Comparative Political Studies

http://cps.sagepub.com/

Is Chvez Populist? : Measuring Populist Discourse in Comparative

Perspective
Kirk A. Hawkins
Comparative Political Studies 2009 42: 1040 originally published online 20
February 2009
DOI: 10.1177/0010414009331721

The online version of this article can be found at:

http://cps.sagepub.com/content/42/8/1040

Published by:

http://www.sagepublications.com

Additional services and information for Comparative Political Studies can be found at:

Email Alerts: http://cps.sagepub.com/cgi/alerts

Subscriptions: http://cps.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Citations: http://cps.sagepub.com/content/42/8/1040.refs.html

>> Version of Record - Jun 29, 2009

OnlineFirst Version of Record - Feb 20, 2009

What is This?

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

 Comparative Political Studies
Volume 42 Number 8
August 2009 1040-1067
2009 Sage Publications

Is Chvez Populist? 10.1177/0010414009331721

http://cps.sagepub.com
hosted at
Measuring Populist Discourse http://online.sagepub.com

in Comparative Perspective
Kirk A. Hawkins
Brigham Young University, Provo, Utah

This article pushes forward our understanding of populism by developing

one of the more underappreciated definitions of populism, populism as dis-
course. It does so by creating a quantitative measure of populist discourse
suitable for cross-country and historical analysis. The article starts by laying
out the discursive definition of populism in the context of existing defini-
tions. It then operationalizes this definition through a holistic grading of
speeches by current chief executives and a few historical figures. The result
is a data set of elite-level populist discourse in more than 40 current and past
governments from a variety of countries across the world, with special focus
on Latin America. This measurement has high reliability comparable to
standard human-coded content analysis, compares well to common under-
standings of actual cases of populism, and is a reasonably efficient technique
even in small samples.

Keywords:populism; discourse; textual analysis; holistic grading;

measurement

Let no one forget that we are confronting the Devil himself. Sunday, 3
December at the ballot box we will confront the imperialist government of
the United States of North America [sic]that is our real adversary, not these
has-beens here, these lackeys of imperialism. . . .
Long live Christ, the first great revolutionary of our time! Martyr of the
peoples, Christ Redeemer, today is his day, the day of Christ the King. . . .
You the people are the giant that awoke, I your humble soldier will only
do what you say. I am at your orders to continue clearing the way to the
greater Fatherland. . . . Because you are not going to reelect Chvez really,

Authors Note: This research was made possible by grants from the BYU College of Family,
Home, and Social Sciences and the Department of Political Science. Special thanks to
Mayavel Amado, Lilia Cruz, and Emily Ekins for their work on this project, and to Jay
Goodliffe, Kurt Weyland, Guillermo Rosas, Dan Nielson, and Carlos de la Torre for their
insightful comments. The conclusions of this article are the responsibility of the author.
1040

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

 Hawkins / Is Chvez Populist? 1041

you are going to reelect yourselves, the people will reelect the people.
Chvez is nothing but an instrument of the people.

Hugo Chvez

When we read the previous quotes from Hugo Chvezs closing cam-
paign speech of December 2006, we may find ourselves agreeing with schol-
ars and journalists who depict him as a populist (The Americas, 2006;
Castaeda, 2006; Roberts, 2006; Weyland, 2003). While no one word really
stands out, except perhaps the frequent reference to the people, these
quotes capture a set of ideas that seem vaguely democratic but violently
opposed to assumptions of pluralist democracy. What is it about these
words that makes them sound populist? Do other politicians that we tradi-
tionally regard as populist sound similar?
These questions get at the heart of an ongoing academic debate over the
definition and nature of populism. Despite the concepts continuing preva-
lence, scholars, journalists, and much of the public are still unsure of what
the word means. Traditional social science definitions of populism focus on
long-term processes of modernization and industrialization or on the macr-
oeconomic policies of particular governments. More recent definitions
focus on populism as a political concept that refers to strategies and insti-
tutions (Roberts, 2006; Weyland, 2001). Still others emphasize discourse
and ideas, which are touched on in the aforementioned quotes from Chvez
(Canovan, 1999; Laclau, 2005; Mudde, 2004).
This continuing conceptual confusion is paralleled by a lack of empirical
rigor. Much as with older accounts of populism, newer ones tend to declare
certain leaders populist by fiat rather than through any kind of systematic
measurement, and analyses that do offer justifications are usually single-
country studies that avoid demonstrating the broad applicability or reliability
of their measure. Scholars have recently begun measuring populism using
quantitative techniques, especially textual analysis (Armony & Armony,
2005; Jagers & Walgrave, 2007). These are exciting advances, but they are
the exception, and the scope of their analyses is still limited across time and
space. Hence, none of the current conceptualizations of populism have been
subjected to any large-scale exercise in quantitative measurement.
This article pushes forward our understanding of populism by develop-
ing one of the more underappreciated definitions of populism: populism as
discourse. It does so by creating a quantitative measure of discourse that is
suitable for cross-country and historical analysis. The article starts by lay-
ing out the discursive definition of populism in the context of existing

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

1042 Comparative Political Studies

approaches. It then operationalizes the discursive definition through a novel

holistic grading of speeches by current chief executives and a few historical
figures. The result is a data set of elite-level populist discourse in more than
40 current and past governments from a variety of countries across the
world, with special focus on Latin America. This measurement has high
reliability comparable to standard human-coded content analysis, compares
well to common understandings of actual cases of populism, and is a
reasonably efficient technique even in small samples.

Defining Populism as Discourse

There are four principal definitions of populism used in the social sci-
ences today: structural, economic, political-institutional, and discursive. The
first three of these are common in the study of populism in the developing
world, particularly Latin America, and have already been extensively cri-
tiqued by other scholars (Roberts, 1995; Weyland, 2001). The structuralist
approach to populism emphasizes its social origins and associates it with
certain stages of development, especially the attempt at industrialization in
countries located at the periphery of the world economy. According to this
view, populist regimes are those using cross-class coalitions and popular
mobilization to support import-substituting industrialization (ISI) (Cardoso &
Faletto, 1979; Di Tella, 1965, 1997; Germani, 1978; Ianni, 1975; Weffort,
1973). The economic approach to populism identifies it with policy
o utputsspecifically, shortsighted economic policies that appeal to the poor
(Dornbusch & Edwards, 1991). And the political approach sees populism as a
phenomena rooted in the basic struggle over control of government, policy,
and core values of the community. This latter approachwhich is also the
most current (see Roberts, 2003; Weyland, 2001)focuses on institutional
or material aspects of populism such as the degree of institutionalization of
the organization embodying the populism, its low esteem for existing insti-
tutions of representative democracy, its emphasis on support from large
numbers of voters, and the presence of a charismatic leader.
The fourth definition, and the focus of this article, is the discursive one.
It sees populism as a Manichaean discourse that identifies Good with a uni-
fied will of the people and Evil with a conspiring elite. This definition is
more common to the study of populism in Western Europe and the United
States but is largely unknown to mainstream political science because of its
association with antipositivist currents within postmodernism.

