1

COMPARATIVE POLITICS
PSCM 2133 [8-713]
Prelims

1. What is Comparative Politics? Why compare?

COMPARATIVE POLITICS - Examine Political Realities in Country all over the world. It looks at many ways the government
operates, and ways people behave in their political lives.

- 1. Concerned with comparative study and analysis of political system.

- 2. The primary focus of both theoretical and empirical work in comparative politics is on the
comparison of institutional practices between states.
- 3. Aims to develop understanding of how and why different institutions have the effects on
political outcomes that they do.

-Five Principles of Politics:

1. All political behavior has a purpose.
* People's goals are achieved through political actions.
2. Cooperation through bargaining or collective action is difficult, and the difficulty mounts as the
number of participants grows. All politics is collective action.
* Incentives are necessary to induce cooperation through informal and formal bargaining.
* Informal bargaining is determined by personal preferences and is referred to as horse-
trading, back-scratching, logrolling, etc.
* Formal bargaining is governed by rules.
3. Rules matter: procedures shape politic
* Rules and practices, which are socially constructed, govern the lives of all political actors.
4. Political outcomes are the products of individual preferences and institutional procedures.
* Individual preferences are determined by ideological, personal, electoral, and institutional
ambitions and interests.
* These preferences for in response to procedures (and, some times, vice-versa)
5. History matters. How did we get to where we are?
* Political memories transmit across time.
* Politicians remember what procedures solved past political problems and which policies do
not work.
* History provides an interpretative framework for political conflicts.
* History provides a context to incorporate current information and experience.
Help from History
Knowing history is a helping hand in shedding light as to why another historical phenom look ena
happen.
* Major turning points in world politics in the 20th Century - Conclusions of major wars:
* WWI
* WWII
* Cold War

* 9/11: American Projection of Power

* Q: How do these international historical timelines shape and/or influence the political history
and culture of my case country?
 2

Challenge of doing comparative studies…

1. There is à lot going on!
2. Events are complicated!
3. People resist ideas that undermine their ways of viewing the world
4. Need for a "verifiable" basis of comparison
How do we intelligently examine these events happening in one country and accurately compare
it with what is taking place in another?

WHY COMPARE? - for better/ easier understanding

– to describe [CHARACTERISTICS],
- to explain [VARIATION],
- to simplify [ABSTRACTING FROM COMPLEXILTY AND NUANCE]
- to evaluate [e.g., POOR, or SUCCESSFUL]

2. Comparative Analysis Within Political Science written by Alexander Stafford

= introduction to the practical, conceptual, and theoretical values of comparative analysis within political science.
= explain the importance of its role and the benefits it brings to the political field of research
= focus on the benefits of comparatively analyzing the collating institutions and processes of two or more countries as
opposed to one

Comparative analysis (CA) is a methodology within political science that is often used in the study of
political systems, institutions or processes. This can be done across a local, regional, national and
international scale. Further, CA is grounded upon empirical evidence gathered from the recording and
classification of real-life political phenomena. Where-by other political studies develop policy via
ideological and/or theoretical discourse, comparative research aims to develop greater political
understanding through a scientifically constrained methodology. Often referred to as one of the three
largest subfields of political science, It is a field of study that was referred to as “the greatest intellectual
achievement” by Edward A. Freeman (Lijphart, ND).

Using the comparative methodology, the scholar may ask questions of various political concerns, such
as the connection, if any, between capitalism and democratization or the collation between federal and
unitary states and electoral participation. CA can be employed on either a single country (case) or
group of countries. For the study of one country to be considered comparative, it is essential that the
findings of the research are referenced into a larger framework which engages in a systematic
comparison of analogous phenomena. Subsequent to applying a comparative methodology in the
collation or collection of data, established hypothesises can then be tested in an analytical study
involving multiple cases (Caramani, 2011).

Patterns, similarities and differences are examined to assess the relationships of variants between the
two or more separate systems. It is this nature of the analysis that renders it comparative. Henceforth
the researcher is subsequently able to isolate the independent variables of each study case. If the
independent variables of “X” and “Y” exist, their relationship to dependant variable “Z” can be
hypothesised, tested and established (Landman, 2008). This isolation is essential for the most defining
and significant strength of CA, that is to establish the hypothetical relationships among variables (Guy,
2011). This empirical analysis can be used to explain a system, present theoretical ideas for
modification and even to reasonably predict the future consequence of the case study in question.

