How Populistic Were the Populists in 19th Century America?

• In Event: Politics and History & MENA Virtual Poster Session
  In Virtual Poster Session: Politics and History Virtual Poster Session

Abstract

Populism in American Historical Context: "populism" has been applied to many politicians and social movements in US history such as Joseph McCarthy, Huey Long, George Wallace, Ronald Regan, Ross Perot, Bernie Sanders, and Donald Trump.

Originally, Populism of the United States began with the People's Party (i.e., the Populists), which was formed early in the 1890s. From 1892 to 1908, the party nominated candidates to run for the US president. However, after the 1896 election, the People's Party limped along and finally collapsed soon after the 1908 election.

Historians have provided excellent qualitative studies of the People's Party. Although there are numerous quantitative researches on Populism in modern America, scholars have not given serious attention to the People's Party using a well-developed scientific methodological apparatus.

Therefore, in this paper, the author applied the modern analytical tool of Populism to the People's Party and demonstrated how populistic they were.

Multiple approaches toward Populism: there are different approaches to Populism, such as (1) political theoretic, (2) structuralist, (3) economic, (4) strategic/discursive, and (4) ideational approach. Those approaches are not mutually exclusive, and it is essential to choose one from those approaches as a relevant analytic tool. In this paper, the author seeks to demonstrate how (and in what way) populistic the politicians of the People's Party was. Therefore, the ideational approach with automated textual analysis is adopted in this article.

Data and Method: the author has collected historical documents left by the People's Party's presidential candidates. The amounts of the texts are as follows. (1) James B. Weaver (1892: 11,255 sentences, 229,984 tokens); (2) William Jennings Bryan (1896 and 1900: 6,525 sentences, 185,234 tokens); (3) Wharton Baker (1900: 3,756 sentences, 125,120 tokens); (4) Thomas E. Watson (1904 and 1908: 17,440 sentences, 358,163 tokens). The author conducted a computerized content analysis, such as dictionary-based sentiment analysis, cosine similarity calculation, word2vec analysis, and a topic models (Latent Dirichlet Allocation: LDA, Structural Topic Model: STM, Keyword Assisted Topic Model, ketATM).

Result: first, the author conducted a sentiment analysis with the populistic dictionary, which the author created. The result showed that Watson and Weaver scored high in 'people-unity sentiment' and scored low in 'anti-elitism sentiment.' Contrary, Bryan scored lower in 'people-unity sentiment' and scored very high in 'anti-elitism sentiment.' Barker scored low in both sentiments. What is more, the results of the cosine similarity calculation showed considerable overlap between Watson and Weaver's documents. Although Bryan and Watson or Weaver were somewhat similar, Barker was distant from all the other presidential candidates from the People's Party. Concerning topic models, the author found that Watson and Weaver's topics contain words related to people's unity. On the other hand, topics for Bryan contain many anti-elite words.

Conclusion: conventional studies assumed that all the presidential candidates from the People's Party were nearly homogenous. However, the results of this study show that they were different sort of populists. Watson and Weaver stressed more the unity of the people rather than anti-elitism. On the other hand, Bryan emphasized more anti-elitism. Finally, the author suggests room for reconsidering the concept of Populism used in studies of American politics and history.

Author

• Masaru Nishikawa, Tsuda University