Populism, Party Ideology, and Economic Expropriations

AbstractWhat is the connection between populism, globalization, and international political economy (IPE) more broadly? Many presume that all populists oppose all forms of globalization, focusing on trade, migration, and international institutions. This article examines the underexplored but important relationship between populism and foreign direct investment, specifically whether populists are more likely than other leaders to expropriate foreign-owned economic assets. The article develops a theory proposing that left populists but not right populists are more likely to expropriate foreign-owned assets, to maintain their domestic political support. The article presents new quantitative panel data on expropriations in all states from 1990 to 2018, finding that left populists but not right populists are significantly more likely than other leaders to expropriate. Further, neither bilateral investment treaties or domestic political institutions, such as democratic constraints, prevent left populists from expropriating. Complementing the quantitative analysis, the article demonstrates these points with an illustrative case study of Bolivia under Evo Morales. The article advances our understanding of the relationship between populism, globalization, and IPE, highlighting the importance of distinguishing between left and right populism and different forms of globalization.¿Cuál es la conexión existente entre el populismo, la globalización y la economía política internacional (IPE, por sus siglas en inglés), en un sentido más amplio? Mucha gente presupone que todos los populistas se oponen a todas las formas de globalización, y se centran en el comercio, la migración y las instituciones internacionales. Este artículo estudia la relación, poco estudiada hasta ahora, pero que resulta de importancia, entre el populismo y la inversión extranjera directa. En concreto, el artículo estudia si los populistas son más propensos que otros líderes a expropiar activos económicos de propiedad extranjera. El artículo desarrolla una teoría que propone que los populistas de izquierdas, pero no los populistas de derechas, cuentan con más probabilidades de expropiar activos de propiedad extranjera, con el fin de mantener su apoyo político interno. El artículo presenta nuevos datos cuantitativos sobre expropiaciones en todos los Estados entre 1990 y 2018. De estos datos, concluimos que los populistas de izquierdas son significativamente más propensos que otros líderes a expropiar, lo que no ocurre con los populistas de derechas. Además, concluimos que ni los tratados bilaterales en materia de inversión ni las instituciones políticas nacionales, tales como las restricciones democráticas, impiden que los populistas de izquierda expropien. Con el fin de complementar este análisis cuantitativo, el artículo demuestra estos puntos con un estudio de caso ilustrativo de Bolivia bajo la presidencia de Evo Morales. El artículo nos permite avanzar con respecto a nuestra comprensión de la relación entre populismo, globalización e IPE, destacando la importancia que tiene distinguir entre populismo de izquierdas y de derechas, así como las diferentes formas de globalización.Quel est le lien entre le populisme, la mondialisation et l’économie politique internationale (EPI) au sens large ? On pense souvent que tous les populistes s’opposent à toutes les formes de mondialisation, en se concentrant plus particulièrement sur le commerce, l’immigration et les institutions internationales. Cet article s’intéresse à la relation sous-étudiée mais importante entre le populisme et les investissements directs à l’étranger et, plus précisément, à si les populistes ont plus de chances que d’autres dirigeants de s’approprier des actifs économiques étrangers. L’article développe une théorie selon laquelle les populistes de gauche, contrairement aux populistes de droite, ont plus de chances de s’approprier des actifs étrangers afin de maintenir leur soutien politique à l’échelle nationale. Il présente de nouvelles données quantitatives de panel sur l’appropriation dans tous les États entre 1990 et 2018, avant de conclure que les populistes de gauche, à la différence des populistes de droite, ont bien plus de chances que les autres dirigeants de s’approprier ces actifs. En outre, ni les traités bilatéraux relatifs aux investissements ni les institutions politiques nationales, comme les contraintes démocratiques, n’empêchent les populistes de gauche d’avoir recours à l’appropriation. En complément de l’analyse quantitative, l’article démontre ces points à l’aide d’une étude de cas de la Bolivie sous la présidence d’Evo Morales. Il enrichit notre compréhension de la relation entre le populisme, la mondialisation et l’EPI, en soulignant l’importance de la distinction entre le populisme de gauche et de droite et des différentes formes de mondialisation.Keywords: Expropriationpopulismforeign