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

 Hawkins / Is Chvez Populist? 1043

The first thing to note about the discursive definition is that it describes
something innately cultural. Culture is used here in the Geertzian sense, as
something rooted in our shared ability to assign meanings to the world
around us (Eckstein, 1996). Scholars who define populism discursively use
a variety of labelsreferring to it as a political style (Knight, 1998), a
discourse (de la Torre, 2000; Laclau, 2005), a language (Kazin, 1998),
an appeal (Canovan, 1999), or a thin ideology (Mudde, 2004)but all
of them see it as a set of ideas rather than as a set of actions isolated from
their underlying meanings for leaders and participants.
What are these ideas that constitute populist discourse? Scholars who study
populism have sometimes disparaged the lack of ideological precision in
populist movements, and these concerns are not without grounds. But analyses
of populist discourse all highlight a series of common, rough elements of
linguistic form and content that distinguish populism from other political dis-
courses. I explain these here by drawing on some of Chvezs words, although
I emphasize that these features are not unique to Chvezs rhetoric.
First, populism is a Manichaean discourse because it assigns a moral dimen-
sion to everything, no matter how technical, and interprets it as part of a cosmic
struggle between good and evil (de la Torre, 2000). History is not just proceed-
ing toward some final conflict but has already arrived, and there can be no fence
sitters in this struggle. Hence, in the previous quotes we find Chvez referring
to the election as a contest between the forces of good and evil. This is no
ordinary contest; the opposition represents the Devil himself while the forces
allied with the Bolivarian cause are identified with Christ. Later in the speech,
Chvez frames the election as a stark choice. What is at stake is not simply
whether Chvez remains in power during the next presidential term but whether
Venezuela becomes a truly strong and free country, independent and prosper-
ous or instead a country reduced once more to slavery and darkness.
Within this dualistic vision, the good has a particular identity: It is the will
of the people. The populist notion of the popular will is essentially a crude
version of Rousseaus General Will. The mass of individual citizens are the
rightful sovereign; given enough time for reasoned discourse, they will come
to a knowledge of their collective interest, and the government must be con-
structed in such a way that it can embody their will. Chvez, for example,
refers to his listeners as el pueblo in the singular and talks about them as the
giant that awoke, and later in the same speech he proclaims to dedicate every
hour, every day of his life to the question of how to give more power to the
poor, how to give more power to the people. The populist notion of the
General Will ascribes particular virtue to the views and collective traditions of
common, ordinary folk, who are seen as the overwhelming majority (Wiles,
1969). The voice of the people is the voice of GodVox populi, vox dei.

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

1044 Comparative Political Studies

On the other side of this Manichaean struggle is a conspiring elite that has
subverted the will of the people. As Hofstadter (1966) eloquently describes
in a classic essay on the paranoid mentality in American politics, for
populists this enemy is clearly delineated: he is a perfect model of malice,
a kind of amoral superman: sinister, ubiquitous, powerful, cruel, sensual,
luxury-loving (pp. 31-32). Populism has a preoccupation with discovering
and identifying this enemy, as this process is what helps negatively consti-
tute the people. Thus, Chvez not only demonizes his opposition but associ-
ates them with sinister conspiracies by international forces led by the United
States. He reminds his listeners of who the real opponent is and that the
opposition leaders are lackeys of imperialism. Only those who reject and
fight against this enemy can be part of the peoples crusade.
This set of discursive premises has two important corollaries. First, at least
in the early stages of a populist movement, the subversion of the peoples will
means that some form of liberation or revolution is required, or what Laclau
(2005) terms a rupture. The old system has been taken over by the forces of
evil and no longer serves the people. This conflict is not over particular policies
or issues but institutions and the system. These must be remade or at least sub-
stantially modified; if not, the forces of evil will regroup and continue their
oppression. References to revolution suffuse the language of Chvez, of course
(we are told that Christ himself is a revolutionary), and in this speech he cele-
brates the institutional changes that have been made over the previous 8 years
in Venezuela. Yet the revolution is ongoing, with new stages always on the
horizon. The first era is ending with this electoral cycle, Chvez declares later
in his speech, and another era will begin, another revolutionary era.
The second corollary of populist discourse is what McGuire (1997), in
his description of the Peronist parties in Argentina, calls an anything goes
attitude. Procedural rights associated with liberal democracy, particularly
minority rights, are seen by populists as instrumental and may be violated
in order to better express the will of the people. The evil minority ceases to
have legitimacy, citizenship, or possibly human rights because it has chosen
to fight against the common good; any respect accorded the opposition is
a generous gift rather than a moral imperative, and the populist is unlikely
to show them the kinds of courtesy that one gives a worthy opponent. In this
particular speech, Chvez repeatedly questions the opposition leaders
patriotism and calls them traitors, implying that they are not true Venezuelans.
The other candidates are pipsqueaks (frijolitos) who are not just incom-
petent but irresponsible, liars, un-patriots, without any sense of honor or
responsibility. Chvez never sees his own use of government funds for the
campaign as a questionable activity; instead, he asserts the governments strict

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

 Hawkins / Is Chvez Populist? 1045

adherence to the rules of the game and insinuates that it is the opposition
that is plotting to use fraud.
Admittedly, this set of ideas lacks the precision of classic ideologies such
as socialism or liberalism. It is because of this that I follow the convention
of much of this literature and use the problematic term discourse in its
largely postmodernist sense, as something that combines elements of both
ideology and rhetoric. Populist discourse is like an ideology in that it is a
set of fundamental beliefs about how the world works and tends to compel
its believers into political action (for a discussion of the concept of ideol-
ogy, see Gerring, 1997; Knight, 2006). But unlike an ideology, populism is
a latent set of ideas or a worldview that lacks significant exposition and
contrast with other discourses and is usually low on policy specifics. It
has a subconscious quality that manifests itself primarily in the language of
those who hold it. Hence, populists like Chvez are usually unaware that
they see the political world differently from other people, and even if they
are aware, they struggle to articulate those differences. A discourse is also
like a rhetoric in that it is manifested in distinct linguistic forms and content
that have real political consequences. But unlike common understandings
of rhetoric, the ideas that constitute populist discourse tend to be sin-
cerely held and embodied in the language of their proponents. In the case
of populism in particular, the language is too inflammatory and too suf-
fused with a radical notion of democracy for most people to be able to or
even want to consistently pretend it; to use populist rhetoric almost cer-
tainly means that we believe in what it represents. Thus, we cannot talk as
Chvez does about confronting the Devil himself and our real adver-
sary without accepting a dualistic, teleological, conspiratorial vision of
politics, nor can we identify Venezuelans who support the Bolivarian move-
ment as the people . . . the giant that awoke without simultaneously
believing that there can be a knowable, common good that overrides our
particular interests and perspectives.
Populist discourse is likely part of a larger typology of core political dis-
courses or worldviews that includes elitism and pluralism, although this
typology has not been fully articulated in the scholarly literature. The typol-
ogy includes roughly two dimensions: (a) whether the discourse is merely a
pragmatic approach to the world or a redemptive one and (b) whether it
accepts or rejects fundamental democratic assumptions about the right and
the ability of citizens to rule (Canovan, 1999; Mudde, 2004). When juxta-
posed with elitism (which rejects the right and the ability of citizens to rule),
populism tends to win the admiration of scholars and activists who favor
democracy. This is probably one of the reasons why radical leftist critics of

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

1046 Comparative Political Studies

liberal democracy frequently become defenders of populist regimes such as

those of Chvez; they see populism as a truly democratic response to the
inequalities of capitalist democracy in developing countries. In contrast,
when populism (a redemptive discourse) is juxtaposed with pluralism (a
pragmatic one), it makes us feel much more ambivalent. It clearly has demo-
cratic aspectsit reaffirms popular sovereigntyand typically emerges as a
response to pluralist regimes that have grown ineffective and corrupt. Yet as
Riker (1982) and others note, its disdain for tolerance and dissent makes it
disregard the procedural norms and minority rights that protect us from the
tyranny of the majority, and it can slip into totalitarianism when coupled with
charismatic leaders who claim to embody the will of the people. Because the
distinction between populist and pluralist discourse is the one that the litera-
ture has worked out the most thoroughly, it is the one I use to operationalize
the discursive definition of populism in the following.1

Critiquing the Discursive Definition

This article cannot provide a thorough logical defense of the discursive
definition with regard to other, better known definitions of populism. Readers
interested in this argument should review the work of Roxborough (1984),
Knight (1998), and Laclau (2005) among others, all of which argue that the
concept of discourse provides a minimal definition that potentially reveals
the underlying logic of populism. Instead, the purpose of this article is to
provide a more empirical or scientific defense by showing that populist
discourse is measureable in ways that are valid and reliable.
Nevertheless, we need to acknowledge that a discursive approach to pop-
ulism is problematic for many social scientists in the positivist tradition. First,
there are epistemological concerns. Can populism or other aspects of political
culture really be measured in ways that are valid, reliable, and efficient, with
results that can be replicated and verified by other scholars (Barry, 1978;
Lichbach & Zuckerman, 1997; Little, 1991)? In other words, can a discursive
concept be subjected to the rules of valid scientific inferences (King,
Keohane, & Verba, 1994, p. 38)? Second, there are ontological concerns. Is
populism really populism if it is spoken but never followed? Doesnt it require
action in order to be important or realsome manifestation in a movement,
party, or candidate that pursues power and seeks to implement policies? Finally,
there is the question of real-world importance. Even if we accept discourse as
a defining attribute of populism, does it matter for politics? Do different dis-
courses actually have consequences for policy outputs and elections, namely,
for the material and moral well-being of the community?

