatigue’. We often assume that support diminishes with the duration of a conflict, especially when the costs rise and success is not forthcoming. As the Kosovo war progressed, some signs of gradual (but limited) erosion of support became indeed visible from mid-April onwards. Although public attitudes certainly contained emotional elements, on the whole the public behaved rationally and in an understandable fashion. It was not the public that constrained their leaders. Mistaken beliefs about public opinion led to a war strategy that almost failed. NATO narrowly escaped. It won, but it won ‘ugly’.67

23

References Brody, R.A. and Shapiro C.R. (1989): ‘Policy Failure and Public Support: The Iran-Contra Affair and Public Assessment of President Reagan’, Political Behavior, 11, 353-369.

Daalder, I.H. and O’Hanlon, M.E. (2000), Winning Ugly: NATO’s War to Save Kosovo, Washington, Brookings Institution.

Everts, Ph.P. (2002), Democracy and Military Force, Basingstoke, Palgrave.

Everts, P. and Isernia, P. (2002), ‘Onlookers or Participants? Public Opinion on the Problem of Terrorism since September 2001 in Countries Outside the US’, paper presented at the Joint Conference of the American Association of Public Opinion Research and World Association of Public Opinion Research, May 14-19, 2002, St. Pete Beach. Florida.

Ignatieff, M. (2000), Virtual War. Kosovo and Beyond, New York, Henry Holt/Metropolitan Books.

Isernia, P. (2000), ‘Where Angels Fear to Tread: Italian Public Opinion and Foreign Policy’. In B.L. Nacos, R.Y. Shapiro and P. Isernia (eds.): Decisionmaking in a Glass House. Mass Media, Public Opinion and American and European Foreign Policy in the 21st Century, Lanham, Rowman & Littlefield, pp.279-296

Kohut, A. (1999): ‘Beware of polls on the war’, New York Times, April 8.

Larson, E.V. (1996): Casualties and Consensus. The Historical Role of Casualties in Domestic Support for U.S. Military Operations, Santa Monica CA: Rand Report.

Lindsay, J.M. (2000), ‘Looking for Leadership. Domestic Politics and Foreign Policy’, The Brookings Review, 18, 1, 40-43.

24 Mueller, J. (1973): War, Presidents and Public Opinion, New York, Wiley.

---------------- (1993): ‘Review: Public Opinion and the Gulf War’ Public Opinion Quarterly. 57, 1, 80-91.

---------------- (1994): Policy and Opinion in the Gulf War. Chicago, University of Chicago Press.

Newport, F. (1999): ‘Public support for US Involvement in Yugoslavia lower than for Gulf war, other foreign engagements’ Gallup Poll releases. March 30 (www.gallup.com/poll/releases/pr990330.asp).

Rusciano, F.L. et al. (1998): World Opinion and the Emerging International Order, Westport, Praeger.

Verba, S. (1971): ‘Cross-National Survey Research: The Problem of Credibility’. In I. Vallier (ed.): Comparative Methods in Sociology, Berkeley, University of California Press, pp.309-356.

Weymouth, A. and Henig, S. (2001): The Kosovo Crisis. The Last American War in Europe?, Harlow, Pearson Education.

25 TABLE 1 - International public opinion and the conflict in Kosovo. Views on NATO actions (in %) 1999, 25 March – 17 Aprila Support Oppose DK/NA Belgium Canada Denmark France Germany Greece Italy Norway Portugal Spain United Kingdom United States

68 68

23 27

9 5

Czech Republic Hungary Poland

35 48 54

57 41 31

8 11 15

64 74 54 57

33 19 34 36

3 7 12 7

47 64

47 23

6 13

1999, Mayb Support Oppose DK/NA Original NATO member 53 35 12 70 68 52 2 37

20 27 40 97 46

10 5 8 1 17

34

48

18

1999, 2 Junec Support Oppose

DK/NA

Averaged

Differencee

47

-9

72 61 54 2 45 64 41 36 67 68

-4 +8 -3 0 +4 na na +5 -1 na

35 48 54

na na na

42

5

13

62 54 2 51

32 44 97 40

6 2 1 9

41 39 67

51 49 41

12 12 6

New NATO members

Other European countries Austria 41 43 16 41 Croatia 82 7 11 82 Finland 50 35 15 44 43 13 47 Ireland 46 42 12 46 Luxemburg 61 30 9 61 Russia 2 94 4 2 Slovakia 21 75 4 21 Ukraine 4 89 7 4 a - As you may know, NATO has recently taken military action in Kosovo. Do you support or oppose NATO's decision to carry out air and missile attacks against Serbian military installations? (AngusReid group/ The Economist, N= about 500 per country) b - Do you personally support or oppose the military actions by NATO in Serbia? (ICM/European Barometer/The Guardian, N= not available) c - Do you personally approve or oppose the military intervention by NATO in Yugoslavia? (IPSOS/Liberation, N= 500 to 900 per country) d - Average level of support across time. Not weighted by the number of cases. e - Difference between the earliest and latest survey available.

na na -6 na na na na na

26 TABLE 2 – US public opinion and the conflict in Kosovo. Support for air strikes (in % support) a 1998, 9-12 October 1999, 19-21 February 1999, 11-14 March 1999, 19-21 March 1999, 23 March 1999, 24 March 1999, 24-28 March 1999, 25 March 1999, 25-26 March 1999, 26-28 March 1999, 28 March 1999, 30 March 1999, 30-31 March 1999, 1 April 1999, 1-2 April 1999, 5 April 1999, 5-6 April 1999, 6 April 1999, 6-7 April 1999, 7-8 April 1999, 8 April 1999, 8-9 April** 1999, 8-13 April 1999, 13-14 April 1999, 14-19 May 1999, 15-18 April 1999, 17-19 April 1999, 21 April 1999, 21-22 April 1999, 22 April 1999, 25-26 April

b

c

d

e

f

g

h

i

j

k

l

m

n

o1

p2

q2

r2

s2

t3

u4

42 43 26 46 47

50

March 24, 1999 - NATO air strikes campaign begins 50 60 60 44 53 53 60 55 55

53 51 58 68

64

58

58

58

55

67 58 54 64 54 61 61

59

55 54 62 60

51

51 49 55 65

56

Row Average 42 43 26 46 47 50 60 49 53 60 55 55 53 51 58 66 57 67 58 54 64 54 61 58 54 62 60 51 49 55 61

27 1999, 26-27 April 1999, 30 April-2 May* 1999, 1-2 May 1999, 7-9 May 1999, 11 May 1999, 12-16 May 1999, 16 May 1999, 19-20 May 1999, 23-24 May 1999, 26-27 May 1999, 2-3 June 1999, 4-5 June 1999, 5-6 June

56 58 53 55 49 53 59

45 49 51 42 53 48 June 10, 1999 - NATO air strikes stop, Serbian forces begin to withdraw from Kosovo 62

1999, 9-13 June 1999, 10 June 1999, 10-15 June 1999, 11-15 June 1999, 23-24 June Column average

54

51

62 55 58 51 55

55 58 51 55 42

45

54

26

47

54

59

44

54

64

62

60

53

58

58

53

55

51

53

49

Cell entries are the percentage of those who support it.

