Do consumers follow the flag? Perceptions of hostility and consumer preferences

Abstract

Do consumers discriminate against foreign products made in countries they deem adversarial? While previous studies have examined how nationalist boycotts influence trade, there is little evidence consumers “follow the flag” more generally. In this study, we employ a conjoint choice survey experiment in the United States and India to assess how individuals’ geopolitical attitudes affect their product preferences. By permitting heterogeneity in perceptions of foreign relations, and examining how these perceptions affect consumer behavior, we reveal one of the micro-level mechanisms at work in the macro-level relationship between trade and conflict. The results show that, when compared to goods made in countries perceived as “neutral” or “friendly,” consumers are 2–6% less likely to select goods made in countries they perceive as being “hostile.” We conclude that, along with organized boycotts, firms, and states, consumers are also partially responsible for the observed correlation between international political relations and trade flows.

¿Los consumidores discriminan los productos extranjeros fabricados en países que consideran adversos? Si bien estudios anteriores han examinado la forma cómo los boicots nacionalistas afectan al comercio, hay pocas pruebas de que, de un modo más general, los consumidores prefieran «lo nacional». En este estudio, empleamos un experimento de encuesta de elección conjunta realizada en los Estados Unidos y en la India para evaluar cómo las actitudes geopolíticas de los individuos afectan a sus preferencias a la hora de elegir un producto. Al permitir la heterogeneidad en las percepciones de las relaciones exteriores, y examinar cómo estas percepciones afectan al comportamiento de los consumidores, desvelamos uno de los mecanismos a nivel micro que actúan en la relación a nivel macro entre el comercio y el conflicto. Los resultados muestran que, en comparación con los productos fabricados en países percibidos como «neutrales» o «amistosos», los consumidores son entre un 2% y un 6% menos propensos a escoger productos fabricados en países que perciben como «hostiles». Concluimos que, además de los boicots organizados, las empresas y los Estados, los consumidores también son parcialmente responsables de la correlación observada entre las relaciones políticas internacionales y los flujos comerciales.

Les consommateurs sont-ils réticents à acheter des produits étrangers provenant de pays qu’ils considèrent comme antagonistes ? Si de précédents travaux ont examiné l’influence de boycotts nationalistes sur le commerce, il est plus difficile de trouver des preuves attestant d’une tendance plus générale au patriotisme de la part des consommateurs. Cet article s’appuie sur une enquête conjointe réalisée aux États-Unis et en Inde, destinée à évaluer dans quelle mesure les postures des individus en matière de géopolitique affectent leurs choix de consommation. Grâce à la pluralité des perceptions des relations internationales reflétée et à l’analyse de l’impact de ces perceptions sur les comportements des consommateurs, nous révélons l’un des mécanismes à l’oeuvre, au niveau micro, dans la relation qui s’observe au niveau macro entre commerce et conflits. Les résultats de nos recherches démontrent que les consommateurs sont 2 à 6 % moins enclins à acheter des produits provenant de pays perçus comme « hostiles » que des produits issus de pays considérés comme « neutres » ou « amis ». Nous en concluons que, tout comme les entreprises, les États et les boycotts organisés, les consommateurs sont partiellement responsables des corrélations observées entre relations politiques internationales et flux commerciaux.

Keywords:

• International trade
• consumer preferences
• commercial peace

Notes

1 See the “consumer animosity” literature, for example (Klein, Ettenson, and Morris 1998; Riefler and Diamantopoulos 2007).

2 See Hainmueller, Hopkins and Yamamoto 2014 for a more detailed explanation of the conjoint method.

3 Marketing research suggests that “country-of-origin” (COO) has the largest effect on more expensive products (Verlegh and Steenkamp 1999). As such, the use of a relatively cheap product in our US sample should prove a hard test for our hypothesis.

4 In the US sample, we initially contacted 2500 respondents on December 2nd 2019. The second wave, fielded on December 6th, yielded 1233 respondents. We had an attrition rate of 48%. We report in the Supplementary Appendix that the respondents that took the second wave do not differ substantively than the respondents that took the first wave. We fielded the Indian survey on December 2, 2019. Note that the surveys predate any reporting of the Covid19 pandemic.

5 This procedure is motivated by an established method. Conjoint survey experiments have been used by many scholars to examine how the attributes of a candidate impact the choice of voters or party leaders (e.g. Doherty, Dowling, and Miller 2019). In such studies, respondents are asked to identify a candidate they would support. Instead of explicitly including the candidates’ race and gender in the survey, they include the candidates’ names. However, when it comes time to perform the statistical analysis, names are replaced with their corresponding race and gender. We are taking a similar approach here, substituting respondents’ perceptions of different countries of origin for the actual countries themselves. This is a novel innovation, but it is not without justification.

6 Also, keep in mind that since we present each respondent with three profiles, interpretations of the substantive effects appreciate that random selection implies a 33% probability a profile is chosen, rather than the 50% probability traditionally assumed when there are two profiles.

7 We also ran a television experiment on our US sample. However, a coding error prevented the experiment from generating reliable data.