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

 Hawkins / Is Chvez Populist? 1047

Regarding the first epistemological issue, one of the purposes of the rest
of this article is of course to demonstrate that populist discourse can be
measured in a way that satisfies scientific criteria. I think that readers will
find this demonstration a compelling one. The answer to the second ques-
tion is a little more evasive. In offering discourse as the defining attribute
of populism, I am not claiming that manifestations of populism can exist with-
out some material component. A discourse is meaningless unless believed and
shared by actual human beings. However, the important point made by
advocates of the discursive definition is that actions aloneraising the
minimum wage, calling for a constitutional convention, repressing the
oppositionare insufficient conditions for populism. Actions are ulti-
mately populist because of the meaning that is ascribed to them by their
participants, not because of any objective quality that inheres in them.
The third questionwhether a populist discourse in the end really mat-
ters for actual politics or whether all politicians and voters respond to similar
sets of preferences rooted in, say, material self-interestis of course one of
the grand theoretical questions that we struggle with as political scientists.
The short answer I provide here is that this is partly an empirical problem.
If we can measure populist discourse and calculate its correlation with
aspects of politics and economics that interest us, then we have shown that
it matters. The question then becomes the more theoretically enriching one
of how or why it matters, and especially how the normative dimensions of
populist discourse fit into the more familiar assumptions of material self-
interest that are common to rational-choice theory. I insist, however, that we
can pursue these questions without abandoning positivist methodology.

Measuring Populist Discourse

For positivists, creating a new definition is not a very impressive feat; the
real test of any concept is our ability to measure it. From this perspective, the
problem with all definitions of populism is that they are either not applied
toward measurement or they are measured in highly imprecise ways that lack
standard tests of reliability and validity or descriptions of how the measurement
took place. Those studies that do offer justifications are usually single-country
ones that avoid demonstrating the broad applicability or reliability of their
measure. Of course, discourse analysts have been particularly reluctant to apply
their concepts to any kind of extensive quantitative measurement. Those
who are more empirically oriented (see de la Torre, 2000; Panizza, 2005) limit

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

1048 Comparative Political Studies

themselves to qualitative case studies or comparisons of just a few leaders, usu-

ally from the same country. Yet proponents of nondiscursive approaches to
populism have been equally reluctant to provide tools for reliably measuring
their own concepts. Many of them also rely on qualitative case studies, and
where they do measure populism in cross-national analyses, as in those of
radical right populism in Western Europe, the label of populist is often applied
without any systematic empirical justification (see Betz, 1994; Taggart, 1996;
for partial exceptions, see Kitschelt & McGann, 1995; Mudde, 2007).
Two recent studies attempt to break from this mold by offering quantita-
tive measurements of populist discourse. Armony and Armony (2005) use a
computer-based technique to measure populist discourse in a large number
of speeches by two Argentine presidents. And Jagers and Walgrave (2007)
perform a human-coded content analysis of television programs by six
Belgian parties. Both studies find significant differences in the discourses of
these leaders and parties that confirm common scholarly depictions. Yet
while these reaffirm the scientific validity of the discursive approach to pop-
ulism, they are still bounded in scope and have natural methodological
limitations I refer to in the following. Our challenge remains finding a way
of measuring the level of populism in the discourse of actual people in mul-
tiple settings, namely, across countries and across time. Doing so will not
only give the discursive definition an added claim to scientific validity but
will allow us to compare leaders and movements of current interest and to
expand our study of the causes and consequences of populism.
In the remainder of this article, I measure populist discourse at the elite
level using a form of content analysis from educational psychology known as
holistic grading, which I apply to a study of more than 200 speeches from 40
chief executives. I measure elite discourse because populism is so often asso-
ciated with the leaders who create and galvanize the movementwe first
want to know how populist Chvez is, and only later do we ask how move-
ment activists and ordinary voters respond to this discourse. Thus, I leave the
measurement of mass discourse for a future exercise, although I refer to this
challenge again in the conclusion.2 I use textual analysis of speeches rather
than a traditional survey technique (such as in the 2006 AmericasBarometer;
see Seligson, 2007) primarily because of accessibility. It is almost impossible
to survey chief executives while they are in office, let alone as large a set of
chief executives as I consider here; in contrast, speeches are widely available.
Textual analysis also tends to respect the culturalist origins of the concept of
populist discourse. Traditional discourse analysts often object to studying
what they consider a highly intersubjective concept using individually

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

 Hawkins / Is Chvez Populist? 1049

subjective measures such as opinion surveys. A study of speeches or similar

texts sidesteps this problem by considering long statements of ideas that
were, potentially at least, widely communicated in a real political setting.
Holistic grading, unlike standard techniques of content analysis (either human
coded or computer based), asks readers to interpret whole texts rather than
count content at the level of words or sentences. Holistic grading is a peda-
gogical assessment technique that is widely used by teachers of writing and
has been extensively developed by administrators of large-scale exams, prin-
cipally Educational Testing Services and the Advanced Placement exams they
administer for the College Board in the United States (White, 1985). Unlike
analytical grading, which tries to break a text down into its parts and then
combine the scores of each of those parts (as a content analysis does), a holis-
tic approach works by assessing the overall qualities of a text and then assign-
ing a single grade without any intervening calculations. The first step is to
design a rubric, or a simplified guide for evaluating a text that identifies the
rough qualities associated with different types or grades. The second step is to
train a set of graders in the use of the rubric, using not only the rubric itself but
also a set of sample or anchor texts that exemplify each type or score
described in the rubric. This combination of rubric and anchor texts is the
hallmark of holistic grading. Finally, the actual grading is conducted using two
to three graders per text, with tests of intercoder reliability calculated along the
way. Analyses of student writing that use holistic grading have been found to
have high levels of intercoder reliability, with correlations typically between
r = .70 and .80, comparable to subjective types of human-coded content
analysis. The costs are often not much greater than that of machine-coded
objective exams, especially when the costs of preparing the exams are taken
into account (Britton, Martin, & Rosen, 1966; Coffman, 1971; Cooper, 1977;
White, 1985).3 Perhaps more important for our purposes, research shows that
small numbers of graders (two to three) and texts (three to four per student
being assessed) have high reliability that improves only marginally with addi-
tional graders and texts (Sudweeks, Reeve, & Bradshaw, 2005).
There are two reasons for using holistic grading to measure populist
discourse. First, we cannot gauge a broad, latent set of meanings in a text
a discoursesimply by counting words. Because the ideas that constitute
the content of the discourse are held subconsciously and conveyed as much
by the tone and style of the language as the actual words, there is no single
word or phrase distinct to populist discourse or a particular location in the
text where we can go to find the speakers statement on the issue, as we
could using party manifestos to measure political ideology (see Budge,

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

1050 Comparative Political Studies

Bara, Volkens, & Klingemann, 2001; Wst & Volkens, 2003).4 This means
that the text must be interpreted by human coders who can quickly analyze
broader, more complex patterns of meaning. Newer computer-based tech-
niques of content analysis offer to solve this problem by generating word
distributions whose broad patterns reveal something about a text (Quinn,
Monroe, Colaresi, Crespin, & Radev, 2006), but in practice these require
considerable interpretation of the resulting distributions. Holistic grading
makes this interpretation more transparent. Second, while it is possible to
use human-coded content analysis at the level of phrases or sections of text,
these techniques are extremely time-consuming and unsuitable for the kind
of cross-country analysis we need to generate large-N comparisons. In con-
trast, holistic grading requires no special preparation of the physical text
and proceeds fairly quickly once the texts are available, and it allows us to
compare texts in multiple languages without any translation so long as cod-
ers speak a common second language that they can use in their training and
in reporting their results.