Questions: a- Based on what you have read or heard, do you think the United States and its Western European allies should or should not conduct military air strikes against the Serbian forces in Kosovo? [Gallup] b- If a peace agreement is not reached between the Yugoslavian Serbs and Kosovo’s ethnic Albanian majority, NATO has said it would carry out air and missile attacks against Serb military installations. Would you favour or oppose the U.S. being a part of that military action? [Gallup] c- As you may know, the military alliance of Western countries called NATO, has launched air and missile attacks against Serbian military targets in Yugoslavia. Do you favour or oppose the United States being a part of that military action? [Gallup] d- The United States has said it may bomb Serbia unless Serbia agrees to a peace plan for Kosovo. If Serbia does not agree to the peace plan, should the United States bomb Serbia or not? [ABC News]

e- The United States and its European allies have said they may bomb military targets in Serbia because it won’t agree to a peace plan for Kosovo. If Serbia does not agree to the peace plan, should the United States participate in bombing Serbia, or not? [ABC News] f- Do you favour or oppose the United States and NATO conducting air strikes against Yugoslavia? [Prior to 5-6 April ‘. . . against Serbia?’] [CBS News]

56 58 53 55 49 53 57 45 49 51 42 53 50

52

28 g.- Do/did you approve or disapprove of NATO forces, including the United States, conducting air strikes against Serbia to force the Serbs to agree to the terms of the peace agreement to end the fighting in Kosovo? [Princeton Survey Research Associates for Newsweek] h- Do you approve or disapprove of the NATO (North Atlantic Treaty Organisation) air strikes against Serbian targets in Yugoslavia and Kosovo? [Yankelovich Partners for Time/ CNN] i- Do you approve or disapprove of the decision to send American military troops to be part of the NATO (North Atlantic Treaty Organisation) air operation against the Serbs (in Kosovo)? [Los Angeles Times] j- Do you support or oppose the United States and its European allies conducting air strikes against Serbia? [ABC News]

k - Do you approve or disapprove of the United States and NATO conducting military air strikes against Serbian targets? [Hart and Teeter Companies for NBC News/Wall Street Journal] l- Do you support or oppose the President's decision to launch military air strikes? [Rasmussen Research] m- Do you approve or disapprove of the President's decision to have U.S. military forces join the NATO air strikes in Yugoslavia over the issues of Kosovo? [Princeton Survey Research Associates for Newsweek] n- Do you approve or disapprove of the President’s (Bill Clinton’s) decision to have US (United States) military forces join the NATO (North Atlantic Treaty Organisation) air strikes in Yugoslavia to support the peace plan for the Kosovo province? [Princeton Survey Research Associates for Newsweek] o- Do you think the U.S. government’s bombing Serb targets was the right or the wrong decision? [Asked only of those who have seen, heard or read about recent events in Kosovo] [Louis Harris Associates] p- Do you think the U.S. made a mistake in getting involved in Kosovo, or not? [CBS News] q- Considering everything, do you think the United States did the right thing in getting involved in the military conflict in Serbia, or do you think it was a mistake? [ABC News/Washington Post] r- Do you think it was a mistake for the United States to become militarily involved with Kosovo, or don't you think so? [Yankelovich Partners/CNN/Time] s- Do you think the United States did the right thing in getting involved in Kosovo, or should we have stayed out?’ [CBS News] t- All things considered, do you think the United States should be involved in Kosovo or should we have stayed out of it? [Opinion Dynamics for Fox News] u- In view of the developments since we entered the fighting in Yugoslavia, do you think the United States made a mistake … Sending military forces to fight in Yugoslavia? [Gallup for CNN/ USA Today] Notes: * Based on half sample **… over the issue of Kosovo 1 right decision 2 right thing 3 should be involved 4 No

29

TABLE 3A – French public opinion and the conflict in Kosovo. Support for air strikes Approval NATO military intervention (1) 57 63

1999, 26-27 March 1999, 2-3 April 1999, 6-7 April 72 1999, 9-10 April 70 1999, 17 April 1999, 17-18 April 1999, 6-7 May 63 1999, 21-22 May 62 1999, 2 June Cell entries are the percentage of those who support it.

Approval NATO air strikes (2) 40

Approval French military participation in NATO (3) 59 65

50

Approval French military participation in air strikes (4) 46 58

73 74 55 50

59 57 67

Questions: 1 Do you personally approve or disapprove of the military intervention by NATO in Yugoslavia? Cells rreport percentage who answer Yes totally, Yes Rather [Canal IPSOS/ Journal du Dimanche] 2 Do you agree (completely) or disagree (completely) with the air strikes against Yugoslavia?(CSA Opinion/Le Parisien) 3 Do you approve or disapprove of the participation of France in this intervention by NATO in Yugoslavia? Cells rreport percentage who answer Yes totally, Yes Rather (Canal IPSOS/ Journal du Dimanche) 4 Do you agree (completely) or disagree (completely with the participation of France in the air strikes against Yugoslavia? (CSA Opinion/Le Parisien)

30 TABLE 3B – German public opinion and the conflict in Kosovo. Support for air strikes a

b

c

63 51 1999, 26-27 March 56 1999, 29 March 60 1999, 29-30 March 1999, 1 April 61 1999, 1-3 April 60 1999, 6-7 April 1999, 9 April 59 1999, 10-12 April 1999, 12 April 62 1999, 13-14 April 1999, 16 April 66 1999, 16-17 April 66 1999, 19-20 April 61 1999, 23-24 April 62 1999, 27-28 April 56 1999, 30 April-3 May 58 1999, 4-5 May 59 1999 6-7 May 52 1999, 10-11 May 61 1999, 14-15 May 57 1999, 18-19 May 59 1999 21-22 May 57 1999, 25-26 May 62 1999, 28-29 May 63 1999, 31 May-1 June Cell entries are the percentage of those who support it.