While some researchers may favour a large amount of countries for their study (large-N) others will
use a smaller amount of units (small-N) (Guy, 1988). The size of the case study is directly collated with
the subject and it must lend the study sufficient statistical power. The researcher decides whether it is
 3

most appropriate to study one or more units for comparison and whether to use quantitative or
qualitative research methods (Guy, 1988). The methodology of utilising multiple countries when
analysing is the closet replication of the experimental method used in natural science (Lim, 2010). A
clear strength of this method is the inclusion of the ability to implement statistical controls to deduct
rival explanations, its ability to make strong inferences that hold for more cases, and its ability to
classify ‘deviant’ countries that contradict the outcomes expected from the theory being tested (Guy,
2011).

Deviant countries or cases are units which appear to be exceptions to the norm of the theory being
analysed. They are most prevalent in studies of processes and institutions involving only one country.
This is due to the fact that there is often a severely limited amount of variability being tested (Lim,
2010). In testing for the relationship between income inequality and political violence in sixty countries,
Muller and Seligson identified which countries collated with their theory and which did not. Brazil,
Panama, and Gabon were found to have a lower level of political violence than was expected for their
national level of income inequality. Alternatively, with a low national level of income inequality, the UK
was shown to have a higher than expected amount of political violence (Harro and Hauge, 2003). This
identification of these ‘deviant’, cases allowed researchers to look for the explanations. They were able
to deduct them from their analysis and increase the accuracy of their predictions for the other cases.
This could not have been achieved in a single-country study and would have inevitably left the findings
unbalanced and inaccurate.

Selection bias is a reductive practice that is most common with single-country studies. It arises through
the deliberate prejudice of countries chosen for examination. The most damaging form of selection
bias to the validity of the research is when only case(s) that support the theory being hypothesised are
analysed. The serious problem of selection bias occurs much less frequently in studies that contain
multiple countries (Lim, 2010). This is because Studies that compare institutions and processes in
multiple countries often rely on a sufficient number of observations that reduce the problem or at least
its effects of selection bias. Using multiple countries reduces the risk of this invalidity causing
phenomena.

However, the research of a single country is clearly valuable; it can produce an insightful exploration
of many domestic institutions such as social healthcare, and also processes such as immigration. The
findings however are mostly applicable to the country of analysis. Whereas the findings of multi-
national CA are also domestically valuable, they also tend to be more inherently valuable to the wider
international field. This is because the comparisons of multi-national institutions and processes that
are functionally collated have an increased global validity and transferability than the findings of
comparison of a single nation (Keman, 2011). The comparative results of a single nation must be
hypothesised with other understandings and predications relying substantially on theoretical
observations, assumptions and past studies. Inevitably, studying more than one country lends the
study a greater field of which to analyse. It is by the CA of subjects from multiple countries that thematic
maps can be developed, national, regional and global trends can be identified, and transnational
organisations can make acutely informed decisions. These practical benefits are not possible when
analysing a phenomenon from one country without cross-case comparison. Analysing multiple cross-
national units also furthers our understanding of the similarities, differences and relationships between
the case study itself, and the geo-political, economic, and socio-cultural factors that would otherwise
escape unaccounted.

The popularity of comparative method of analysing two or more countries has steadily increased
(Landman, 2008). Indeed it can be regarded as essential to the understanding and development of
modern day political, and international relation’s theory. With the constant dissolvent of countries
around the world such as Yugoslavia and Czechoslovakia combined with the potential for the creation
of new states such as Palestine; the CA which involves different countries offers a wealth of information
and most importantly, prediction for their futures. This theoretical framework for prediction is invaluable
to society.

Alternatively analysis applied to a single-nation case is less applicable on a global scale (Lim, 2010).
For example; studying the process of democratization in one Latin American country, although it offers
important inferences that can be examined in other countries with a similar set of circumstances, is
arguably insufficient to develop a theory of democratization itself that would be globally applicable.
 4

Quite simply, the singular analysis of an institution or process involving only one country often fails to
provide a global set of inferences to accurately theorise a process (Harro and Hauge, 2003).

Comparing and contrasting processes and institutions of two or more countries allows the isolation of
specific national variants (Hopkin, 2010). It also encourages the clear revealing of common similarities,
trends and causation and the deduction of false causation. This means that established hypothesis
are continually ripe for revaluation and modification. It enables the researcher to minimise the reductive
phenomenon of having ‘too many variables not enough countries’, this occurs when the researcher is
unable to isolate the dependant variable of the study because there are too many potential variables
(Harro and Hauge, 2003). This problem is far more associated with single-country studies because it
results from a surplus of potential explanatory factors combined with an insufficient amount of countries
or cases in the study (Harro and Hauge, 2003).