The Analysis
My assistants and I began by devising a rubric that captures the core ele-
ments of populist discourse. We did so primarily by drawing on the literature
on populist discourse, but also by reading the speeches of several Latin
American politicians who seemed to be widely regarded as populist and com-
paring these with speeches of leaders who were considered more pluralist.
These included speeches by Chvez and several other Latin American leaders.
A copy of the rubric is found in the appendix; it essentially juxtaposes the ele-
ments of populist discourse noted earlier with their pluralist counterparts.
I next recruited and trained a set of native speakers of the languages of
each country. All of these were undergraduate students at my university,
many without political science background. The training familiarized the
students with the discursive definition of populism and the use of the rubric,
including an analysis of anchor speeches that exemplified different catego-
ries of populist discourse.5 I then had the students perform the actual coding,
which they did by reading each speech, taking notes for each of the elements
of populist speech in the rubric (as a check on their work and also as a way
to find relevant quotes), and assigning an overall grade. For the sake of speed
and to use a more holistic approach, which requires having an anchor text for
each potential grade, I had readers use a simple 3-point scale of 0 (nonpopu-
list or pluralist), 1 (mixed), or 2 (populist).

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

 Hawkins / Is Chvez Populist? 1051

The research proceeded in two phases. In the first, we analyzed the speeches
of 19 current presidents (as of fall 2005) of Latin American countries and 5
historical chief executives from the region. In a few cases where changes in
power took place during the study (e.g., the election of Morales in Bolivia),
we considered both chief executives. In this and the subsequent phase of the
analysis, my assistants and I considered four speeches selected quasi-randomly
from four categories: a campaign speech, a ribbon-cutting speech, an interna-
tional speech, and a famous speech, typically an inaugural address or an annual
report to the nation. (For comparison, I later analyze a random selection of
speeches for two of these leaders.) The specific criteria for the four categories
are found on the authors Web site, but the general rule was to select the most
recent available speech within each category that met certain standards of
length (1,000 to 3,000 words). Our purpose in using these particular four cat-
egories was to test the consistency of the discourse while ensuring that we had
not overlooked key classes of speeches. We expected that the famous and
campaign speeches would have a stronger populist discourse than the ribbon-
cutting or international ones because they represented contexts where there
were larger audiences and an appeal to the nation as a whole. The readers
researched and selected the speeches themselves, most of which were availa-
ble on government Web sites, and I reviewed and approved the final selections.
Only two graders were used for all speeches, and each speech was read by the
same two graders, both of whom spent no more than 30 to 45 minutes per
speech. As will be seen later, we were able to significantly reduce this time in
subsequent analyses.
In the second phase, I considered an additional 15 countries outside of
Latin America. These countries were drawn from several regions, including
Western and Eastern Europe, North America, Asia, and Africa, and again I
considered only the current chief executive as of approximately March 2006.
This phase was more challenging because it required training another 35
readers, 2 or 3 from each country. For a better check of intercoder reliability
and to avoid problems of attrition, I tried to recruit 3 readers per country,
although in some cases (noted in Table 2) I ended up with only 2. The
analysis used the same four speech categories, selection criteria, and coding
procedure as the Latin American study. All graders were again native speak-
ers of the language and read the speeches in their original language. Readers
located the speeches themselves (again, usually from government Web
sites), but in this case most of the final selections were approved by two
student assistants. Grading again took 30 to 45 minutes per speech.

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

1052 Comparative Political Studies

Reliability of the Technique

The first question is whether this technique generated data that were
consistent across graders, or reliable. Because the analysis proceeded in
two parts with very different sets of graders, I calculate several measures of
covariation and agreement.
In the first phase of the analysis (the Latin American study), which
included a total of 85 speeches each coded by the same two graders,6 the
level of reliability was high and gave us great confidence in the method.
The Pearson correlation coefficient between the two sets of scores is r = .79
for all individual speeches and r = .87 for the average scores for each
president (N = 24). Alternatively, because the ordinality of the grading scale
is rather rough, I calculated a Spearmans rho of 0.70 for the individual
speech scores, which allows us to reject the null hypothesis of no relation-
ship between the two coders at the p < .000 level. The content analysis lit-
erature generally regards these levels of covariance as high (Neuendorf,
2002). The level of agreement, which tells us if the coders actually had the
same scores rather than if they simply moved in the same direction (Tinsley
& Weiss, 1975), was also quite strong. Within the data from the first phase
of the analysis there was a raw 78% agreement between our two graders
(that is, 78% of the time they assigned exactly the same grade); if we cal-
culate agreement as any time in which the graders are within one grade of
each other, then there was 100% agreement. To account for the possibility
of agreement due to chance, I also calculated Cohens kappa, a statistic that
adjusts the percentage agreement by taking into account the size of the
original scale and the actual observed agreement; the resulting scale ranges
from 0 to 1. The kappa statistic for this first phase of the analysis was .68,
a level generally regarded as substantial (StataCorp, 2003).
The level of reliability in the second phase of the analysis was not as
high but was still encouraging, especially given the large number of
graders, their lack of experience in political science, and the small
number of speeches they each had the chance to read. We cannot calcu-
late covariance figures for these data because I used a different set of
graders for each country and often had three graders rather than two.
However, we can calculate the level of agreement. If we consider agree-
ment to be when all three readers give exactly the same grade, then we
had only 70% agreement in this second phase. If we instead consider
agreement to be when two readers give the same grade and a third
reader differs by no more than one point, then we had 86% agreement.

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

 Hawkins / Is Chvez Populist? 1053

Table 1
Average Populism Scores for Latin
American Chief Executives
Chief Average Standard Number of
Country Executive Populism Score Deviation Speeches

Venezuela Chvez 1.9 .25 4

Ecuador Velasco Ibarra 1.7 .58 3
Bolivia Morales 1.6 .71 4
Argentina Pern 1.5 .71 4
Brazil Vargas (I) 1.0 .54 4
Brazil Vargas (II) 0.9 .57 4
Argentina Menem 0.8 .50 4
El Salvador Saca 0.6 .25 4
Mexico L. Crdenas 0.6 .43 4
Paraguay Duarte 0.5 .50 3
Ecuador Palacio 0.4 .14 3
Peru Toledo 0.3 .29 3
Dominican Fernndez 0.3 .50 4
Republic
Mexico Fox 0.3 .50 4
Brazil Lula 0.3 .29 4
Uruguay Vsquez 0.3 .50 4
Argentina Kirchner 0.2 .29 4
Costa Rica Pacheco 0.2 .29 3
Panama Torrijos 0.2 .29 3
Bolivia Mesa 0.1 .25 4
Chile Lagos 0.1 .25 4
Guatemala Berger 0.0 .00 3
Nicaragua Bolaos 0.0 .00 4
Honduras Maduro 0.0 .00 3
Colombia Uribe 0.0 .00 4
Speech category
Campaign 0.6 .67 12
Ribbon-cutting 0.3 .48 19
International 0.4 .62 19
Famous 0.4 .62 19
Average 0.4 .60 (All categories
equally weighted)

The kappa statistic for these data is .44, indicating a moderate level of agree-
ment, although this figure is somewhat reduced by our inability to weight the
calculation for the ordinal nature of the scale.

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

1054 Comparative Political Studies

Table 2
Average Populism Scores for NonLatin
American Chief Executives
Chief Average Standard Number of
Country Executive Populism Score Deviation Speeches
Belarus Lukashenko 1.7 .27 4
United States Bush 1.2 .32 4
Iran Ahmahdinejad 1.2 .58 4
Ukraine Yushchenkoa 1.1 .85 4
Philippines Arroyoa 0.5 .41 4
Russia Putin 0.4 .50 4
United Blair 0.3 .50 4
Kingdom
Ghana Kufuor 0.2 .32 4
Norway Stoltenberg 0.2 .33 4
Mongolia Enkhbayar 0.1 .17 4
Bulgaria Stanishev 0.1 .17 4
Canada Harpera 0.0 .00 3
Finland Halonen 0.0 .00 4
South Africa Mbekia 0.0 .00 4
Spain Zapateroa 0.0 .00 4
Sweden Persson 0.0 .00 4
Speech category
Campaign 0.6 .67 15
Ribbon-cutting 0.3 .58 16
International 0.3 .52 16
Famous 0.6 .72 16
Average 0.4 .62 (All categories
equally weighted)

a. Only two graders participated.