d

e

f

62

63 53 49

64

Questions: a- Do you think it is right that NATO has intervened militarily in the Kosovo conflict with air strikes? [EMNID] b- Do you support the air attacks of NATO on Serbian military targets in Yugoslavia ? [Der Spiegel] c- Do you support the use of NATO forces against the Serbians? [Focus] d- Do you support that NATO carries out air attacks against Serbian positions? [Focus] e- [favor air strikes?] [Dimap/ Die Welt] f- Do you think it right that the Bundeswehr also takes part in the air attacks of NATO? [Die Woche]

31 TABLE 3C – British public opinion and the conflict in Kosovo. Support for air strikes a

b

c

56 1999, ? 65 1999, end of March 58 1999, 24-28 March 1999, 24 March-14 April 1999, 26 March 69 1999, 26-27 March 1999, 28 March 1999, 30-31 March 75 1999, 1-2 April 1999, 2 April 1999, half April? 1999, half April? 1999, 16-17 April 1999, 30 April-1 May Cell entries are the percentage of those who support it.

d

e

f

g

h

68 56 55 65 76 57 55 57 70

Questions: a- Do you support or oppose the air strikes? [ICM/ Guardian] b- Do you support or oppose the air strikes? [Gallup/ Daily Telegraph] c- Do you support or oppose the air strikes? [Marplan/ Sunday Times] d- Should air strikes continue? [ICM/ Guardian] e- Should air strikes continue? [NOP/ GMTV] f- Do you support or oppose the recent US/NATO military action over Kosovo? [USIA] g- Do you support or oppose the involvement of British troops in the NATO bombing campaign? [ICM/ Observer] h- As you may know, NATO forces are currently taking part in air raids in Yugoslavia. On balance, do you believe that Britain is right or wrong to have joined in the NATO bombing of Yugoslavia? [MORI/ Mail on Sunday]

32 TABLE 3D – Italian public opinion and the conflict in Kosovo. Support for air strikes a1

b

1999, 25 March 38 1999, 26 March 1999, 26-28 March 52 1999, 29 March 1999, 31 March 49 1999, 7 April 1999, 12 April 48 1999, 13 April 1999, 14 April 1999, 15-18 April 35 1999, 15-25 April 43 1999, 17 April 1999, 18 April 46 1999, 19 April 1999, 24 April 45 1999, 27 April 44 1999, 5 May 43 1999, 10 May 39 1999, 17 May 37 1999, 24 May 42 1999, 25 May 39 1999, 31 May 40 1999, 3 June 46 1999, 9 June 55 24 1999, 25 June Cell entries are the percentage of those who support it.

c

d

37

27

37 40 62 49 45*

42 43

Questions: a- NATO is now bombing Serbia. What do you think, NATO should be supported until the end, Italy should take a critical position within NATO or Italy should reject the position of NATO? [SWG] b- Are you much in favour, somewhat in favour, neither for nor against, somewhat against or much against the bombing actions by NATO on Serbia and Kosovo? [SWG] c- Do you support or oppose the NATO air strikes against Yugoslavia? [Datamedia, USIA and Pragma]

d- Do you think that air strikes by NATO are justified or not? [Ispo/Cra Neilsen survey] Notes: 1 NATO should be supported until the end; * Pragma survey

33 TABLE 4 – Support for air strikes and ground operations in five countries (in %)

Country France Germany Italy United Kingdom United States

NATO Intervention1 Support Oppose DK 54 57 47 68 68

34 36 47 23 27

12 7 6 9 5

Ground troops2 Support Oppose 55 27 33 54 52

32 63 63 30 45

DK 13 9 5 15 3

Difference in support +1 -30 -14 -14 -16

1 ‘As you may know, NATO has recently taken military action in Kosovo. Do you support or oppose NATO's decision to carry out air and missile attacks against Serbian military installations? (AngusReid group/ The Economist, in the period March 25-April 17, N= about 500 per country) 2 ‘Now, if there is no settlement to end the war in Kosovo, would you support or oppose NATO sending ground troops in to fight against Yugoslav forces?’ (Angus Reid group for CNN, in the period 22-25 April 1999, N= 305 (France), 476 (Germany), 302 (Italy), 300 (UK), 300 (USA)).

TABLE 5 – Support for air strikes and ground operations in twelve countries (in%)

Country

Support

NATO Intervention1 Oppose

DK

Support

Ground troops2 Oppose

DK

Difference in support -14 -12 -18 -4 -15 -34 +1 -1 -9 -23 -6 -3

41 43 16 27 58 15 Austria 53 35 12 41 48 11 Belgium 70 20 10 52 37 11 Denmark 44 43 13 40 45 15 Finland 68 27 5 53 42 5 France 52 40 8 18 78 4 Germany 2 97 1 3 96 1 Greece 46 42 12 45 41 14 Ireland 37 46 17 26 59 15 Italy 61 30 9 38 52 10 Luxemburg 34 48 18 28 54 18 Spain 54 33 13 51 36 13 United Kingdom 1 ‘Do you personally support or oppose the military actions by NATO in Sebia? (ICM/European Barometer/The Guardian, in the period May, N= not available) 2 ‘Do you personally support or oppose the use of ground forces in Kosovo?’ (ICM/European Barometer/The Guardian, in the period May, N= not available).

34 TABLE 6 – US public opinion and the conflict in Kosovo. Support for ground operations a 1999, 14 March 1999, 19 March 1999, 23 March 1999, 24-28 March 1999, 25 March 1999, 25-26 March 1999, 28 March 1999, 30 March 1999, 30-31 March 1999, 1 April 1999, 1-2 April 1999, 5 April 1999, 5-6 April 1999, 6 April 1999, 6-7 April 1999, 8 April 1999, 8-13 April 1999, 13-14 April 1999, 15-18 April 1999, 15-16 April 1999, 15-18 April 1999, 21-22 April 1999, 23-28 April 1999, 26 April 1999, 26-27 April 1999, 29 April 1999, 30 April- 2 May 1999, 1-2 May 1999, 5-6 May 1999, 12-16 May 1999, 14-19 May 1999, 16 May

b

c

d

e

f

g

h

i

j

k

l

m

n

o

p

q

r

s

50 30 37 44 31 47 33 24 39 41 54 55 46 57 47

41 57 62

52

43

45

47` 57 51 38

45

46

42 56 40

36

33 40 45 45

27 44

56 52

Row average 50 30 37 44 31 47 33 24 39 41 54 55 46 57 44 57 62 47 47 57 51 43 42 56 38 33 40 45 36 44 56 52

35 30

1999, 19-20 May 1999, 23-24 May 1999, 26-27 May 1999, 2-3 June 1999, 5-6 June Column average