Studies involving multiple countries assists in the defining of results as being idiographic in nature or
nomothetic (Franzese, 2007). It also assists in making the important distinction between causation,
positive correlation, negative correlation and non-correlation. When analysing only one case or country
it is harder to correctly make a distinction between these relationships especially one that is not only
subject to the one country.

It is by studying institutions and processes of different countries by use of an empirical methodological

framework, that the researcher is able to realise inferences without the ambiguity of generalisations.
The separation of the cases being compared offers the researcher a richer study ground of variables
that assist in acutely testing hypothesises and in the creation of others. It is through CA that correlating,
dependent and independent relationships can be identified (Lim, 2010). The inclusion of multiple
countries in a study lends the findings wider validity (Keman, 2011). For example, Gurr demonstrated
that the amounts of civil unrest in 114 countries are directly related to the existence of economic and
political deprivation. This theory holds true for a majority of countries that it is tested with (Keman,
2011).

It should also be noted that all countries, to differing degrees, are functioning in an interdependent
globalized environment. Because of immigration, economic and political interdependence, the study
of an institutions and/or processes within a single country inevitably gives a reduction in the
transferability of the findings. This is because the findings at least are only as applicably transferable
as their counterparts are functionally equivalent. It also somewhat fails to account for transnational
trends (Franzese, 2007). Alternatively, comparison involving multiple nations, especially using
quantitative techniques, can offer valuable empirically-based geopolitical and domestic
generalizations. These assist in the evolution of our understanding of political phenomena and produce
great recommendations into how to continue particular research using the same form of analysis or a
different method all together.

The study of processes and institutions within two or more countries has been criticised for producing
less in-depth information compared to studies involving one country (Franzese, 2007). While this is
appears to be a substantial criticism; there is not always agreement between scholars that this trade-
off between quantity and quality is substantial, or indeed extremely relevant. Robert Franzese claims
that the relative loss of detail which results from analysing large amounts of cross-national cases, does
not justify retreating to qualitative study of a few cases (Franzese, 2007). This is because most
generalizations from single-country studies will inevitably be limited, since the country as a unit is
bound by unique internal characteristics.

It is clear that both single-nation and multinational studies play an important role in CA. Yet as
evidenced above, the strengths of encompassing multiple countries into comparative research far
outweigh any reduction in the quality of the findings. Indeed, multinational studies work to reduce
selection bias, and encourage global transferability, assist in variable deduction and receive
recognition as being empirically scientific.

