The role of military casualties

Before continuing this analysis, we want to look in some more detail into the impact of references in the wording of poll questions referring to the likelihood of military casualties (Table 6.7). For the US, the hypothesis that - if taken separately from other aspects - the incidence of casualties exercises a downward pressure on the size of support seems to find conformation across all issues for which data are available. For the group of EU countries, however, the evidence is contradictory with confirmation only for the case of Kosovo and, rather strong, contrary evidence for the three other cases, where a reference to (the risk of) casualties among one's own forces even tends to increase rather than diminish support.15 Apparently, for Europeans the case of Kosovo was an issue

Table 6.7 The impact of references to military casualties in different cases (in % support, N = 151)

Groups

Issues

Mean support

Std. deviation

diff*

N

All

Kosovo

39,8

18,8

-10

21

Terrorism

63,0

12,4

4

20

Afghanistan

60,0

13,8

11

5

Iraq

47,8

11,3

-1

104

Total

49,1

14,4

-

150

US

Kosovo

36,3

13,4

-16

9

Terrorism

64,8

12,5

-8

16

Afghanistan

59,0

15,7

-17

4

Iraq

48,0

11,2

-9

99

Total

49,6

13,4

-10

128

EU

Kosovo

42,4

22,2

-5

12

Terrorism

56,3

10,7

10

4

Afghanistan

64,0

-

23

1

Iraq

43,8

16,2

7

5

Total

46,2

19,2

6

22

*difference between mean support in the case of military casualties and mean total support for each subgroup concerned.

worth going to war over provided that it would remain a 'war without bloodshed', at least on the part of the NATO forces. Unfortunately, no data are available for the impact of casualties on the support of the use of force for the group of 'Other countries'.

We should remember, however, that Table 6.7 only shows a bivariate relationship. The literature on the impact of casualties is in wide agreement that a bivariate relationship may be misleading because the impact of casualties can or will often be compensated by the impact of other factors. The significance of the present findings will have to wait, therefore, for the more detailed, multivariate analysis to be presented in the next paragraphs.

 
Source
< Prev   CONTENTS   Source   Next >