30 40 43 27 54

40 43 27 54 44

30

31

29

45

47

40

42

38

28

55

40

40

43

44

44

51

52

59

Cell entries are the percentage of those who support it. Questions: a- Would you support or oppose the US, along with its allies in Europe, sending in ground forces to try to enforce a peace treaty in Kosovo? [ABC News] b- If the Serbs launch an armed offensive against the Kosovo Albanians, should the United States intervene by sending ground troops into battle? [Rasmussen] c- If the current NATO air and missile strikes are not effective in achieving the United States’ objectives in Kosovo, would you favour or oppose President Clinton sending U.S. ground troops into the region to stop the Serbian attack on Kosovo [Gallup] d- Would you favour or oppose sending American ground troops into Kosovo? [Rasmussen] e- If the current NATO air and missile strikes are not effective in achieving the United States’ objectives in Kosovo, would you favour or oppose President Clinton sending U.S. ground troops into the region along with troops from other NATO countries? [Gallup] f- If the air strikes do not stop Serbian military attacks in Kosovo, would you favour or oppose sending U.S. ground troops to Kosovo along with troops from other NATO countries [to try to end the conflict in Kosovo]?, [half sample each] [Princeton Research Associates for Pew Research Center] g- If NATO (North Atlantic Treaty Organisation) decides to send ground troops to Kosovo, do you favour or oppose sending US (United States) ground troops as part of the NATO force? [Opinion Dynamics/ Fox News] h - Suppose US (United States) and NATO (North Atlantic Treaty Organisation) goals are not reached through the bombing campaign, would you favour or oppose the United States and its NATO allies sending in ground troops to fight in Kosovo? [ICR Survey Research Group for National Public Radio/ Henry J. Kaiser Family Foundation, Harvard University's Kennedy School of Government] i- If the air strikes are not effective in stopping the Serbian attacks, would you favour or oppose President Clinton sending U.S. ground troops into Bosnia to join ground troops from other Western European countries? [Based on half sample] [Gallup] j- Should the U.S. send ground troops into combat against the Serbian Army in Kosovo? [Opinion Dynamics/ Fox News] k- Suppose the bombing does NOT stop Serbia’s military action in Kosovo. Would you support or oppose the United States and its European allies sending in ground troops to try to end the conflict in Kosovo? [ABC News] l - Now thinking about the current situation in Kosovo, would favour or oppose sending U.S. ground troops, along with troops from other NATO countries, to serve in a combat situation in the region right now? [Gallup] m- Do you favour or oppose sending US (United States) ground troops, along with troops from other NATO (North Atlantic Treaty Organisation) countries, in to the current situation in Yugoslavia? [Gallup] n- Do you favour or oppose sending US (United States) ground troops along with troops from other NATO (North Atlantic Treaty Organisation) countries to serve in a combat situation in the region (of Kosovo)? [Yankelovich Partners for CNN/Time]

36 o- In order to try to end the fighting in Kosovo, would you favour or oppose the United States sending in ground troops as part of a NATO peacekeeping force? [CBS News] p -There's been some talk about sending 4,000 U.S. troops, as part of a NATO peacekeeping force of about 30,000 troops, to help secure peace in Kosovo, Serbia. Do you approve or disapprove of sending U.S. troops to Kosovo? [Princeton Research Associates for Pew Research Center] q- Please tell me whether or not you would support sending US (United States) ground troops to Yugoslavia in each of the following circumstances: What about ... to force Milosevic to comply fully with the peace plan? [Princeton Survey Research Associates for Newsweek] r- Please tell me whether or not you would support sending US (United States) ground troops to Yugoslavia in each of the following circumstances: What about ...to force Milosevic to agree to a NATO (North Atlantic Treaty Organisation) peace plan? [Princeton Survey Research Associates for Newsweek] s- If NATO (North Atlantic Treaty Organisation) decides to send in a multinational force to protect the people of Kosovo from the Serbs and the Yugoslav army, would you support or oppose sending US (United States) troops into Kosovo as part of this NATO force? [Louis Harris Associates]

37 TABLE 7A – French public opinion and the conflict in Kosovo. Support for ground operation a 41 1999, 26-27 March 1999, 29 March 1999, 1-2 April 58 1999, 2-3 April 1999, 9 April 65 1999, 9-10 April 1999, 11 April 64 1999, 17 April 1999, 19 April 1999, 25 April 55 1999, 21-22 May Cell entries are the percentage of those who support it.

b

58 55 63 63 63 56

Questions: a- If the air strikes turn out to be insufficient to solve the crisis, should NATO decide to intervene in Kosovo with ground forces? [Canal IPSOS/ Journal du Dimanche] b- If the air strikes turn out to be insufficient to solve the crisis, should NATO decide to intervene in Kosovo with ground forces? [IFOP/ L’Express/ dimanche Ouest-France]

38 TABLE 7B – German public opinion and the conflict in Kosovo. Support for ground operations a 1999, Early March 1999, 26 March

b

c

d

e

f

28

g

h

47 28

35 1999, 26-27 March 38 1999, 29-30 March 1999, 1 April 36 1999, 1-3 April 35 1999, 6-7 April 33 1999, 8-10 April 1999, 9 April 27 1999, 10-12 April 1999, 12 April 31 1999, 13-14 April 1999, 16 April 30 1999, 16-17 April 31 1999, 19-20 April 33 1999, 23-24 April 34 1999, 27-28 April 27 1999, 30 April-3 May 28 1999, 4-5 May 31 1999 6-7 May 30 1999, 10-11 May 30 1999, 14-15 May 24 1999, 18-19 May 30 1999 21-22 May 29 1999, 25-26 May 32 1999, 28-29 May 25 1999, 31 May-1 June Cell entries are the percentage of those who support it.