Descriptive Results

Tables 1 and 2 present the results of these two phases of our research.
The first result that readers will notice is that Chvez indeed has a very
populist discourse. More important, Chvez stands together with a few
other current chief executives who are often regarded as populist, such
as Evo Morales in Bolivia, Alexander Lukashenko in Belarus, Victor
Yushchenko in Ukraine (at least when he first came to power), and inter-
estingly, George Bush in the United States. These results probably fit
the expectations of scholars and the public. Morales is an important ally of

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

 Hawkins / Is Chvez Populist? 1055

Chvez and leads a popular indigenous movement for revolutionary change

that has polarized the population since his election in 2005; Lukashenko has
long had a strongly nationalistic outlook and a rapport with poor voters,
especially in rural areas; and Yushchenko came to power at the head of a
popular, prodemocracy movement, the Orange Revolution. In a moment I
will explore the results for Bush.
Another important finding is that the measure of populist discourse
reflects more than just the discourse of current leaders. Several key his-
torical Latin American presidents who are usually considered populist
also have a strongly populist discourse, including Juan Domingo Pern in
Argentina, Jos Mara Velasco Ibarra in Ecuador, and to a lesser degree
Getlio Vargas in Brazil. While space here prevents a full analysis, my
assistants and I found that these classic populists generally incorporated all
of the elements of discourse identified in our rubric. Thus, there is consider-
able continuity between the populists of the past and those of today, a
continuity that suggests our rubric has tapped into something deep and
fundamental.
Another important finding is that populist discourse is a fairly rare phe-
nomenon. Despite recent talk of a wave of populism, we found only two
clear current examples of populist discourse in Latin America as of late 2005
and early 2006 (Chvez and Morales) and potentially 3 in our sample of 15
countries outside of Latin America (Lukashenko, Bush, and Yushchenko).
One of these 3 nonLatin American casesLukashenkowas not origi-
nally in the sample and was only included after we realized that the sample
had too little variance for the causal analysis I hoped to carry out in subse-
quent research. If we exclude this case, then only about 1 out of every 7
leaders in either sample had a strongly populist discourse.7 This may contra-
dict popular perceptions of developing regions such as Latin America, but it
fits the arguments of scholars who study the relationship of populism to
democracy (Canovan, 1999; Mudde, 2004) as well as the rough data given
in the few multicountry studies of populism (Betz, 1994; Conniff, 1999;
Kitschelt & McGann, 1995; Taggart, 1996). Populist parties and leaders are
almost always present in every country, but they typically rise to prominence
in moments of crisis and tend to be short-term, cyclical phenomena. If they
achieve their goals, they often become routinized and lose their fervor; and
if they fail to achieve their goals, they are removed from power by elitist or
pluralist forces (and sometimes new populists). Even countries thought of as
consistently populist, such as Argentina or Ecuador, often have nonpopulist
interregnums.

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

1056 Comparative Political Studies

Where populist discourse is more common in our data is precisely where

we expected it: the famous and campaign speeches. As the data in Tables 1
and 2 indicate, these categories of speeches were nearly twice as populist as
the international and ribbon-cutting speeches. This pattern manifested itself
in both of our samples, although it was somewhat stronger in the nonLatin
American sample.8 This has special implications for the few instances where
we had missing data, which in most cases was the campaign speech. The
strong overall average for this category suggests that in instances of missing
data, the actual average score for the leader may be a little higher.
A few of our findings are more unusual and deserve explanation. First,
we find that two current leaders who are sometimes considered lukewarm
populists, Nestor Kirchner in Argentina and Luis Igncio Lula da Silva in
Brazil, do not in fact have much of a populist discourse. Kirchner in par-
ticular uses a populist discourse very inconsistently. His campaign speech
is the only one that is somewhat populist (graded around a 1). In this speech
he frequently hints at a popular will and a notion of a romanticized common
man. The speech also has a kind of Manichaean quality that features limited
cosmic proportionality and bellicosityincluding brief mentions of heroes
such as Jos de San Martn, Simn Bolvar, Mariano Moreno, and the
Pernsas well as a sense of dualism. And there is a clear set of enemies,
mostly former politicians from the 1990s. However, both graders of his
speeches agreed that Kirchner did not use these elements with the same
consistency and power as Morales or Chvez, who frequently apply cosmic
proportions to even the most mundane issues. Moreover, the discourse he
used in this election is not the same as what he uses at other times; his three
other speeches were all graded 0 by both readers. These other speeches
lacked the bellicosity that was evident in the campaign speech and avoided
references to any romanticized will of the people or enemy.
With Lula, it is harder to find any speech that is strongly populist. We
focus on him more in the following, but for now I will simply point out that
none of his speeches have much in the way of a Manichaean quality.
Instead, Lula tends to focus on narrow issues and avoids any kind of cosmic
proportionality or the mention of historical figures. He consistently empha-
sizes consensus and negotiation and, while briefly criticizing some indi-
viduals or opposition groups (e.g., former president Cardoso and wealthy
Brazilians), he avoids characterizing these as evil. He does make brief men-
tion of a popular will in some of his speeches, reminding the audience of
his own working-class origins and telling them that he understands their
needs. Thus, his discourse is more consistent than Kirchners, but it is not
strongly populist.

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

 Hawkins / Is Chvez Populist? 1057

An important historical instance where we found an absence of strong

populist discourse is in the speeches of Lzaro Crdenas. Crdenas,
president of Mexico from 1934 to 1940, is routinely considered a populist
in older academic studies, most of which emphasize his expropriation of
U.S. oil companies in 1938, his land redistribution and pro-labor policies,
and his charisma (Conniff, 1999; Knight, 1998). However, Crdenass
discourse in the speeches we examined lacks the tone and consistently
populist content found in those of even his close contemporaries, such as
Pern, Velasco Ibarra, and Vargas; in fact, none of his speeches ever score
higher than a 1. This includes the speech he delivered shortly after the oil
expropriation, which was used as his famous speech, as well as a much
larger set of nearly 60 speeches that I analyze in the following. The breadth
of our analysis in this case gives us great confidence in these findings. Of
course, the accepted wisdom in this literature is generally based on nondis-
cursive definitions of populism, and the results we find here may reflect a
bit of historical timing. Crdenas represented a consolidated revolution, and
while the speeches we examined include strong, frequent references to a
General Will or romanticized people that are tinged with socialism (the
proletariat, the workers, complete emancipation of the people), and a
strong sense of cosmic proportions, they lack parallel references to a conspir-
ing elite or a notion of dualism. Instead, Crdenas emphasizes consolidating
or institutionalizing the gains of the Mexican Revolution and respect for law.
The last instance where we find a noticeable absence of populism is in
the speeches of Carlos Menem of Argentina, a noted neo-populist of the
1990s. During the decade when he was president, scholars debated whether
he and a few other politicians in his generation represented a new form of
populism, one relying on pro-market reforms and cross-class coalitions that
included wealthy entrepreneurs (Roberts, 1995; Weyland, 1996). In fact, our
analysis revealed that this particular leader had only a moderately populist
discourse. All four of Menems speeches utilize a kind of Manichaean dis-
course that frames issues as stark, morally weighty choices and that offers a
bold vision for a new future for Argentina, often coupled with references to
great figures and movements. For example, in the international speech he
talks about creating a new State, on solid moral foundations and trans-
forming the world into something more humane, and he insists that the
moment for change is now or never. Likewise, his ribbon-cutting speech
(the inauguration of a new fish cannery) waxes eloquent with references to
God and the Bible and their lessons for the appropriate model of economic
development. But his speeches rarely express any notion of a romanticized
popular will or a conspiring elite. Thus, it may be that Menem represents a
charismatic leader but not a populist one.9

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

1058 Comparative Political Studies

One final finding that I should comment on is the high level of populist
discourse in the speeches of Bush. Bush incorporates most of the elements
listed in the grading rubric: He presents issues in a broad, moral, dualistic
framework that ascribes cosmic proportions to his topic (Either you are
with us, or you are with the terrorists); he speaks about a common people
that represents the good (Americans and friends of liberal democracy every-
where); and he describes a conspiring threat that embodies evil (fundamen-
talist Islamic terrorism). Critics of Bush would also argue that he displays a
disregard for the rule of law and the accuracy of his dataan anything
goes attitudein defense of what he perceives to be a just cause.
That said, I suggest that it is probably not helpful to consider Bushs
discourse as populist, although it can certainly be considered as antagonistic.
Bushs discourse is not about rectifying past injustices suffered by the people
at the hands of an oppressive elite. His cause is the defense against a common
external enemy rather than revolution or systemic change. Nowhere in the
speeches we examined does he call for radically reforming the political and
economic system that governs liberal democracies and the United States.
Bush has urged changing key constitutional rights and the provisions of
international law in the fight against terrorism, but it is hard to imagine him
ever claiming that these institutions are the product of a subversive Islamic
cabal. Nor does he publicly demonize Muslims for supposedly undermin-
ing American values; instead, in the texts that we examined he reaffirms a
pluralist notion of religious tolerance.
Of course, the fact that our coders failed to make this distinction sug-
gests that our training could have been more careful. These elements of
populist discourse (the need for liberation or revolution from an elite that
has subverted the popular will) are stated in the rubric, but in retrospect they
could have been emphasized more. This seems to have been a problem only
in the larger, comparative phrase of our study where the training and expe-
rience of the graders were more condensed.

