Desktop version

Home arrow Communication arrow Public Opinion, Transatlantic Relations and the Use of Force

Source

Results of the analysis

We ran three separate kinds of analyses. We first examined an additive model of all relevant variables and the control, run separately for the US and for the five European countries (France, Germany, Italy, the Netherlands and the United Kingdom) together. We then explored the relationship between Atlanticism, political ideology and support for the use of force with different interactive models, in which Atlanticism was first interacted with a dummy variable contrasting the EU with the US, and then a three-way interaction to explore the conditional relationship between Atlanticism and political ideology, in the EU and the US, using clustered standard error. Finally, we ran two separate path analyses for Europe and the US.9 The latter results, although less robust than those based on a structural equation model if data conditions had permitted, are still indicative of the similarities and differences among the different predictors in explaining support for the use of force.

Starting with our additive model, Table 4.2 r eports the results of separate runs for the five European countries (EU-5) and the US, using the three different dependent variables: support for the use of force in different hypothetical conditions, for Iraq and for Afghanistan. The results are quite clear and impressive. As expected, both Atlanticism

Table 4.2 Attitudes toward the use of force (omitting socio-demographics controls and country dummies; odds ratio coefficient in logistic regression; robust Huber-White standard errors in parentheses)

Use of force in generalt*

Iraq**

Afghanistan**

EU-5

US

EU-5

US

EU-5

US

Atlanticism

  • 0.304***
  • (0.0307)
  • 0.315***
  • (0.083)
  • 1.589***
  • (0.078)
  • 1.165
  • (0.143)
  • 1.607***
  • (0.0680)
  • 1.279*
  • (0.158)

Hawk

  • 0.462**
  • (0.151)
  • 0.651*
  • (0.321)
  • 1.351
  • (0.343)
  • 10.41***
  • (5.243)
  • 1.087
  • (0.230)
  • 4.186**
  • (1.935)

Pragmatist

  • 0.545***
  • (0.120)
  • 0.259
  • (0.306)
  • 1.287
  • (0.246)
  • 6.608***
  • (3.163)
  • 1.464*
  • (0.227)
  • 3.312**
  • (1.437)

Dove

  • -0.165
  • (0.120)
  • -0.313
  • (0.353)
  • 0.610*
  • (0.118)
  • 4.674**
  • (2.506)
  • 0.733*
  • (0.110)
  • 1.444
  • (0.706)

Realism Threat

  • 0.299***
  • (0.064)
  • 0.432*
  • (0.185)
  • 0.956
  • (0.093)
  • 2.458***
  • (0.626)
  • 0.886
  • (0.0759)
  • 1.582
  • (0.405)

Global Threat

  • 0.441***
  • (0.063)
  • 0.339*
  • (0.139)
  • 1.122
  • (0.108)
  • 0.994
  • (0.205)
  • 0.989
  • (0.0845)
  • 0.785
  • (0.163)

Feelings US

  • 0.0058***
  • (0.0013)
  • 0.0158***
  • (0.0036)
  • 1.013***
  • (0.0021)
  • 1.036***
  • (0.0066)
  • 1.007***
  • (0.0017)
  • 1.015**
  • (0.0056)

Feelings EU

  • 0.0004
  • (0.0013)
  • -0.0082**
  • (0.0032)
  • 1.000
  • (0.0020)
  • 0.996
  • (0.0050)
  • 1.006***
  • (0.0017)
  • 1.002
  • (0.0053)

Left-Right

  • 0.111***
  • (0.0227)
  • 0.0593
  • (0.0507)
  • 1.222***
  • (0.0457)
  • 1.630***
  • (0.136)
  • 1.057
  • (0.0328)
  • 1.262**
  • (0.0950)

Constant

  • 2.764***
  • (0.264)
  • 3.081***
  • (0.829)

0.0697 (0.0282) ***

0.00004 (0.00009) ***

  • 0.115
  • (0.039)***
  • 0.0045
  • (0.007)**

R2

0.2260

0.2100

0.3104*

0.2417*

0.1010*

0.1512*

Adjusted R2

0.2207

0.1826

Observations

3398

567

3861

707

3734

697

UK and Isolationists as baseline; EU5 includes France, Germany, Italy, Netherlands, United Kingdom. Controlling for gender, age, education and country dummies. t Index ranging from 0 = low to 8 = high support for the use of force.