36 28

47

31

48

46

29 33

46

Questions: a- Would you support or oppose if the Bundeswehr would, if necessary, take part with ground forces in enforcing the peace? [EMNID] b- Do you favour or oppose to commit ground troops to Kosovo (if air strikes are not sufficient to restore peace)? [Forsa] c- Would you support in an emergency the commitment of NATO ground forces in Kosovo? [Der Spiegel] d- Should NATO ground troops enter into Kosovo? [Die Woche] e- Do you support that NATO also commits ground forces? [Focus] f- Should the Bundeswehr take part in a ground offensive? [Die Woche] g- If this should nevertheless be decided, should Germany then participate in this action? [Forsa] h- Should German troops also participate if ground forces are engaged in the war in Kosovo? [Der Spiegel]

39

TABLE 7C – British public opinion and the conflict in Kosovo. Support for ground operations a

b

c

52 1999, 24-28 March 1999, 26 March 1999, 26-27 March 62 1999, 26-27 March 1999, 30-31 March 1999, 1-2 April 47 1999, 2 April 1999, 16-17 April 51 1999, 30 April-1 May 1999, half April ? Cell entries are the percentage of those who support it.

d

e

39 51 58 66 50 48

Questions: Britain should send ground troops into the conflict [as part of a NATO ground force] a Gallup/ Daily Telegraph b MORI/ Mail on Sunday c Marplan/ Sunday Times d ICM/ Guardian e NOP/ GMTV

TABLE 7D – Italian public opinion and the conflict in Kosovo. Support for ground operations

1999, 7 April 1999, 13 April 1999, 17 April 1999, 19 April 1999, 27 April 1999, 5 May 1999, 10 May 1999, 17 May 1999, 25 May 1999, 31 May 1999, 25 June

a 33 31 32 30 33 28 32 27 41

b 33 32 31 28 32 27 30 30 31 32 39

Cell entries are the percentage of those who support it. Questions: a- There is a discussion on the possible intervention of NATO soldiers in Kosovo. Are you in favour or against such an initiative? [SWG] b- Are you in favour or against the participation of Italian forces in such an operation? [SWG, except June 25 Nomesis]

40 TABLE 8 - Which role should NATO play in the conflict around Kosovo/Yugoslavia? (in %)

‘Thinking now of the crisis in Kosovo in former Yugoslavia, which of the following comes closest to your views of what kind of role N play there? Do you think NATO should send more armed forces in to the area; NATO should do more to seek a diplomatic solution but send more troops; NATO should keep its involvement at the current levels; NATO should reduce its involvement in Kosovo; or NATO completely end any involvement in Kosovo?’

Send more armed forces Canada Denmark France Germany Italy Norway United Kingdom United States

30 26 12 12 11 24 29 27

Keep present level

More diplomatic effort Original NATO members 19 35 27 35 17 51 29 47 12 59 22 37 23 33 23 30

Reduce involvement 4 1 2 4 2 3 2 4

New NATO members Czech Republic Hungary Poland

11 19 18

12 14 10

45 38 48

8 6 6

Other European Countries Croatia Finland Russia Slovakia Ukraine

39 20 4 1

21 7 1 8 4

Source: Angus Reid group, 25 March-17 April 1999 N=about 500 per country

19 56 40 55 43

1 4 4 5 5

End in

41 TABLE 9 - Bombing or return to diplomacy? Opinions in Ital (in %) Prime Minister D’Alema has said that after the first bombing the moment has come to return to diplomacy, while Mr Clinton and Mr Blair want to continue the bombing until Serbia sign the peace agreement. Which position do you support? Date 26/3 29/3 7/4 13/4 17/4 19/4 27/4 5/5 10/5 17/5 72 66 60 63 67 65 63 67 62 69 Return to diplomacy 22 31 34 30 28 29 31 33 34 28 Continue the bombing 6 4 6 7 6 6 6 5 4 3 Don’t Know 100 100 100 100 100 100 100 100 100 100 Total Source: SWG-Trieste

TABLE 10 - The goals of the NATO operation (in % ) “[Finally], do you feel that one of the goals of NATO's campaign should be to create an independent state of Kosovo, Or should NATO's goal be to simply keep Kosovo a separate governed province of Yugoslavia?” France Germany Italy United United All Kingdom States Independent state 40 29 41 31 31 33 Separately governed 42 45 47 37 53 48 province Don’t know 18 25 12 31 16 19 Total 100 100 100 100 100 100 Source: Angus Reid Group for CNN, 22-24 April 1999 (N= 305 (France), 476 (Germany), 302 (Italy), 300 (UK), 300 (USA))

42 TABLE 11 - Attitudes toward Milosevic (in %) Milosevic a war criminal?: “Some people have accused President Milosevic of being a war criminal. They say that he started the war and that he has persecuted the Albanian people of Kosovo. President Milosevic strongly denies these charges. In your view, should President Milosevic be charged and tried as a war criminal?” France Germany Italy United United All Kingdom States Yes 84 81 83 82 79 81 No 6 5 11 7 15 11 Don’t know 10 14 6 10 6 8 Total 100 100 100 100 100 100 Milosevic to be removed from office?: “Now, as you may know, President Milosevic heads the government of Yugoslavia, which NATO accuses of starting the war in Kosovo. In your view, should a settlement in Kosovo require that Pres. Milosevic be removed from office?” France Germany Italy United United All Kingdom States Yes 77 86 78 83 82 82 No 13 6 16 12 15 13 Don’t know 10 7 5 6 3 5 Total 100 100 100 100 100 100 Source: Angus Reid Group for CNN, 22-24 April 1999 (N= 305 (France), 476 (Germany), 302 (Italy), 300 (UK), 300 (USA))

TABLE 12 - Has NATO done a good job? (in %) Overall, do you feel that NATO has done a very good, good, poor or very poor job of managing its military campaign against Yugoslavia? France Germany Italy United United All Kingdom States Very Good 7 6 8 9 13 10 Good 53 53 51 55 58 56 Poor 20 17 21 19 19 19 Very Poor 6 4 13 7 6 6 Don’t’ know 13 20 8 10 4 9 Total 100 100 100 100 100 100 Source: Angus Reid Group for CNN, 22-24 April 1999 (N= 305 (France), 476 (Germany), 302 (Italy), 300 (UK), 300 (USA))

43 TABLE 13 - Air strikes are enough? (in %)

1999, 25-26 Marcha 1999, 1-2 Aprila 1999, 5 Aprilb 1999, 5-6 Aprilb 1999, 8-9 Aprila 1999, 15-18 Aprila

Air strikes enough

Ground forces will be required

Don’t know/ No answer

N

21 12 25 27 15 19

60 71 68 66 71 65

19 17 7 7 14 16

752 750 509 1011 751 1000

Question wording: a “Do you think the air strikes will be enough to make Yugoslav President Milosevic to agree/to fully comply to the NATO peace plan, or do you think ground forces will be required?” (Princeton Survey Research Associates for Newsweek) b “Just your best guess, do you think air strikes will be enough to stop Serbia from continuing its attack in Kosovo, or do you think ground troops will have to be sent to try to end the conflict?” (ABC News for Washington Post)

TABLE 14 - NATO actions as a success or a failure under different time frames? (in %) From what you've heard or read, do you think the current NATO military action in Yugoslavia has been a success or a failure? Success Failure Too soon to Don’t know/ tell (vol.) No answer 37 41 15 7 1999, 6-7 Aprila 37 46 9 8 1999, 13-14 Aprila 35 47 9 9 1999, 26-27 Aprila Source: Gallup (N= respectively 1.055, 1.069 and 1.073)