3. Latin America and Beyond: The Case for Comparative Area Studies written by Bert Hoffmann
 5

Focusing on Latin America and the Caribbean as a field of study (or the focus of ajournai, as in the case of
ERLACS) is a statement in itself: It conveys the argument that there is something specific about these societies which
makes it useful to study them together, and to distinguish them from other possible groupings (Whitehead, 1994, p.
1). In this sense, area studies - even if focused on a single world region - seem inherently comparative by nature.1
This article proposes to make the comparative nature of Latin American Studies more explicit than it often is. It argues
for a notion of Comparative Ar- ea Studies (CAS) that endorses an approach of explicit intra-, inter- and cross area
comparisons which builds on the traditional strengths of area studies while at the same time enhancing their impact
on broader conceptual and methodo- logical debates. Such an approach, however, requires the systematic coopera-
tion between scholars with in-depth knowledge of different world regions and their respective global connections; and
it requires adequate organizational and institutional structures to support their cooperation.
The long shadow of the area studies debate
During its 50 years of existence, ERLACS has been part of the remarkable suc- cess story of Latin American
Studies. Taking just one indicator to illustrate the rapid growth of these studies, the meetings of the Latin American
Studies As- sociation - turning 50 next year - have become truly massive events with the congress in 2015 drawing
more than 5,000 participants. However, a constant companion to this development has been the criticism and at
times outright disrespect from the so-called systematic disciplines against all area studies, such as Latin American
Studies. In what some have called the 'area studies war' (Waters, 2000) area studies have been criticized for their
supposed 'horizontal ignorance' - for being parochial, immersed in region- or country-specific knowledge but unable
to look beyond it, lacking in theoretical and methodological rigor, and hence having an inability to arrive at
generalizable findings that 'speak' to a broader scholarly community. Area studies scholars have rebutted such
criticism many times, starting with the fact that they themselves usually have a solid disciplinary background and that
it is more the combination of disciplinary and regional expertise which is at the base of area studies, not an either/or.
From this basis, area studies scholars have been bashing the 'generalists' from mainstream disciplines for their 'ver-
tical ignorance' - their lack of language and cultural understanding, shallow historic depth, and ignorance of the
scholarly work emanating from the coun- tries or areas themselves (e.g. Szanton, 2004). (These examples are drawn
from political science as this is the field I know best. While similar debates have surfaced in other disciplines as well,
a more detailed look than this brief article allows would shed light on important differences between the different disci-
plines.)
While the aggressiveness of the 'area studies war' may now have given way to some sort of peaceful co-
existence (Bates, 1997; Harbeson, et al. 2001), the underlying challenges remain. This article argues for a
Comparative Area Stud- ies (CAS) approach as a promising way (by no means the only one, of course) for Latin
American and Caribbean studies scholars to move forward.
Making comparisons explicit
'And what should they know of England who only England know?', Rudyard Kipling (1891) once famously
asked. For social scientists, the need to compare seems inevitable. To understand what is specific about a given
polity one needs to know others. To identify the distinctive characteristics of culture, politics, society or economy in
Latin America and the Caribbean, we need to see these in the light of culture, politics, society or economy elsewhere.
The comparative area studies approach argues to make this comparative perspective explicit rather than leaving it
implicit. In the case of the latter, all too often the authors' own background tends to be the built-in comparative horizon
by which they structure, measure and judge the empirical realities they study. In area studies this has a long pedigree.
If we think of Tacitus as one of the early forefathers of western area studies scholarship, his 'Germania' describes the
Germanic tribes along the categories of the author's imperial Roman socie- ty - from public affairs to family law, from
social hierarchies to wealth and trade - and then describes all he finds to deviate from the Roman template.
Similarly, since the nineteenth century, modern area studies have emerged as a 'child of empire', often driven
by political and commercial interests or the perceived 'civilizing mission' of the colonial powers. They have ever since
been criticized for their built-in Euro-centrism: that the colonial powers (or later the OECD world) set the norms and
yardsticks to which the rest of the world had to measure up. In this tradition, area studies have been about 'the others'.
There are no 'German studies' in Germany, nor are there 'Brazil stud- ies' in Brazil. At home, it seems, the multi-
disciplinary approach that charac- terizes area studies scholarship is blocked by the deeply rooted separation of
academic disciplines and departments. Remarkably, this concept of area studies as being about 'the others' has also
been replicated outside the traditional 'West'; the current expansion of area studies centres in China provides ample
illustration.
From its origins, the perspective of area studies on 'the others' was attached to profound asymmetries of
power and knowledge production. Over the past decades, however, scholarship from Latin American authors on Latin
Ameri- can issues has been absolutely key in enriching and dynamizing the field in recent years, and it has proven
 6