  • * OLS with White-corrected standard errors (Sandwich estimation). Reported coefficients are unstandardized b.
  • ** Logistic regression with White-corrected standard errors (Sandwich estimation). Reported coefficients are odds ratios. ф Pseudo R2
  • * p < 0.05, ** p < 0.01, *** p < 0.001

Source: GMFUS, Transatlantic Trends Survey, 2004.

and the general orientation toward the use of force are important and statistically significant predictors of support for the use of force, both in the hypothetical scenario and in the two specific circumstances of Afghanistan and Iraq. In Europe, Atlanticism is always highly significant. In the US, it is significant in the hypothetical scenario and in the Afghanistan case, but not in the case of Iraq, where it is insignificant statistically, although the sign takes the Right direction. Next to Atlanticism, there is the impact of the general orientation toward the use of force. Being a Hawk and a Pragmatist is always a highly significant predictor of support for the use of force in the US, but much less so in Europe. Quite interestingly, Doves in the US behave differently, depending on the situations. In Europe, Doves behave as expected, although the coefficient is not statistically significant in the hypothetical scenario. In the US, in contrast, they are more supportive than the baseline American (who, we remember, is an Isolationist). Threat perception is significant only in some conditions and only when respondents are concerned about Realist threats. Threat perception has a clear impact only in the US and in the Iraq war, with those most concerned about Realist threats also more likely to support the use of force there. Pro-American sentiments, and to a lesser extent pro-European sentiments, are also significant predictors of higher support.

Finally, the respondent's position on the Left-Right scale is systematically important in all situations (except for Europeans in the case of Afghanistan) and in the expected direction. All in all, in different combinations, all four dimensions of the Atlantic community are important predictors of support for the use of force, with Atlanticism, and orientation toward the use of force and pro-American sentiments standing out as influential in predicting support for the use of force.

We argued before, and the data confirm, that Europeans and Americans closely resemble one another in the elements that structure their way of thinking about transatlantic relations, but they do differ in the way these common elements combine. In particular, we suggest that Atlanticism and political ideology work differently in Europe and the United States. To explore these interactions we ran a set of pooled analyses, using a two- and a three-ways interaction term between Atlanticism and ideological orientation, comparing the EU and the US. More specifically, we explored first how the impact of Atlanticism on support for the use of force is moderated in Europe and the United States, modeling an interaction term between a dummy that distinguishes the five European countries from the US. Second, we tested a three-way interaction of Atlanticism and political ideology, always contrasting the EU versus the

US. Table 4.2 reports the sets of results for the three dependent variables used throughout this section.10

It is reassuring that the overall results for the other dimensions of Atlantic community do not change using a different, more efficient estimation. Orientation toward the use of force, feelings toward the US and EU, threat perception and political ideology all behave as they did in the previous model, in both the hypothetical situations and in the specific cases of Iraq and Afghanistan. Threat perception, as in the previous analysis, is less relevant, especially when it comes to Iraq and Afghanistan. Coming more specifically to Atlanticism, Model 1 reports the differential impact of Atlanticism on support for the use of force in Europe and in the US. We find that the interaction coefficient is significant in the Iraq and Afghanistan case, while it is not in the hypothetical case. These two real cases are, however, the most interesting ones for our discussion since these are wars in which the very sense of Atlantic community has been called into question. The results confirm the differential impact of Atlanticism in Europe and the US. In Europe, the impact is strongly positive, making those who are more Atlanticist also more willing to support the use of force, while in the US the relationship is weaker (Table 4.3).