Looking ahead to when the NATO military action in Yugoslavia has ended, do you think it will be a success or a failure? Success Failure Too soon to Don’t know/ tell (vol.) No answer 50 35 8 7 51 35 8 6 49 37 6 8

44 TABLE 15 - Worries about casualties in USA (in %) How worried are you that US troops in Kosovo might suffer casualties: very worried, somewhat worried, not too worried, or not at all worried? March, 24-28, 1999 April, 15-18, 1999 May, 12-16, 1999 Very worried 55 66 56 Somewhat worried 31 25 31 Not too worried 9 5 7 Not at all worried 3 3 3 Don’t know 2 1 3 Total 100 100 100 Source: Pew Research Center Poll conducted by Princeton Survey Research Assoc. March 24-28 1999. N=1488

TABLE 16 - Would it be worth to suffer casualties? (in %) Do you think peace in Yugoslavia is worth the loss of American life and the other costs of attacking Yugoslavia, or not? 1 March 24, 1999 March 28, 1999 April 1, 1999 April 5-6, April 13-14, 1999 1999 Worth it 34 36 31 44 32 Not worth it 52 54 52 45 52 Don’t know 14 10 17 11 16 Total 100 100 100 100 100 (N) (527) (na) (na) (na) (878) Please tell me if you agree or disagree with the following statement: It would be worth the loss of some American soldiers' lives if the United States could help bring peace to Kosovo. 2 March 23, 1999 March 26, 1999 April 5-6, 1999 Agree 37 45 45 (31)* Disagree 59 49 52 Don’t know 4 6 4 Total 100 100 100 (N) (518) (na) (1011) Sources: 1 CBS News Poll. 2 ABC News Poll/Washington Post * Between parentheses: if answered ‘worth’ and ‘worth if 100 soldiers killed’.

45 TABLE 17 - Casualties and support for sending ground troops (in %) Suppose the bombing does NOT stop Serbia’s military action in Kosovo. Would you support or oppose the United States and its European allies sending in ground troops to try to end the conflict in Kosovo? Support Oppose Don’t know Total 55 41 4 100 April 5, 1999 57 39 5 100 April 6, 1999 57 39 4 100 April 8, 1999 44 50 5 100 If some casualties* 37 57 6 100 If 100 casualties** 31 62 7 100 If 500 casualties*** 26 66 7 100 If 1,000 casualties**** 56 40 4 100 April 26, 1999 45 52 3 100 If some casualties 52 46 2 100 May 16, 1999 43 56 1 100 If some casualties * (If ‘support’ in question): “Would you support or oppose sending in ground troops if there was a good chance that some US soldiers would be killed in the fighting?” ** (If ‘some’): Would you support or oppose sending in ground troops if there was a good chance that up to 100 US soldiers would be killed?” ***(If ‘100’): … up to 500? **** (If ‘500’): …up to 1,000? Source: ABC/Washington Post

TABLE 18 - Support for sending ground troops (in %) Given what you know right now, which would you prefer the United States and NATO to do in the conflict in Kosovo: Send ground troops 21 If the answer was ‘send ground troops’: “What if sending ground troops meant there would be a lot of American or NATO casualties? Then, what would you prefer?” Send ground troops 12 Continue the air strikes but don’t send ground troops 7 withdraw forces before becoming more deeply involved 1 Don’t know 1 Total Continue the air strikes but don’t send ground troops Withdraw its forces before becoming more deeply involved Don’t know Total CBS News

44 30 5 100

21

46 TABLE 19 - Support for sending ground troops under different conditions (in %) “ Is it better to restrict the bombing/restrict ground troops even if it means a longer war or is it better to increase the bombing/increase ground troops even if it means risking American and NATO casualties in order to bring a quick end to the war?” Bombing Troops Restrict 48 57 Increase 42 33 Don’t know 10 9 Total 100 100 CBS News

TABLE 20 - Reactions to casualties in Kosovo (in %) ‘Imagine that in the course of carrying out such a ground war [ten OR fifty] Americans were killed in a battle. Imagine that you saw their dead bodies on television. How do you imagine that you would react? Would you want to:’ Ten Americans killed Fifty Americans killed Withdraw all American troops 21 20 Strike back hard at the attackers 19 19 Bring in reinforcements so that future attacks can be met 33 35 with overwhelming force Not react in any of these ways, but stay the course 21 19 Don’t know 7 8 Total 100 100 Source: PIPA, May 1999

TABLE 21 Casualties and success (in %) ‘Imagine that over the course of the ground war c) 25/ d) 250 Americans were killed but the operation succeeded in driving Serb forces out of Kosovo so that ethnic Albanian refugees could return. Do you imagine that you would think that NATO did or did not do the right thing by going into Kosovo with ground troops?’ 25 Americans killed 250 Americans killed Did the right thing 65 60 Did not do the right thing 32 33 Don’t know/ No answer 3 7 Source: PIPA, May 1999

47

FIGURE 1 – Support for air strikes in the United States

100 90 80

% in favor

70 60 50 40 30 20 10 0 09/02/99

01/03/99

21/03/99

10/04/99

30/04/99

20/05/99

09/06/99

29/06/99

Time c

f

k

r

c- As you may know, the military alliance of Western countries called NATO, has launched air and missile attacks against Serbian military targets in Yugoslavia. Do you favour or oppose the United States being a part of that military action? [Gallup] f- Do you favour or oppose the United States and NATO conducting air strikes against Yugoslavia? [Prior to April, 5-6 “. . . against Serbia?”] [CBS News] k- Do you support or oppose the United States and its European allies conducting air strikes against Serbia? [ABC News] r- Considering everything, do you think the United States did the right thing in getting involved in the military conflict in Serbia, or do you think it was a mistake? [ABC News/Washington Post]

48 FIGURE 2 - Support in Italy for sending ground forces, Italian participation in such an operation and the incidence of casualties

50 45

35 30 25 20 15

NATO Intervention

Italian intervention

Italian with casualties

20-5-99

15-5-99

10-5-99

5-5-99

30-4-99

25-4-99

20-4-99

15-4-99

10-4-99

10 5-4-99

Percent support

40

49 FIGURE 3 Support for the air strikes, German participation in sending ground forces and expected success in Germany 70 60 50 40 30 20 10

Support air stikes Milosevich will be forced Bundeswehr in ground force

28-5-99

21-5-99

14-5-99

7-5-99

30-4-99

23-4-99

16-4-99

9-4-99

2-4-99

26-3-99

0

50 NOTES Chapter 1 1

NATO issued another ultimatum on September 24, 1998 to stop Serbian repression in Kosovo, to which eventually Milosevic complied on October 27, 1998. But violence erupted again with the new year and the discovery of 45 bodies of Albanians in Racak.