to be one of the most innovative and most rapid- ly expanding elements of organizations such as the Latin American
Studies Association or in international journals dedicated to Latin American affairs. These scholars often do not
identify themselves as 'area studies', just as few
U.S. scholars working on U.S. politics would. As a consequence, seeking clear delimitations between 'area
studies' and 'the discipline' is becoming ever more futile. This, however, should not be seen as a problem, but rather
as a strength of area studies. An ability to bring together, on equal footing and without pre- established biases different
disciplines, different traditions of scholarship and different perspectives, united by a consciously chosen regional
focus is key.
Still, some who see the blurring lines of area studies as an inherent problem opt for seeking refuge in the so-
called systematic disciplines and declare area studies obsolete. However, the problem of doing research on 'the
others' is by no means exclusive to area studies, but also very much present (though less reflected upon) in the
disciplines. Take political science with one of its core sub-disciplines - comparative politics - carrying the notion of
comparing in its very name. In practice, however, its mainstream reproduces the traditional area studies focus on 'the
others', as Adam Przeworski points out when he describes the dominant understanding of comparative politics in the
U.S. as 'one where Americans go out and study other countries' (Przeworski 2003, p. 59). If U.S. scholars work on
elections or social movements or health policy in the U.S., they do political science or government. However, if U.S.
scholars work on elections or social movements or health policy in Brazil, they do comparative politics. Przeworski
goes on to say: 'Now, I ask myself: "What do Brazilians do when they study Brazil?"' (Przeworski, 2003, p. 59).
Rather than declaring area-specific expertise obsolete, the area-specific context of research should also be
made explicit where most of the time it is not. Studies on the U.S. political system cannot pretend to be on government
per se, but on a specific, contextual ized version of it - just as Brazil's, area studies have been said to provide 'bounded
generalizations' (Bunce, 2000), which are valid within the confines of their local or regional context; this, however, is
just as true of many political science studies, even if they treat their cases as if they were universal models.
Intra-, inter-, cross-area comparisons
This article does not claim to invent something totally new. Quite the contrary, the case for comparative area
studies can build on the impressive body of work that Latin American Studies scholars have undertaken over the past
decades. To structure the field it is useful to distinguish three types of Comparative Area Studies (Basedau & Köllner,
2007, p. 11):
o Intra-regional comparisons which compare entities within a specific area, e.g. different Latin American
countries. This can also apply to sub-national units such as provinces or cities, and, of course, it can also compare
differ- ent actors, institutions or practices within an area,
o Cross-regional comparisons, which compare cases from different world regions, such as the different
development trajectories of East Asia and Latin America (e.g. Kay, 2002) or the politics of taxation and their impact
on race in Brazil and South Africa (Liebermann, 2003).
o And finally inter-regional comparisons which take world regions as a whole as the unit of analysis and
explore the differences and commonalities be- tween them; examples are the classic comparative studies on the
transitions to democracy in Southern Europe and Latin America (O'Donnell et al., 1986) or the World Value Survey
which maps the different support for values and attitudes across world regions (Inglehart & Wenzel, 2005).
Regions, it must be noted, are no given fact but socially constituted along, but not only, geographic, historic
or cultural lines. As such, what constitutes a re- gion or sub-region and what is the appropriate term for it will always
be sub- ject to debate. While it is easy to see that the precise limitations and concepts demarcating 'Europe' or the
'Middle East' are difficult to pin down, does this also apply to the study of Latin America - or Latin America and the
Caribbe- an? South America? The Americas?
In Latin American Studies intra-regional comparisons have by far been the most common comparative
framework. In contrast, cross- and inter-regional comparisons have been much less frequent. The reasons are easy
to see. Inter- regional comparisons tend to wipe over the diversity existing within a region; they build on large-N
studies and aggregate data which have been the domain of mainstream economics and comparative politics rather
than traditional area studies. Cross-regional comparisons, too, are a difficult setting for area special- ists as few will
have the same profound knowledge of culture, language and society for more than one world region.
These problems, however, should not make Latin American Studies schol- ars desist from venturing into
comparisons beyond the region. This simply is too important. If focusing scholarship on a specific area - however that
may be defined - is to be more than just an arbitrary parcelling out the world, then the category of 'area' needs to be
substantiated. The question of the specificities of Latin America and the Caribbean then is not one amongst many,
but is at the core of legitimating the field of study as such. This, however, cannot be achieved by intra-regional studies
 7

