Populism and the economics of globalization

Abstract

Populism may seem like it has come out of nowhere, but it has been on the rise for a while. I argue that economic history and economic theory both provide ample grounds for anticipating that advanced stages of economic globalization would produce a political backlash. While the backlash may have been predictable, the specific form it took was less so. I distinguish between left-wing and right-wing variants of populism, which differ with respect to the societal cleavages that populist politicians highlight. The first has been predominant in Latin America, and the second in Europe. I argue that these different reactions are related to the relative salience of different types of globalization shocks.

Appendix: Sources and Methods on Measuring Populist Party Strength32

The populism database used to construct the figures in this paper is based on the Global Elections Database (GED) and the Constituency-Level Elections Archive (CLEA).³³ We define populist parties loosely as those which pursue an electoral strategy of emphasizing cleavages between an in-group and an out-group. Parties are coded as populist in the dataset if they are labeled as such in the academic or journalistic literature at some point in their history and fit this definition. The full list of populist parties included in the dataset is shown in Table 3 with sources.

Table 3 Countries, parties, and years in the dataset

The data are unique by state, year, district, and ideology. For each state, year, and district, there are three ideologies possible: left-wing populist, right-wing populist, and not populist. The vote totals represent the votes received by populist parties of that ideological representation in that district. Therefore, the total number of valid votes cast in a district for one election is the sum of the votes for right-wing populist parties, the votes for left-wing populist parties, and the votes for nonpopulist parties.

In general, populist parties fell cleanly into either the left-wing or right-wing categories. Only two parties were difficult to place. The Movimento Cinque Stelle in Italy was identified as a right-wing party because of its Euroskepticism and the Civic Radical Union in Argentina was identified as left wing.

When the data provide multiple rounds of voting per election only results from the first round are used. The reasoning here is that many of the small populist parties have the most visibility early in the election cycle. Furthermore, small populist groups are often eliminated from subsequent rounds of voting so voters do not even have them as a choice in later rounds. Only lower house and presidential (where applicable) elections are included in the dataset. CLEA only maintains information on lower house elections, so to ensure comparability between the datasets the upper house elections in GED are not included. Presidential elections are included because of their prominence in Latin America.

For reasons of continuity, the GED dataset is supplemented with the CLEA dataset as infrequently as possible. Thus, if the GED and CLEA have election data for the same election, the GED data are always prioritized.

The figures in the paper show only the countries with at least one populist party in their history. The populist share in countries which have never recorded a populist party is always zero. The vote shares for these charts are calculated as the sum of the number of votes for populist parties divided by the total number of votes.

The dataset only identifies parties, not individual politicians, as populist. This could be an important limitation in Latin America, where populism is more often associated with an individual leader (usually a presidential candidate) rather than a party. We identify Latin American populist leaders with their party. Furthermore, a party is treated as populist if it has been populist at any point in its history. This creates a particular anomaly in the case of Chile. Allende, considered a populist candidate, won the presidency as the candidate of the Socialist Party of Chile. Allende was later deposed by Pinochet: twenty-seven years after Allende’s candidacy Ricardo Lagos also won the presidency as the Socialist Party’s candidate. Despite having represented the same party as Allende, Ricardo Lagos is not often considered a populist. We adjusted the particular case of Chile by coding the Socialist Party of Chile as populist only before Pinochet’s coup.