A Test of the Sampling Technique

Readers may still question the small sample size (four speeches) for
each chief executive. After all, perhaps the level of populism varies widely
across speeches, making it necessary to have a very large sample or even
making any single measurement of populism meaningless.
The variance in our results suggests that this is not a significant concern.
Tables 1 and 2 include the standard deviations of the average scores for each
leader. As can be seen, few if any leaders in the set have a standard deviation

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

 Hawkins / Is Chvez Populist? 1059

Table 3
Analysis of Random Samples of Lula
and Crdenas Speeches
Lula Crdenas

Grader 1 Grader 2 Grader 1 Grader 2

Mean 0.48 0.64 0.32 0.20

Standard deviation 0.51 0.48 0.50 0.44
Overall mean 0.56 0.24
Overall standard 0.39 0.35
deviation
N 42 60
Percentage 64.4 75.0
agreement
Kappa 0.27 0.33

over .50, and many have considerably lower. These figures are not very
large. By comparison, a leader with average scores of 2, 2, 2, and 1that is,
with all scores identical but onewould have a standard deviation of .50,
while the maximum possible standard deviation (associated with the set of
scores 2, 2, 0, 0) would be 1.2. The one outlier in the sample is Yushchenko
in Ukraine, with a standard deviation of .85; this result makes sense in light
of the extraordinary experience that brought him to power. Thus, most lead-
ers in the samples are fairly consistent in their use of populist discourse.
To further increase our confidence in these results, I analyzed a large,
random sample of speeches by two leaders in the set, Crdenas and Lula.
I selected these presidents because their discourses were harder to measure
(both were perceived as only mildly populist, somewhere between 0 and 1),
thereby presenting us with a more challenging test, and because my assist-
ants and I were reasonably certain that we had the entire universe of their
speeches. We then randomly selected one speech from each month of their
terms in office, 42 speeches for Lula and 60 for Crdenas. Because of the
large numbers of speeches, I asked the graders to dispense with any note
taking or written analysis besides a short set of comments and a grade.
This new grading technique was much faster (10 to 15 minutes per speech
instead of 30 to 45 minutes). Two native Portuguese speakers and two native
Spanish speakers conducted the grading.
Table 3 provides the results. In the case of Crdenas, the level of inter-
coder reliability is about as high as in previous phases of the project, with

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

1060 Comparative Political Studies

75% absolute agreement and a kappa of .33 (the kappa is low because
Crdenas never receives above a 1, thereby generating a high level of
expected agreement). In the case of Lula the level of intercoder reliability
is somewhat lower. The absolute agreement is only 64% and the kappa
statistic is only .27. This may be a result of having to grade in-between
speeches using a 3-point ordinal scale. Our graders indicated afterward that
many of Lulas speeches were right between a 0 and a 1 (hence the aver-
age scores of around .50), a pattern that forced them to make many hard
decisions. The fact that the average scores of each grader across the entire
Lula sample were almost indistinguishable from each other suggests that
the lack of agreement was not a problem of bias or inadequate training but
of small differences in judgment magnified by the scale.10 In future rounds
of analysis we may want to try a more continuous scale.
The more specific question, however, is whether these results show
that we were justified in using a small sample in our two previous phases.
One indicator of the robustness of our sampling criteria is whether the
average scores for Lula and Crdenas from the first phase of our project
were close to the average scores from the new analysis. Indeed, the actual
differences between these two phases are not very large, only about .31 in
the case of Lula and .32 in the case of Crdenas, differences that are sig-
nificant at only the p < .12 level and p < .23 level, respectively (t test with
unequal variance). Given the fact that we used different sampling criteria in
these two different phases (one a nonrandom sample from four speech cat-
egories that took context into account, the other a random sample from all
available speeches), these similarities are actually quite striking. Again,
they suggest remarkable continuity in each leaders discourse.
The other important indicator of the effectiveness of our sampling criteria
is the size of the variance in our data and especially the difference in vari-
ances across the samples. If the larger, random sample yields a dramatic
change in the variance of our estimates, we may not be justified in relying on
such small samples in the first phases of the analysis. As it turns out, the
variances of these two samples are nearly identical and not very large. In the
original analysis, the four scores for Lula had a standard deviation of .29 and
those for Crdenas had a standard deviation of .35, while in the second, larger
samples, the standard deviations are only .36 and .43, respectively. Using the
Levene test for difference in variance, the difference between the earlier and
later standard deviations is not significant by common standards for either
president (p < .71 for Lula and p < .69 for Crdenas).11

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

 Hawkins / Is Chvez Populist? 1061

Conclusion

In this article I have enhanced our understanding of populism by pre-

senting a technique for measuring populist discourse. After laying out the
discursive definition in the context of other approaches to populism, I
measure populist discourse quantitatively across a broad sample of leaders
from different countries and times. I do this by holistically analyzing sam-
ples of speeches of chief executives. This novel technique produces mean-
ingful results with reasonable face validity and moderate or high reliability,
even with small sample sizes. It shows a small percentage of highly popu-
list leaders today (including Chvez) and finds that their discourses are
very similar to well-known populists from the past. This implies that pop-
ulism defined as discourse is a reasonably coherent and consistent phe-
nomenon that we cannot brush aside for methodological reasons. The
burden is now on critics to clarify their definitions and apply them to
similar empirical tests.
This technique is fairly efficient in comparison with other standard tech-
niques of textual analysis. Although initial runs of the exercise took 30 to
45 minutes per speech, the later version using a random sample of speeches
required only 10 to 15 minutes. These are respectable times for human-coded
content analysis. When we consider that the exercise naturally incorpo-
rated the interpretation of the text into the measurement process, these
times even begin to compare favorably with some of the newer computer-
coded content analyses, which leave most of the interpretation for afterward
and incorporate fewer checks of reliability into this stage.
The article has bracketed the question of mass-level reception or recrea-
tion of elite populist discourse, that is, whether activists and voters actually
respond to the discourse or if perhaps they follow populist leaders for other
reasons. However, the results here have implications for mass-level analysis
and measurement. Currently, mass-level texts comparable to political
speeches are not readily available, making surveys one of our best alterna-
tives. Survey research typically shows a preference for short, precise,
closed-ended questions because of c oncerns about coding (especially cost)
and the reliability of results. However, the use of holistic grading here
suggests that we can make use of longer answers to questionnaires.
Surveys that hope to gauge such broad, latent cultural phenomenon as
political discourse might include questions soliciting a long, recorded
response rather than trying to break the discourse down into its constituent

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

1062 Comparative Political Studies

elements and measuring these separately. The text can still be graded
efficiently and is more likely to generate valid results.
The analysis has implications not only for the study of populism but for
the broader study of culture in comparative politics. Political scientists have
long used surveys and content analysis to gauge the presence of certain
ideas in the minds of politicians and voters, but we have yet to face chal-
lenging concepts such as political discourse that reflect intersubjective
views of culture. The method I have used hereholistic gradingis one
that tries to be faithful to the theoretical insights and ontological assump-
tions of the interpretivist approach while still maintaining positivist stand-
ards of measurement. Traditional discourse analysts may or may not be
happy with these results. After all, this is still an attempt to quantify what
some may see as unquantifiable, and it glosses over important qualitative
distinctions that we can only see by closely analyzing particular speeches.
My point in this analysis is not to discredit qualitative techniques but to
complement them with quantitative ones that can enhance our understand-
ing while still respecting culturalist insights.