alone but requires the confrontation with out-of-area realities through inter- or cross-regional comparisons. For this,
the often implicit assumptions underlying the area studies approach need to be made explicit.
If intra-regional studies can arrive at bounded generalizations for the validi- ty (or not) of theoretical
propositions under the scope conditions of their re- gional context, this leaves two interrelated questions: What
specifically is it that binds this regional cluster? And what happens if such a 'bounded generali- zation' is being tested
beyond its bounds? In this sense, comparative area stud- ies takes up the idea of making concepts travel - however,
freeing it from the bias of the past, when all too often conceptual travels meant a one-way road, exporting concepts
generated in the OECD countries to the rest of the world. In this sense, it is part of the challenge of comparative area
studies to overcome the hierarchic order of countries and regions underlying much of conventional scholarship.
In the past, area studies have excelled in qualitative research and small-N analyses or case studies in which
in-depth knowledge of the empirical realities is key to gain scholarly insights. In contrast, they have tended to be
sceptical about quantitative and large-N approaches, often leaving these to scholars from comparative politics or
other disciplines, criticizing their superficial knowledge of the Latin American cases and their unreflected use of
supposedly universal categories and coding which were seen as hardly fitting reality.
As Ahram (2011, p. 72-77) has shown, quantitative cross-area studies by and large fail to account for regional
variation. Surveying 741 articles from seven leading political science journals he concludes that even in the minority
of cases where 'they even bother to test for regional variation, large-N studies remain at a loss to explain it' (Ahram,
201 1, p. 77). Large-N analyses may find regional clustering of correlations, but in many cases it will be up to scholars
with area expertise to make sense of them.
In part reacting to the limitations of quantitative regression analyses, mixed method-approaches have
become popular. Approaches such as Lieberman's (2005) 'nested analysis' seek to bring the strengths of both to
bear. They can be a fruitful field for Latin American Studies scholars with a background in quali- tative empirical
research to work hand in hand with their colleagues from the large-N quantitative side.
In recent years area studies scholars have made important contributions to a research agenda that stresses
trans-national and trans-regional connections, interdependencies and entanglements, challenging the idea of the
nation-state as a 'container', to be studied separately from its environment. It would be short-sighted to see this as
contrary to comparative approaches. These in no way have to be wed to what some have called 'methodological
nationalism' (Wimmer & Glick-Schiller, 2002). Quite to the contrary: Precisely to identify the impact of international
and trans-regional relations a comparative perspec- tive can be extremely helpful, as it allows the identification of
factors that con- tribute to similar or different outcomes in the phenomena under scrutiny.
Institutional requirements for Comparative Area Studies
Cross-area comparisons can, but of course do not have to resort to quantitative methods but can also build
on the strengths of in-depth qualitative studies. Sil (2009) has eloquently made the case for the crucial role qualitative
cross- regional small-N comparisons can play in helping area specialists make the empirical and theoretical value of
their research more obvious to comparatists focused on general theories and models (Sil, 2009, p. 26).
However, the call for Latin America-focused scholars to engage more in inter- and cross-area studies runs
up against the dilemma of 'but nobody does area studies'. By and large, area studies-scholars are single-area studies
schol- ars. There is a Latin American Studies Association and similar organizations for scholarship on Africa, Asia,
the Middle East or Eastern Europe, but there is no internationally relevant 'Area Studies Association' as such.
Moreover, there are good reasons for it: for one researcher to become a specialist in many areas is a tall order. Sil' s
argument that 'there is no inherent reason, other than the pressure to publish more quickly, why one cannot patiently
increase one's fa- miliarity with cases' (Sil, 2009, p. 29) will not convince many if this implies learning Japanese,
Russian and Mandarin at the same time.
Rather than burdening the individual scholar with oversized ambitions, the key is collaboration. If it makes
sense to comparatively study, say, the impact of the Internet on state-society relations in China, Vietnam and Cuba,
then the organizational response would be for scholars with expertise in these three countries to team up in a joint
research endeavour. However, this also requires institutional facilities that can make it possible - which in turn means
overcom- ing the traditional division into single-area studies containers with little need to speak to each other.
In the past area studies found institutional answers to the challenges of mul- ti-disciplinarity. Similarly,
cultivating the institutional grounds for comparative area studies requires a wide range of activities. Universities could
create fo- rums to promote systematic intellectual exchanges between their different re- gion-focused studies centres;
area-specific graduate or post-graduate pro- grammes could include seminars that foster comparative perspectives
to other world regions; area studies associations could organize joint conferences or call for shared panels to develop
comparative area studies; journals could work together on special issues; it would be up to funding bodies to establish
 8