2

To mention an example, even in a country in which public opinion is claimed to exert no role in foreign policy, such as Italy, the prime minister at that time, Massimo D’Alema, found the time, at the end of the conflict, to write a short book titled ‘L’opinione pubblica e la Guerra in Kosovo’ (Public Opinion and the Kosovo war) in which he tried to justify his action in front of the ‘tribunal’ of public opinion. For a discussion on the role of public opinion in the Italian parliamentary system see Isernia (2000).

3

The September 11th events confirm this impression. For evidences bearing on this point see Everts and Isernia (2002).

4

Interviews were carried out in seventeen countries (Canada, Croatia, Czech Republic, Denmark, Finland, France, Germany, Hungary, Italy, Norway, Poland, Russia, Slovakia, Ukraine, United Kingdom and United States).

5

Interviews were carried out in five countries (France, Germany, Italy, United Kingdom and United States).

6

Interviews were carried out in six countries (Croatia, Czech Republic, Hungary, Poland, Rumania, Slovenia).

7

Interviews were carried out in eight countries (Belgium, France, Germany, Greece, Italy, Portugal, Spain and United Kingdom).

Chapter 3a 1

One other poll (NIPO, end of March (N=?) and 17-18 April 1999 (N=1436)) even measured 78% agreement with the actions.

2

Croatian public opinion was also the more optimistic, among the six Eastern European countries polled by Gallup Eastern Europe in early March 1999, in assessing NATO ability ‘to prevent the military conflict in Kosovo and at the Albanian-Serb border’, with 84 percent answering yes.

3

Another poll by RFE/RL (N=?) of late March found 48 percent in favour and 36 percent opposing NATO air strikes.

4

On April 8, however, a Gallup Hungary survey (N=506) found 61 percent in support of the operation and 33% opposing it.

5

Another Public Opinion Foundation survey in Russia on April 9 1999 (N=1500) shows only 2% in favour of NATO air strikes and 92% opposing them.

6

But a INRA/Le Soir survey of April 9-12 1999 (N=502) found 64% in support and 20% opposing the NATO air strikes.

7

On April 22 1999 The Times reported a survey’s results in Spain with 55% in support and 39% against NATO air strikes. The exact date, the survey agency and the sample size were not reported.

8

SIFO Research for Aftonbladet, 31 March 1999, N=1000.

9

The questions were respectively: ‘Do you think it is right that NATO has intervened militarily in the Kosovo conflict with air strikes?’ [EMNID] and ‘Do you think it right that the Bundeswehr also takes part in the air attacks of NATO?’ [Die Woche].

51

10

Canal IPSOS, 26-27 March, 2-3, 9-10 and 17 April and 21-22 May 1999.

11

In still another survey 58 percent expressed support for French participation in the international action, even though the question referred explicitly to ‘on the side of NATO’, which is traditionally controversial in France (IFOP for Dimanche Ouest-France, 1-2 April (N=967)). Text of the question: ‘Vous personellement, approuvez –vous ou désapprouvez-vous l’engagement militaire de la France aux côtés de l’Otan dans l’affaire du Kosovo?’

12

CSA Opinion for Le Parisien, 26-27 March (N=1003), 6-7 April (N=1033), 17-18 April (N=?) and 6-7 May (N=1002).

13

NIPO for 2Vandaag TV, 7 April 1999 (N=251).

14

SWG for Palazzo Chigi, 23 March 1999 (N=??).

15

INRA Belgium for Le Soir, 9-12 April 1999 (N=502).

16

According to The Times, 22 April 1999.

17

21 and 24% respectively in other polls by Lagendijk for Studiecentrum Vredesvraagstukken Nijmegen, early May 1999.

18

According to The Times, 22 April 1999.

19

NIPO, 6 April 1999 (N=251).

20

Canal IPSOS, 26-27 March, 2-3, 9-10 and 17 April and 21-22 May 1999 (N= 1033, ?, 946, 971, ?). The figures are confirmed by a series of surveys held by IFOP in the same period (29 March, 1-2 , 9, 11 and 25 April 1999 (N=830, 1000, 1000, 961, 1000)) with 58, 55, 63, 63 and 66% support respectively even though the question again referred explicitly to NATO. Text of the question: ‘Dites-moi si vous seriez plutôt favorable ou plutôt défavorable à une intervention terrestre de l’Otan si les frappes aériennes de se révélaient insuffisantes pour faire cesser les massacres au Kosovo?’ See for an example of decline CSA 17-18 April and 6-7 May 1999.

21

Angus Reid, 25 March - 17 April 1999.

22

Archivio Disarmo/SWG, 15-25 April 1999 (N=1007).

Chapter 3b 1

But apparently not enough, if only 59% of the Americans knew the American were bombing the Serbs (with 32% not sure and 9% mentioning the Albanians) on the 24-25 of March (the day after the launch of air strikes) and this percentage was up only to 69% on April 7 (with 24% not sure and 7% mentioning the Albanians) (Opinion Dynamics/Fox News, 24-25 March 1999, N=921 and 7 April 1999, N=921).

2

To confirm the lack of a perceived serious direct threat to US interests, the day after the initiation of air strikes, Rasmussen Research asked ‘Which situation poses a greater threat to the United States, the Kosovo situation or the theft of nuclear secrets by the Chinese government?’. Only 13% mentioned Kosovo and 68% the theft of secrets by the Chinese government (Rasmussen Research, March 25, N=1000).

3

NIPO, 6 April 1999 (N=251).

4

ICM/The Observer (N=517).

5

MORI for Mail on Sunday, 26-27 March 1999 (N=606).

52

6

Interestingly, no such a question was asked in the United States this time. At the time of the Gulf war, between 44 and 47 percent were in favour of planning Saddam Hussein’s assassination (Mueller, 1993: 264).

7

CSA Opinion for Le Parisien, 26-27 March (N=1003).

8

CSA for L’Evenement du jeudi, 2-5 April 1999 (N=957).

9

Infratest dimap, Deutschlandtrend, 31 March-1 April 1999 (N=1300). Text of the question: ‘Who carries the main responsibility for the extension of the Kosovo conflict: Serbia, the independence movement of Kosovo Albanians UCK, or NATO?’

10

Gallup for Daily Telegraph, 24-29 March (N=700).

11

MORI for Mail on Sunday, 26-27 March 1999 (N=606).

