Abstract

The term populism has been used extensively in the last two decades; political theorists have spent a lot of effort to define the concept unambiguously. Yet the measurement of populism - particularly over time and space - has received far less attention. This is problematic because the most relevant research questions about populism are empirical and not conceptual in nature. In this paper it is demonstrated that despite its conceptual vagueness, populism is empirically measurable. The paper compares two different methods to measure populism: a classical content analysis and a computer-based content analysis. The methods are applied to election manifestos of political parties in the United Kingdom and the Netherlands in three election periods in each country. Both methods turn out to yield valid and reliable measurements of populism. Based on an assessment of the quality of the methods, recommendations are formulatedfor how to measure populism in future comparative research.

Over the last two decades the term “populism” has increasingly been used in Western Europe – both in academic and vernacular language. The term has been applied to a wide range of different political parties such as the Front National in France, Die Linke in Germany, the British National Party in the United Kingdom and the Lijst Pim Fortuyn in the Netherlands. As a result of this wide-covering application, there exists a lot of disagreement about how populism should be defined. Unfortunately, this lack of conceptual clarity has hitherto prohibited scholars from trying to develop systematic methods to empirically measure populism. This is problematic, because the most relevant populism-related questions are empirical - and not conceptual - in nature. A good example of such an empirical question is whether we are living in a populist Zeitgeist (see Mudde 2004). Have mainstream political parties in Western Europe become more populist over the years? In order to answer this and other empirical questions, we have to pay more attention to the empirical measurement of populism - particularly in comparative settings. Although there have been some empirical investigations into populism, most of these studies are (mostly qualitatively oriented) single case studies (see Meny and Surel 2002a; Albertazzi and McDonnell 2008a; Taggart 2000). A more systematic and comparative perspective is still lacking.

In this paper it is demonstrated that despite its conceptual vagueness, the concept of populism is measurable over time and across cases, in both valid and reliable ways. Moreover, the paper provides the reader with recommendations about how to measure populism in future comparative research. We have compared two different measurements of populism and assessed their validity and reliability. More concretely, we have compared two different content analysis methods with which we have measured the degree of populism of political party manifestos in the United Kingdom and the Netherlands in three different election periods in each country. On the one hand, a classical content analysis was conducted in which extensively trained coders analyzed election manifestos by means of a codebook. On the other hand, a computer-based technique was executed in which a computer carried out the actual analysis. As expected, the computer-based method turns out to be more reliable, while the classical approach yielded more valid results. The more interesting result of our comparison is that both methods turn out to be both sufficiently valid and reliable. This is good news for empirical researchers who want to go beyond conceptual debates and who want to measure populism in comparative settings.

It has often been argued that populism is a vague concept. It deals with “an essential impalpability, an awkward conceptual slipperiness” (Taggart 2000: 1), and points in different contexts to different phenomena (Canovan 1981: 3; Taguieff 1995: 17). Some scholars have therefore proposed to discard the term completely (see Canovan 1981). We do not agree with this vision. The concept of populism is highly relevant, because it helps to understand the common characteristics of phenomena that at first sight seem to be completely different from each other. The concept is an analytical construct that detects commonalities among divergent phenomena that occur in different times and different places. This makes it particularly useful for comparative research. The only way to compare the United Kingdom Independence Party with the Dutch Lijst Pim Fortuyn, is to go beyond specific cases and study these parties simultaneously and by means of comparable standards.

Although scholars do not agree on a definition of populism, and many different definitions circulate simultaneously, there is a common denominator that most scholars seem to share. The populist core, most of them concur, consists of an antagonistic relationship between “the (good) people” and “the (corrupt) elite” (Canovan 1981; 1999; 2002; Meny and Surel 2002b; Mudde 2004; 2007; Laclau 2005; Panizza 2005; Jagers 2006; Jagers and Walgrave 2007; Stanley 2008). Populists mostly position “the people” at the center of politics. At the same time, they are hostile towards (corrupt) political, economical and cultural elites because they stand in the way of the centrality of the people. In this paper we have departed from this scholarly agreement about populism as an antagonistic relationship between the people and the elite. Moreover, inspired by Mudde (2004) and others, we have conceptualized populism neither as a particular form of organization, nor as a political style, but as a “thin ideology”.

In the following section attention is paid to three different perceptions of populism: populism as an organizational form, populism as a style and populism as a thin ideology. Next, two different dimensions of populism are expounded and it is argued why populism should be conceived of as a scale. In the following section the focus is on the content analysis approach as an appropriate way of measuring the thin ideology of populism. Before digging into our own two methods, attention is paid to two recent content analysis studies of populism. Then, drawing on analyses of party manifestos, the classical and computer-based content analysis methods are compared with each other and their validity and reliability are assessed. After this comparison, the advantages and the disadvantages of the proposed methods are evaluated. Finally, guidelines for how to measure populism in future comparative research are provided.

Дата добавления: 2015-07-10