To illustrate, we take the cases of Iraq and Afghanistan as appropriate examples, given the greater transatlantic conflict that these two wars provoked. Figures 4.4 and 4.5 report the probability of supporting the use of force among Atlanticists in Europe and the United States, comparing Hawks and Doves, for Iraq and Afghanistan respectively.11 The results clearly show that Atlanticism works quite differently in the US and in Europe in a way that makes attitudes on these two wars strikingly similar once they are seen through the lenses of Atlanticism. Not surprisingly, Hawks are always more likely than Doves to support the use of force in both Iraq and Afghanistan, in Europe as well as the United States. What is remarkable, however, is the differential impact of Atlanticism in Europe and the United States. In Europe, to be an Atlanticist helps to overcome the transatlantic gap on the conditions under which force is seen as acceptable or necessary, even in such a controversial case as Iraq. Surely, Doves are systematically less enthusiastic than Hawks, but even among them the likelihood of supporting the Iraq war almost doubles as we move up along the index of Atlanticism. Nothing similar occurs in the US. In fact, in the US it is not Atlanticism but rather the general orientation toward the use of force that makes all the difference in explaining support.

In Model 2 (Table 4.3), we add a further element, political ideology, and we report a three-way interaction between Atlanticism, ideology

Table 4.3 Attitudes toward the use of force - interaction terms (omitting socio-demographics controls; odds ratio coefficient in logistic regression; robust Huber-White clustered standard errors in parentheses)

Model 1

(Pooled EU and US with Atlanticism interaction)

Model 2

(Pooled with 3-way interaction: Atlanticism, Left-Right, EU-US)

Use of force in general f *

Afghanistan**

Iraq**

Use of force in general f*

Afghanistan**

Iraq**

Atlanticism

  • 0.279**
  • (0.0501)
  • 1.526***
  • (0.0853)
  • 1.524***
  • (0.152)
  • 0.417
  • (0.175)
  • 1.549*
  • (0.265)
  • 1.344
  • (0.206)

Hawks

  • 0.596*
  • (0.215)
  • 1.448
  • (0.441)
  • 2.466***
  • (0.634)
  • 0.590*
  • (0.224)
  • 1.383
  • (0.401)
  • 2.385***
  • (0.589)

Pragmatists

  • 0.468*
  • (0.165)
  • 1.448*
  • (0.241)
  • 1.904***
  • (0.350)
  • 0.474*
  • (0.172)
  • 1.446*
  • (0.234)
  • 1.888***
  • (0.341)

Doves

  • -0.266
  • (0.111)
  • 0.831
  • (0.197)
  • 0.792
  • (0.197)
  • -0.265
  • (0.114)
  • 0.828
  • (0.193)
  • 0.785
  • (0.194)

Realism Threat

  • 0.281**
  • (0.0545)
  • 1.011
  • (0.125)
  • 0.985
  • (0.154)
  • 0.281**
  • (0.0540)
  • 1.011
  • (0.126)
  • 0.989
  • (0.157)

Global Threat

  • 0.483*
  • (0.157)
  • 0.864
  • (0.0655)
  • 0.932
  • (0.213)
  • 0.481*
  • (0.158)
  • 0.874
  • (0.0633)
  • 0.946
  • (0.218)

Feelings US

  • 0.00623*
  • (0.00206)
  • 1.007**
  • (0.00221)
  • 1.017***
  • (0.00319)
  • 0.00610*
  • (0.00204)
  • 1.007**
  • (0.00221)
  • 1.017***
  • (0.00320)

Feelings EU

  • -0.00250
  • (0.00239)
  • 1.007*
  • (0.00280)
  • 0.994
  • (0.00427)
  • -0.00255
  • (0.00249)
  • 1.007**
  • (0.00271)
  • 0.995
  • (0.00427)

Left-Right

  • 0.123*
  • (0.0383)
  • 1.089
  • (0.0612)
  • 1.300***
  • (0.0890)
  • 0.169
  • (0.100)
  • 1.039
  • (0.0807)
  • 1.169*
  • (0.0775)

Dummy EU-US

-0.736*

1.729

2.352*

-0.0159

0.0910***

0.117***

(0 = EU)

(0.197)

(0.657)

(0.889)

(0.497)

(0.0285)

(0.0367)