specific programmes that specifically seek to enable collaborative cross-area research or to honour cross-area
approaches in their evaluation criteria.
To draw on personal experience, the institution I work at is an example for the institutional transformations
that are required for a comparative area studies to take root. Ten years ago the German Institute of Global and Area
Studies, GIGA for short, had been merely loose umbrella for four rather disconnected centres on Asia, Africa, Latin
America and the Middle East, respectively. It required a major institutional overhaul to establish a structure with cross-
cutting research programmes in which the scholars from different area back- grounds continuously interact. This
provides the indispensable breeding- ground for generating collaborative research proposals or working on joint
publications that strive for comparisons beyond the confines of any one of the regional units.
As each academic is embedded as much in one of the four area institutes as in one of the four thematic
research programs, the intellectual mind-sets had to change. Scholars working on the Internet's impact on state-
society relations in China and Vietnam discovered that the Cuban experience on the matter was of much more
relevance to them than that of the regional neighbors. Colleagues from different backgrounds teamed up to see if the
analytical frame- work of neo-patrimonialism, broadly applied to Africa in the past, could be usefully applied to Latin
American cases. Others drew up a research project to study the role of supreme and constitutional courts in politics
in three South American as well as in three African countries (Llanos et al., 2015). To further promote the idea, an
international award was established which bi-annually honours articles that excel at living up to the promises of
Comparative Area Studies.2
Of course, approaches of this type face their own set of problems. There are limits on how symmetric
research designs can be when, for instance, in Argen- tina, Chile and Paraguay, a study on the role of courts builds
on a broad literature on the matter, whereas in Benin, Madagascar and Senegal much is pio- neering work, and the
researchers are often the first ever to interview the judg- es or go through the court's archives. Language issues also
matter. The domi- nance of Spanish and Portuguese makes it comparatively easy for scholars of Latin America to
look beyond single country experiences, whereas Asia's lan- guage diversity makes many scholars more strongly
attached to the countries in which the language of their expertise is spoken. For Latin America and the Caribbean a
small team can take on a continent-wide, primary sources-based analysis of, for instance, legislation on migration
policies, whereas in Asia this would require knowledge of more than a dozen languages and require a correspondingly
high staff effort.
Comparative Area Studies, to repeat, is no panacea; cross-area comparisons are not always feasible and do
not always make sense. There is no claim that Comparative Area Studies are in any way superior to other forms of
scholar- ship. But the argument in its favour is that such an approach contributes to new insights, and enables area
studies to have a stronger impact on overarching conceptual debates. Neither is Comparative Area Studies meant to
undermine Latin American Studies. Quite the contrary! By contrasting Latin American experiences with those of other
world regions it should sharpen our under- standing of what is specific about the region, and what the study of the
region can contribute to our general understanding of the workings of the world we live in. In doing so it can also
show why we should continue developing Latin American and Caribbean Studies as a distinct field of scholarly
attention which, hopefully, will have a forum like ERLACS for another 100 issues.

4. Regime Legacies and Levels of Democracy: Evidence from Latin America written by Aníbal Pérez-Liñán and Scott
Mainwaring

WEvery country in Latin America except Cuba either began 1978 as a democracy (Costa Rica, Colombia,
and Venezuela) or experienced a transition to a competi- tive regime during the third wave of democratization
that started in 1978 (sixteen countries). But by the early twenty-first century the outcomes of these transitions
varied widely. At one pole, countries such as Nicaragua, Paraguay, and Guatemala established competitive
regimes that remained fraught with limited accountability, weak judiciaries, frequent state abuses of power, and
weak protection of politi- cal rights. At the other end of this spectrum, Chile after 1990 and Uruguay after 1985
joined Costa Rica as stable, robust democracies with solid mechanisms of intrastate accountability, effective rule
of law, and solid respect for civil and political rights.
An even more dramatic dispersion of regime outcomes after an initial transition to competitive regimes
has occurred in other parts of the world in the third and fourth waves of democratization. Many transitions to
competitive regimes have failed, resulting in a burgeoning number of competitive authoritarian regimes that
sponsor controlled elections. Other transitions (Poland, Hungary, the Czech Republic) have produced robust
democracies. In light of the wide variation in regime outcomes after transitions, a new question has emerged on
 9

the political science agenda. Why have some countries blossomed into stable and robust democracies, while
other regimes are best characterized as semidemocratic or even authoritarian? This question has assumed
importance as a large number of hybrid regimes, semidemocracies, and competitive authoritarian regimes in
Latin America, Africa, Asia, and the post-Soviet region have come into existence.1
We contribute to this literature by examining the level of democracy in post- 1977 Latin America. We
show that countries with stronger histories of democracy between 1900 and 1977 are more democratic today.
The only countries that have attained a very high level of democracy in contemporary Latin America – Chile,
Costa Rica, and Uruguay - had the region's strongest democratic legacies from 1900 until 1977. Most countries
with highly authoritarian pasts have transitioned to competitive political regimes, but the level of democracy in
these regimes is much lower.
This phenomenon can be explained by regime legacies that were partly repro- duced over time through
political parties and legal institutions. This finding apparently echoes recent work on path dependence in that it
underscores the impact of long regime histories on the contemporary level of democracy. However, using the
term "path dependence" to characterize our argument excessively stretches this concept. Latin American political
regimes in many countries underwent a profound transfor- mation in the post-1977 period. Because the notion
of path dependence implies too much linearity, we instead develop the concept of regime legacies.
The Level of Democracy in Latin America Since 1978
The level or quality of democracy (these terms are hereafter used interchangeably) is a complex and
multidimensional concept.2 This article focuses on the most con- ventional dimensions, civil liberties and political
rights, and assesses the level of democracy achieved in the post-transition era using Freedom House scores.3
Although not without problems, in recent years Freedom House has had some advantages over Polity IV, and
provides more variance than dichotomous or tri- chotomous regime classifications. For present purposes, this
variance is essential, and we also employ alternative measures of democracy to assess past experiences with
democratization. The post-1977 wave of democracy profoundly transformed Latin America. For the first time
ever, almost all countries in the region have had competitive political regimes for a prolonged period. But the
level of democracy has varied widely. Table 1 provides information on the level of democracy measured by
Freedom House scores since the inauguration of competitive regimes after 1977 (and since 1978 for the three
countries that were democratic at that time). Countries enter the sample in the year a competitive political regime
was first established in the post-1977 period (Column 2). Because we focus on post-transition levels of
democracy, we do not include Cuba, the only country that did not undergo a transition after 1977. Because data
are missing for some variables, we also do not include Haiti. In the third wave of democratization, Chile, Costa
Rica, and Uruguay stand out as the Latin American countries with the highest levels of democracy. They are the
only countries in Latin America that have ever registered the highest possible Freedom House score. Nicaragua
and Guatemala anchor the other end of the spectrum with much lower mean Freedom House scores. Table 1
also reports the average Polity scores prior to the third wave of democratization, in 1900-1977. The Polity index
ranges from -10, indicating an institutionalized autocracy, to 10, indicating an institutionalized democracy
Table 1
 10