direct investment AcknowledgementsWe received very helpful feedback when we presented earlier versions of this article at the Emory International Relations Journal Club, the UC San Diego Political Science Department, and the 2021 annual meeting of the American Political Science Association. We offer special thanks to Eric Reinhardt and Umberto Mignozzetti for their very helpful comments.Disclosure statementNo potential conflict of interest was reported by the author(s).Data availability statementReplication materials for this article are available at https://dataverse.harvard.edu/dataverse/internationalinteractions.Notes1 See also Johns and Wellhausen (Citation2016), Wellhausen (Citation2015), Kerner (Citation2009) and Jensen et al. (Citation2020).2 Hartwell and Devinney use the term “political personalities” broadly, inclusive of the political “orientation of the parties in power,” political preferences.3 Mudde defines populism as “an ideology that considers society to be ultimately separated into two homogeneous and antagonistic groups, ‘the pure people’ versus ‘the corrupt elite,’ and which argues that politics should be an expression of the volonté générale (general will) of the people” (Citation2004, 543).4 On populists and international institutions see Carnegie, Clark, and Kaya (Citation2021), Von Borzyskowski and Vabulas (Citation2019), Söderbaum, Spandler, and Pacciardi (Citation2021). Acemoglu, Egorov, and Sonin (Citation2013) use the label “populism,” but are de facto describing left populist leaders.5 This definition also includes partial expropriations, i.e., expropriations of part of a company’s assets.6 Besides the Kobrin data, there are other expropriation data (Arel-Bundock, Peinhardt, and Pond Citation2020; Graham, Johnston, and Kingsley Citation2018). These data are, however, either limited in scope or only capture the more abstract concept of “expropriation risk.” Another potential source is investor-dispute settlement cases (ISDS). While ISDS records are public , we do not use ISDS cases because they are much less visible and hence do not serve the political signaling function described in our theory.7 The news outlets are: New York Times, Associated Press, The Times (London), Washington Post, The Guardian, and United Press International. We chose these outlets because of their temporal scope and reputation as providing global coverage of political economy news. Further, our theory predicts that left populists expropriate to signal, meaning they will seek broad media coverage of their actions, justifying our use of leading global media outlets.8 The online appendix includes more details on the coding approach.9 The coding rules for our expropriation data correspond to the “overt” events in Kobrin’s dataset (Esberg and Perlman Citation2023).10 See online appendix.11 The data cover not only the ruling party but also the ideology of coalition partners and opposition parties.12 The correlation between DPI and V-Party is relatively strong (−0.62). The correlation is negative because the coding scheme in DPI is the opposite of V-Party, i.e, leftist is the highest value.13 We also include robustness tests of rare events logit and regular logit regression. For the main results in the text, we use the linear probability model because it is easier to interpret.14 For robustness, we also run the main model without the lagged dependent variable; the coefficients remain similar.15 The full regression table is displayed in Supplemental Appendix B.16 The sample mean is 0.007 (see Supplemental Appendix).17 The V-Party Populist measure is an index variable consisting of two variables. The first variable is about the importance of anti-elite rhetoric for a party. The other variable captures people-centrism, i.e., do parties glorify ordinary people and do they see them as part of them? This rhetoric-based measure includes similar elements as the Blair Institute measure. One advantage is that it places parties on a continuous populist scale.18 We use the Kobrin data from 1990 to 2014 based on the extension by Esberg and Perlman (Citation2023). Our findings replicate using the entire Kobrin data set and when we only subset the data set to “overt” expropriation events, which maps most closely to how we code our data.19 Liu, Wang, and Xu (Citation2022) discuss two other types of estimators. Given the lack of evidence supporting their identifying assumptions, we prioritize the matrix completion estimator.20 We use a treatment dummy for left populism that is based on the V-Party score. As a robustness test, we also use the DPI left-right measure to code the left populism dummy. The ATT using the DPI left populism measure is 0.136 (p < 0.05). The Supplemental Appendix includes more detailed results as well as tests validating the no pre-trends assumption.21 Jamal and Milner (Citation2022) discovered that Tunisian subjects with stronger affiliation with Islam were often more likely to oppose FDI.