Atlanticism x

-0.0559

0.695***

0.662**

-0.471*

1.494*

1.544*

Dummy EU-US

(0.0573)

(0.0636)

(0.0877)

(0.166)

(0.270)

(0.338)

Left-Right x

-

-

-

-0.0345

1.000

1.034

Atlanticism

(0.0351)

(0.0416)

(0.0225)

Left-Right x Dummy

-

-

-0.165

1.974***

1.998***

EU-US

(0.122)

(0.142)

(0.0998)

Atlanticism x

0.0975*

0.843***

0.827***

Dummy EU-US x

Left-

Right

(0.0371)

(0.0336)

(0.0233)

Constant

2.956**

2.789**

(0.506)

(0.675)

R2

0.1809

0.0931*

0.1876*

0.1818

0.0965*

0.1903*

Adjusted R2

0.1765

0.1763

Observations

3965

4431

4568

3965

4431

4568

Pooled EU-5 (France, Germany, Italy, Netherlands, United Kingdom) and US. Controlling for gender, age, education; coefficients not reported. Isolationists baseline.

f Index ranging from 0 = low to 8 = high support for the use of force.

OLS with White-corrected standard errors (Sandwich estimation). Reported coefficients are unstandardized b.

Logistic regression with clustered White-corrected standard errors (Sandwich estimation). Reported coefficients are odds ratios.

$ Pseudo R2

* p < 0.05, ** p < 0.01, *** p < 0.001

Source: GMFUS, Transatlantic Trends Survey, 2004.

Probability of support of the Iraq and Afghanistan war by level of Atlanticism and orientation on use of force

Figure 4.4 Probability of support of the Iraq and Afghanistan war by level of Atlanticism and orientation on use of force

Probability of support of the Iraq and Afghanistan war by level of Atlanticism and orientation on use of force

Figure 4.5 Probability of support of the Iraq and Afghanistan war by level of Atlanticism and orientation on use of force

and Europeans versus the US. The results corroborate our expectations. The three-way interaction is always significant, although some effects still remain to be attributed to the differential impact of ideology and Atlanticism in Europe and the US. This testifies to the relevance of the way the different components of the Atlantic belief structure vary depending on the degree of politicization of issues.

To discuss the results in a more effective way, we focus again on the case of Iraq, comparing Europeans and American Hawks, as they vary in ideological position and level of Atlanticism (Figures 4.6 and 4.7). In Europe, Atlanticism always has a positive impact on support for the Iraq war, and its impact grows stronger as we move from Left to Right.

Probability of support of the Iraq war by level of Atlanticism and political ideology

Figure 4.6 Probability of support of the Iraq war by level of Atlanticism and political ideology

Probability of support of the Iraq war by level of Atlanticism and political ideology

Figure 4.7 Probability of support of the Iraq war by level of Atlanticism and political ideology

In the US, the picture is completely different. On the Left, the relationship closely resembles the European pattern, but as we move along the ideological spectrum, the relationship slowly turns negative, with those on the extreme Right most supportive of the war when they are least Atlanticist.

Last, in Figure 4.8, we show a recursive path model of the general support for the use of force using the hypothetical questions. This model confirms several results already discussed, but it also stresses interesting new points which have to do with the relationship between Atlanticism and its antecedents in Europe as well as the US. As one could expect, the relationships between support for the use of force and Atlanticism, sense of we-feelings and Left-Right ideological orientation all work in the same direction in the case of the European countries. Those who are pro-American are also pro-Atlanticists and those on the Right are more strongly both. In the US, the relationship does not hold and it is more complicated.12 Of course, being pro-European means being also pro-Atlanticist, but the ideological makeup of those who are both pro-Atlanticist and pro-European is completely different. In the US, the Liberals are also likely to be pro-European and pro-Atlanticist, while for the Conservatives it works in the opposite direction.

A path analysis of Atlantic community policy coordination on use of force (first coefficient EU, second coefficient US)

Figure 4.8 A path analysis of Atlantic community policy coordination on use of force (first coefficient EU, second coefficient US)

 
Source
Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >

Related topics