Appendix
Populist Speech Rubric
Name of politician:
Title of Speech:
Category:
Grader:
Date of grading:
Final Grade (delete unused grades):
2 = A speech in this category is extremely populist and comes very close to
the ideal populist discourse. Specifically, the speech expresses all or
nearly all of the elements of ideal populist discourse and has few ele-
ments that would be considered nonpopulist.
1 = A speech in this category includes strong, clearly populist elements
but either does not use them consistently or tempers them by including
nonpopulist elements. Thus, the discourse may have a romanticized
notion of the people and the idea of a unified popular will but avoid
bellicose language or references to cosmic proportions or any particu-
lar enemy.
0 = A speech in this category uses few if any populist elements.
(continued)

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

 Hawkins / Is Chvez Populist? 1063

Appendix (continued)
Populist Pluralist

It conveys a Manichaean vision of the world, The discourse does not frame issues in moral
that is, one that is moral (every issue has a terms or paint them in black and white.
strong moral dimension) and dualistic Instead, there is a strong tendency to focus
(everything is in one category or the other, on narrow, particular issues. The
right or wrong, good or evil). The discourse will emphasize or at least not
implicationor even the stated ideais that eliminate the possibility of natural,
there can be nothing in between, no fence justifiable differences of opinion.
sitting, no shades of gray. This leads to the
use of highly charged, even bellicose
language.
The moral significance of the items mentioned The discourse will probably not refer to any
in the speech is heightened by ascribing reified notion of history or use any cosmic
cosmic proportions to them, that is, by proportions. References to the spatial and
claiming that they affect people everywhere temporal consequences of issues will be
(possibly but not necessarily across the limited to the material reality rather than
world) and across time. Especially in this any mystical connections.
last regard, frequent references may be
made to a reified notion of history. At the
same time, the speaker will justify the
moral significance of his or her ideas by
tying them to national and religious
leaders that are generally revered.
Although Manichaean, the discourse is still Democracy is simply the calculation of votes.
democratic, in the sense that the good is This should be respected and is seen as the
embodied in the will of the majority, foundation of legitimate government, but it
which is seen as a unified whole, perhaps is not meant to be an exercise in arriving at
but not necessarily expressed in references a preexisting, knowable will. The major-
to the voluntad del pueblo; however, the ity shifts and changes across issues. The
speaker ascribes a kind of unchanging common man is not romanticized, and the
essentialism to that will, rather than letting notion of citizenship is broad and legalistic.
it be whatever 50% of the people want at
any particular moment. Thus, this good
majority is romanticized, with some notion
of the common man (urban or rural) seen
as the embodiment of the national ideal.
The evil is embodied in a minority whose The discourse avoids a conspiratorial tone
specific identity will vary according to and does not single out any evil ruling
context. Domestically, in Latin America it is minority. It avoids labeling opponents as
often an economic elite, perhaps the evil and may not even mention them in an
oligarchy, but it may also be a racial elite; effort to maintain a positive tone and keep
internationally, it may be the United States; or passions low.
the capitalist, industrialized nations; or
international financiers; or simply an ideology
such as neoliberalism and capitalism.

(continued)

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

1064 Comparative Political Studies

Appendix (continued)
Populist Pluralist

Crucially, the evil minority is or was recently The discourse does not argue for systemic
in charge and subverted the system to its change but, as mentioned above, focuses
own interests, against those of the good on particular issues. In the words of
majority or the people. Thus, systemic Laclau, it is a politics of differences
change is/was required, often expressed in rather than hegemony.
terms such as revolution or liberation
of the people from their immiseration or
bondage, even if technically it comes
about through elections.
Because of the moral baseness of the threaten- Formal rights and liberties are openly
ing minority, nondemocratic means may be respected, and the opposition is treated
openly justified or at least the minoritys with courtesy and as a legitimate political
continued enjoyment of these will be seen as actor. The discourse will not encourage or
a generous concession by the people; the justify illegal, violent actions. There will
speech itself may exaggerate or abuse data be great respect for institutions and the
to make this point, and the language will rule of law. If data is abused, it is either an
show a bellicosity toward the opposition that innocent mistake or an embarrassing
is incendiary and condescending, lacking the breach of democratic standards.
decorum that one shows a worthy opponent.
Overall Comments (just a few sentences):

Notes
1. That this typology is incomplete can be seen in the fact that it is threefold, while the exist-
ence of two dimensions suggests something fourfold. Elitism seems to be the problematic type, in
that it fails to distinguish between redemptive and pragmatic nondemocratic discourses.
2. The fact that a populist leader has a large popular following does not necessarily mean
that activists and voters are populist. As de la Torre (2007) points out, citizens may follow
populists to send a message to traditional politicians because of material inducements or sim-
ply because the politician is entertaining.
3. Psychometricians now generally acknowledge the inadequacy of correlations coeffi-
cients for measuring intercoder reliability as they fail to indicate actual agreement (Ebel, 1951;
Tinsley & Weiss, 1975). Consequently, recent studies make use of much more sophisticated
techniques for assessing reliability of holistic testing while finding similar, positive results (cf.
Sudweeks, Reeve, & Bradshaw, 2005).
4. We found this out early in the analysis when we attempted to measure populism using a
standard content analysis; no single word or sentence ever conveyed the broader set of meanings.
5. The anchor texts included an international speech by Robert Mugabe of Zimbabwe
(generally graded a 2 by the readers), an international speech by Evo Morales of Bolivia (also
generally graded a 2), a campaign speech by James Harper of Canada (graded a 1), and an
international speech by Tony Blair (graded a 0). Of course, none of these anchor texts were
included in the subsequent analysis.
6. While 92 Latin American speeches were coded in the first phrases of the analysis, as
indicated in Table 1, some of these were speeches by Brazilian presidents that were coded by
other readers in the second phase.

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

 Hawkins / Is Chvez Populist? 1065

7. The Latin American sample is almost the entire universe of cases in this region, and
this overall result should be seen as representative. To choose cases for the second, nonLatin
American sample, we used a decision rule that seems unlikely to be correlated with populism,
namely, we considered countries for which a sizeable population of undergraduate students
was available to work as readers at my university. Thus, the overall descriptive results here
may be representative as well.
8. If we consider the two samples separately, the difference between the campaign speeches
and either the ribbon-cutting or international speeches is statistically significant at the p < .05 to
p < .10 level in both samples (one-tailed test); however, the difference between the famous
speeches and either the ribbon-cutting or international speeches is generally not significant.
9. This confirms the characterization of Menems discourse by Novaro (1998), who argues
that Menems notion of the people was more inclusive than that of Pern. It may also substantiate
some of the qualifications of the concept of neopopulism subsequently made by Weyland (2001).
10. Although the difference in means is statistically significant for Lula, with a t test of
p < .09, the actual difference between the two graders in this phase is only .16. Similar results
prevail with the Crdenas analysis, where a t test is significant at the p < .07 level but the actual
difference is only .12.
11. Because the distributions of scores are not symmetric, the Levene test is calculated
using the median in place of the mean.

