ABSTRACT
Even as the world’s sole superpower, the United States requires the cooperation of other states to achieve many of its foreign policy objectives. The President of the United States thus often serves as ‘Diplomat in Chief’ in public diplomacy efforts to appeal directly to publics abroad. Given Donald Trump’s antagonistic approach to foreign relations and widespread lack of popularity, what are the implications for support for US policy among publics abroad – particularly among middle power states allied to the US? While previous research on public opinion relying on observational data has found that confidence in the US President is linked to support for American foreign policy goals, the mechanisms at work remain unclear. Using original data from survey-based experiments conducted in Canada and Australia, this article seeks to clarify the effect of ‘presidential framing’ (presenting a policy goal as endorsed or not endorsed by Trump) on attitudes toward key policy issues in the Canada–US and Australia–US relationships. Results point to a negative ‘Trump framing’ effect in Canadians’ and Australians’ trade policy attitudes, but such an effect is not observed in other policy domains (energy policy in Canada, and refugee policy in Australia).
Disclosure Statement
No potential conflict of interest was reported by the author.
Notes on contributors
Timothy B. Gravelle is Lecturer in Political Science and a Member of The Policy Lab at the University of Melbourne. This article was originally presented at the 2017 annual conference of the Australian Society for Quantitative Political Science in Wellington, New Zealand and the 2018 annual conference of the American Association for Public Opinion Research in Denver, Colorado, USA. Thanks are due to the participants of those conferences, and especially to the two anonymous AJIA reviewers for their constructive and prompt reviews. Thanks are also due to Sean Simpson (in Toronto) and Dheepan Shatheeshkumaran (in Melbourne) with Ipsos Public Affairs for their assistance in collecting the survey data used in this article.
ORCID
Timothy B. Gravelle http://orcid.org/0000-0002-7091-206X
Notes
1. Given that the functional form of the ordinal logit model assumes proportional odds (Fox Citation2008), it is worth noting that tests of the proportional odds assumption are statistically insignificant – NAFTA model: χ2 = 0.556, d.f. = 2, p = 0.757; Keystone XL model: χ2 = 1.366, d.f. = 2, p = 0.505. This is to say that proportional odds can be assumed to hold with these data.
2. It is again worth noting that tests of the proportional odds assumption are statistically insignificant – AUSFTA model: χ2 = 3.126, d.f. = 2, p = 0.210; refugee deal model: χ2 = 3.812, d.f. = 2, p = 0.149.