12

Archivio Disarmo/SWG, 15-25 April 1999 (N=1007.

13

Gallup Hungary, 8 and 15 April 1999 (N=510 and 601).

14

Text of the question: ‘All in all, do you consider that three/four weeks/two months after its beginning the military intervention by NATO is going to succeed or fail?’ (CANAL IPSOS).

15

Text of the question: ‘Do you think today that the NATO action to find a just and durable solution in Kosovo has turned out to be positive or negative?’ (CSA Opinion for Le Parisien, 6-7 April (N=1033)).

16

EMNID, 26 February-2-3 March 1999 (N=2000).

17

Like the French, the Germans made no distinction between support in general and German participation in the air strikes (Politbarometer Forschungsgruppe Wahlen – Mannheim, 12-15 April and 14 May 1999 (N=1267, ?))).

18

IFOP for Dimanche Ouest-France, 11 April (N=930) and idem, 15 May 1999 (N=952) for Libération (20 May 1999).

19

Poll for Der Spiegel, 29-30 March 1999 (N=1000).

20

NIPO for Stichting Maatschappij en Krijgsmacht. 4-6 June 1999 (N=?).

21

Polls for Bundesministerium der Verteidigung, 11-12, 15-16 and 18-19 June 1999 (N=?).

22

Princeton Research Associates for PEW, 15-18 April 1999 (N=1000).

23

NIPO/ 2 Vandaag TV (N=?).

24

Gallup/CNN/USA Today, 6-7 April 1999 (N=1055).

25

Hart and Teeter Research Companies for NBC News/ Wall Street Journal, 18-19 April 1999 (N=1006).

26

NIPO, end of March 1999 (N=?).

27

Source of this poll unknown.

28

Polls by ICM for The Observer of 26 March (N=517) and by MORI for Mail on Sunday of 1 April 1999 (N=604).

29

MORI, 28 March (N=606), 2 April (N=604) and 23-26 April 1999 (N=?).

53

30

SWG, 15-25 April (N=1007), 24 May (N=465) and 25 June 1999 (N= 502).

31

Metron, 7 April 1999 (N=?).

32

V-PRC/ Ta Nea, ca 15 April 1999 (sample of Athens area, N=?).

33

MORI for Mail on Sunday, 1 April 1999 (N=604). Cf also Gallup poll for Daily Telegraph, 24028 March 1999 (N=700) with a similar result.

34

In the Louis Harris polls with a similar question (8-13 April, 14-19 May and 10-15 June 1999 he did slightly better.

35

Various polls by CBS/The New York Times, Opinion Dynamics/Fox News, Princeton Survey Research for PRC and ICR Survey Research Group for National Public Radio/Heny J.Kaiser Family Foundation and Harvard University.

36

Los Angeles Times, 25 March 1999 (N=544), Opinion Dynamics/Fox News, 7-8 April (N=921) and Rasmussen Research, 5 May 1999 (N=1000).

37

CBS News, 24 March (N=527), Gallup/CNN/USA Today, 25 March (N=675), Opinion Dynamics/Fox News, 14-25 March (N=921), 7-8 April (N=921), 21-22 April (N=942) and 5-6 May 1999 (N=901).

38

Gallup, 30-31 March (N=1078), 6-7 April (N=1055),13-14 April (N=1069), 26-27 April (N=1073) and 11-13 June 1999 (N=1072).

39

CBS for New York Times,(N=811).

40

Hart and Teeter for NBC News/ Wall Street Journal (N=1006).

41

Opinion Dynamics for Fox News, 7-8 April (M=921), 21-22 April (N=942), 5-6 May (N=901) 2-3 June (N=924) and 23-24 June (N=908).

42

ABC News/Washington Post, 5-6 April (N=1011). When this question was repeated on 25-26 April (N=757) these percentages were 23 and 36 respectively.

43

NIPO, October 1998 (N=?).

44

NIPO for Stichting Maatschappij en Krijgsmacht, 3-5 April (N=980) and 4-6 June 1999 (N=?).

45

Gallup Poll of 24-28 March 1999 (N=700) for Daily Telegraph (Gallup Organization, Poll releases, 30 March 1999.

46

Survey by MORI on 26-27 March (N=604) and 1 April 1999 (N=604) for Mail on Sunday.

47

SWG, 24 May 1999 (N=465).

48

Source: CBS polls.

49

CBS, 5-6 April 1999 (N=811).

50

The Harris Poll. April 8-13 1999 (N=1006). Perhaps this outcome was due to the double negation.

51

The Los Angeles Times Poll. March 25 1999 (N=544).

52

Gallup, 6-7 April 1999 (N=1055).

53

Fox News/Opinion Dynamics Poll, 24-25 March 1999, N=900. See also another outcome (Rasmussen, 8 April 1999 (N=1000)) ‘Suppose you knew that to accomplish our mission in Kosovo, military action would lead to the

54

death of 100 American soldiers. If this was the case, would you favour or oppose sending American ground troops into the region? 29% ‘favour’, 54% ‘oppose’, 17% d.k/n.a.. 54

Andrew Kohut, Pew Research Center, cited by Dan Balz, Washington Post, 6 April 1999, p.A01.

55

G. Langer and B. Fitzpatrick: http://abcnews.go.com/sections/world/DailyNews/kosovopoll990409.htm.

56

CBS, 5-6 April 1999 (N=811).

57

Poll of 13-17 May 1999 for Program on International Policy Attitudes (www.pipa.org/kosovq.html) (N=1206).

58

See for a similar conclusion also Giacometti, P. and Dupin, E. (1999), Débat: intervention de l’OTAN en Yougoslavie: les determinants de l’opinion publique (www.ipsos.com).

59

See for a comparable conclusion Kohut, A. (1999), ‘Beware of polls on the war’, New York Times, 8 April.

60

Frank Newport, ‘Public support for US Involvement in Yugoslavia lower than for Gulf war, other foreign engagements’, Gallup Poll releases, March 30 1999 (www.gallup.com/poll/releases/pr990330.asp).

61

Ignatieff (2000), p. 183.

62

Lindsay (2000).

63

Kohut (1999).

64

See the EMNID polls reported in figure 3 and the data from Politbarometer Forschungsgruppe Wahlen, Mannheim, 16 April and 17-20 May 1999).

65

One wonders, however, what would have happened had there been a right-of-centre coalition government at the time with either SPD or the Greens, or both in opposition.

66

Hofland, H.J.A. (1999) ‘De NAVO-bombardementen als investeringsruïne’, NRCHandelsblad (Rotterdam), 29 May.

67

Daalder and O’Hanlon (2000).