1
Anti-Americanism in North America: Canada and Mexico
Brian Bow, Peter J. Katzenstein, Arturo Santa-Cruz
Paper presented at the 48th Annual International Studies Association Convention,
Hilton Chicago, Chicago, Ill. February 28-March 3, 2007
The 2003 invasion of Iraq changed the way that people all over the world saw,
and felt about, the United States. While President Bush and his policies were strongly
disliked before, worldwide many more respondents held favorable rather than
unfavorable opinions of the United States. In the Pew 2002 poll, for example, pluralities
in 35 of 42 countries expressed favorable views of the United States (Pew Global
Attitudes Project 2002).
1
This situation changed dramatically in 2003 and 2004. The
societies most hostile to the United States, by far, are located in the Islamic Middle East,
North Africa, and Pakistan. Yet Anti-American views have also spread and deepened in
Canada and Mexico. This upsurge in neighboring countries so intimately tied to the
United States and Americans is intriguing. It is one thing for devout Muslims in Pakistan
to dislike intensely the United States and its policies. But why should anti-American
sentiments be shared by so many Mexicans and Canadians? And what precisely do anti-
American sentiments refer to in these two countries?
We preface our discussion with one methodological note. In the analysis of anti-
Americanism, the efficiency of public opinion surveys is both seductive and dangerous.
Polls are about opinions, an important but hardly the only aspect of anti-Americanism.
Critical opinion itself is not a marker of anti-Americanism. Far from it. It may actually
be a marker of a deep kind of pro-Americanism abroad and at home. Only where critical
opinion hardens into skepticism and further into systematic bias or prejudice do we see
the full force of anti-Americanism in action. And only if we understand that anti-
Americanism describes not merely an aggregation of individual opinions but a collective
understanding of a society can we fully grasp its potentially far-reaching political effects.
Furthermore, as a matter of research practice, reducing anti-Americanism to the study of
polls would severely limit systematic analysis to the last few years.
2
This is not to deny
that the kinds of historical evidence we have relied in this paper also has severe
limitations. Historical, qualitative evidence does not offer a particularly welcoming
foundation for the development of a clear deductive argument or the systematic testing of
theories. This paper thus is primarily an exercise in descriptive inference.
1
In the analysis of anti-Americanism opinion is an imperfect measure and must be enriched by an
analysis of systematic bias or prejudice as well as of distrust. These conceptual distinctions are
developed by Katzenstein and Keohane 2007a, 19-24.
2
Only in Europe do time-series data stretch back to the 1950s.
Page 1 of 50 
Convention