References
The Americas: The return of populism; Latin America. (2006, April 15). The Economist, p. 58.
Armony, A. C., & Armony, V. (2005). Indictments, myths, and citizen mobilization in Argentina:
A discourse analysis. Latin American Politics & Society, 47(4), 27-54.
Barry, B. (1978). Sociologists, economists and democracy. Chicago: University of Chicago
Press.
Betz, H. G. (1994). Radical right-wing populism in Western Europe. New York: Macmillan.
Britton, J. N., Martin, N. C., & Rosen, H. (1966). Multiple marking of English compositions:
An account of an experiment. London: Her Majestys Stationery Office.
Budge, I., Bara, J., Volkens, A., & Klingemann, H. D. (2001). Mapping policy preferences:
Estimates for parties, electors and governments, 1945-1998. New York: Cambridge University
Press.
Canovan, M. (1999). Trust the people! Populism and the two faces of democracy. Political
Studies, 47, 2-16.
Cardoso, F. H., & Faletto, E. (1979). Dependency and development in Latin America. Berkeley:
University of California Press.
Castaeda, J. G. (2006). Latin Americas left turn. Foreign Affairs, 85(3), 28-43.
Coffman, W. E. (1971). Essay examinations. In R. L. Thorndike (Ed.), Educational measure-
ment (pp. 271-302). Washington, DC: American Council on Education.
Conniff, M. L. (Ed.). (1999). Populism in Latin America. Tuscaloosa: University of Alabama
Press.
Cooper, C. R. (1977). Holistic evaluation of writing. In C. R. Cooper & L. Odell (Eds.),
Evaluating writing: Describing, measuring, judging (pp. 3-32). Urbana, IL: National Council
of Teachers of English.
de la Torre, C. (2000). Populist seduction in Latin America: The Ecuadorian experience.
Athens: Ohio University Center for International Studies.

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

1066 Comparative Political Studies

de la Torre, C. (2007). The resurgence of radical populism in Latin America. Constellations,

14, 384-397.
Di Tella, T. S. (1965). Populism and reforms in Latin America. In C. Vliz (Ed.), Obstacles to
change in Latin America (pp. 47-74). New York: Oxford University Press.
Di Tella, T. S. (1997). Populism into the twenty-first century. Government and Opposition, 32,
187-200.
Dornbusch, R., & Edwards, S. (Eds.). (1991). Macroeconomic populism in Latin America.
Chicago: University of Chicago Press.
Ebel, R. L. (1951). Estimation of the reliability of ratings. Psychometrika, 16, 407-424.
Eckstein, H. (1996). Culture as a foundation concept for the social sciences. Journal of
Theoretical Politics, 3, 471-497.
Germani, G. (1978). Authoritarianism, fascism, and national populism. New Brunswick, NJ:
Transaction Books.
Gerring, J. (1997). Ideology: A definitional analysis. Political Research Quarterly, 50, 957-994.
Hofstadter, R. (1966). The paranoid style in American politics and other essays. New York:
Alfred A. Knopf.
Ianni, O. (1975). La formacin del Estado populista en Amrica Latina [The creation of the
populist state in Latin America]. Mxico, DF: Ediciones Era.
Jagers, J., & Walgrave, S. (2007). Populism as political communication style: An empirical
study of political parties discourse in Belgium. European Journal of Political Research,
46, 319-345.
Kazin, M. (1998). The populist persuasion: An American history (Rev. ed.). Ithaca, NY:
Cornell University Press.
King, G., Keohane, R. O., & Verba, S. (1994). Designing social inquiry: Scientific inference
in qualitative research. Princeton, NJ: Princeton University Press.
Kitschelt, H., & McGann, T. (1995). The radical right in Western Europe: A comparative
analysis. Ann Arbor: University of Michigan Press.
Knight, A. (1998). Populism and neo-populism in Latin America, especially Mexico. Journal
of Latin American Studies, 30, 223-248.
Knight, K. (2006). Transformations of the concept of ideology in the twentieth century.
American Political Science Review, 100, 619-626.
Laclau, E. (2005). On populist reason. London: Verso.
Lichbach, M. I., & Zuckerman, A. S. (Eds.). (1997). Comparative politics: Rationality, cul-
ture, and structure. New York: Cambridge University Press.
Little, D. (1991). Varieties of social explanation: An introduction to the philosophy of social
science. Boulder, CO: Westview.
McGuire, J. W. (1997). Peronism without Pern: Unions, parties, and democracy in Argentina.
Palo Alto, CA: Stanford University Press.
Mudde, C. (2004). The populist Zeitgeist. Government and Opposition, 39, 541-563.
Mudde, C. (2007). Populist radical right parties in Europe. New York: Cambridge University Press.
Neuendorf, K. A. (2002). The content analysis guidebook. Thousand Oaks, CA: Sage.
Novaro, M. (1998). Populismo y gobierno. Las transformaciones en el peronismo y la consolidacin
democrtica argentina [Populism and government: The transformations of Peronism and the
consolidation of Argentine democracy]. In F. Burbano de Lara (Ed.), El fantasma del populismo:
Aproximacin a un tema (siempre) actual [The phantom of populism: Approach to an (Always)
Current Topic] (pp. 9-24). Caracas, Venenzuela: Editorial Nueva Sociedad.
Panizza, F. (2005). Introduction: Populism and the mirror of democracy. In F. Panizza (Ed.),
Populism and the mirror of democracy (pp. 1-31). London: Verso.

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012

 Hawkins / Is Chvez Populist? 1067

Quinn, K. M., Monroe, B. L., Colaresi, M., Crespin, M. H., & Radev, D. R. (2006). An automated
method of topic-coding legislative speech over time with application to the 105th-108th U.S.
Senate. Unpublished manuscript.
Riker, W. H. (1982). Liberalism against populism: A confrontation between the theory of
democracy and the theory of social choice. Prospect Heights, IL: Waveland Press.
Roberts, K. M. (1995). Neoliberalism and the transformation of populism in Latin America:
The Peruvian case. World Politics, 48, 82-116.
Roberts, K. M. (2003). Social correlates of party system demise and populist resurgence in
Venezuela. Latin American Politics and Society, 45(3), 35-57.
Roberts, K. M. (2006). Populism, political conflict, and grass-roots organization in Latin America.
Comparative Politics, 38, 127-148.
Roxborough, I. (1984). Unity and diversity in Latin American history. Journal of Latin American
Studies, 16, 1-26.
Seligson, M. (2007). The democracy barometers: The rise of populism and the left in Latin
America. Journal of Democracy, 18, 81-95.
StataCorp. (2003). Stata base reference manual, Vol. 2, Release 8. College Station, TX: Stata
Corporation.
Sudweeks, R. R., Reeve, S., & Bradshaw, W. S. (2005). A comparison of generalizability theory
and many-facet Rasch measurement in an analysis of college sophomore writing. Assessing
Writing, 9, 239-261.
Taggart, P. (1996). The new populism and the new politics: New protest parties in Sweden and
in a comparative perspective. New York: St. Martins.
Tinsley, H. E. A., & Weiss, D. J. (1975). Interrater reliability and agreement of subjective
judgments. Journal of Counseling Psychology, 22, 358-376.
Weffort, F. C. (1973). Clases populares y desarrollo social [Popular classes and social develop-
ment]. In F. C. Weffort & A. Quijano (Eds.), Populismo, marginalizacin y dependencia:
Ensayos e interpretacin sociolgica [Populism, marginalization, and dependency:
Essays and sociological interpretation] (pp. 17-169). San Jose, Costa Rica: Editorial
Universitaria Centroamericana.
Weyland, K. (1996). Neopopulism and neoliberalism in Latin America. Studies in Comparative
International Development, 31, 3-31.
Weyland, K. (2001). Clarifying a contested concept: Populism in the study of Latin American
politics. Comparative Politics, 34, 1-22.
Weyland, K. (2003). Economic voting reconsidered: Crisis and charisma in the election of
Hugo Chvez. Comparative Political Studies, 36, 822-848.
White, E. M. (1985). Teaching and assessing writing. San Francisco: Jossey-Bass.
Wiles, P. (1969). A syndrome, not a doctrine. In G. Ionescu & E. Gellner (Eds.), Populism: Its
meaning and national characteristics (pp. 166-79). New York: Macmillan.
Wst, A. M., & Volkens, A. (2003). Euromanifesto coding instructions (Working Paper No. 64).
Mannheim, Germany: Mannheimer Zentrum fr Europische Sozialforschung.

Kirk A. Hawkins is an assistant professor in the Department of Political Science at Brigham

Young University. He received his PhD in comparative politics from Duke University in 2003.
He has previously published articles on Chavismo, populism, and Venezuelan politics in Latin
American Research Review and Third World Quarterly and has published book chapters analyz-
ing Christian Democratic party formation in Latin America and the social spending programs of
the Chvez government. He is currently finishing a book-length manuscript on Chavismo
and populism in comparative perspective.

Downloaded from cps.sagepub.com at University of Alabama at Birmingham on December 6, 2012