suggests an intriguing relationship between early

democratic experiences and the average level of
democracy after the most recent democratic
transitions initiated in 1978. Average Freedom
House scores for 1978-2010 correlate at .69
(p < .01) with Polity scores for 1900-1977 and
at .64 (p < .01) with earlier Polity scores for
1900-1944.
Although revealing, Table 1 does not provide conclusive evidence of long- term regime legacies for two
reasons. First, average Freedom House scores mask within-country variation in levels of democracy during the
post-1977 period. A few countries (Peru, Venezuela, and Nicaragua) exhibit pronounced shifts over time. Second
and most important, apparent regime legacies may result from long-term forces driving latent continuities at the
national level. It is possible that some stable conditions, such as enduring social cleavages or cultural traits, have
affected the level of democratization in a consistent manner over the past century. Likewise, a consistently higher
level of development, rather than legacies of the political regime, may explain why some countries have been
on average more democratic than others both since 1977 and from 1900-1977. Focusing exclusively on Latin
America rules out some likely candidates for this type of explanation (colonial legacies, religious worldviews) that
show little variance within this region.
Testing for Regime Legacies
To verify the legacy of past political regimes on the current level of democracy in Latin America, a time-
series, cross-section model is employed. Each country in a given year is one observation. Because we focus on
the third wave of democratization, Costa Rica, Colombia, and Venezuela enter the dataset in 1978 even though
they inaugurated competitive political regimes well before that (1949, 1958, and 1959, respectively). Figure 1
depicts the evolution of Freedom House scores for Latin American countries between the year of the transition
from authoritarian rule (or 1978 for Costa Rica, Colombia, and Venezuela) and 2010. Stars represent the
observed Freedom House scores, and horizontal lines indicate mean values for the period. Although some
countries have shown considerable fluctuations, differences among countries are more important than change
over time within countries for understanding the level of democ- racy in the region. Because democracy scores
are stable for many countries, entrenched country characteristics may shape the overall level of democracy in
the long run.
 11

Figure 1 suggests two related

questions. Why do some countries
enjoy high levels of democracy, on
average, while others are less
democratic? And why do countries
rise above or fall below those
historical averages during periods?
While the latter question can be answered by looking at time-varying explanatory factors (for instance, a period
of economic decline may trigger an erosion in the level of democ- racy), the former calls for the analysis of stable
country characteristics that help explain variation across countries.
Invoking country characteristics as causal factors does not mean that some countries are culturally
predetermined to be more or less democratic. Rather, it is important to identify stable conditions that affect the
level of democracy in the long run. This approach precludes the use of a cross-sectional analysis or a standard
fixed-effects model. A cross-sectional model would not properly capture the effect of the time-variant predictors,
while a fixed-effects model would not capture the effect of stable-country characteristics. Therefore, the impact
of our independent variables is estimated using a "hybrid" fixed-effects estimator developed by Paul Allison.5 To
estimate this model, we perform two tasks. First, the deviation of all time-varying predictors are computed from
their country averages (that is, we group-center or de-mean the variables). Second, a random-effects model is
estimated in which the centered variables are included in the equation along with their country- level averages
and other time-invariant predictors.6
The model therefore includes cross-section as well as time-varying covariates. Every time-varying
predictor is decomposed into two variables. The first represents the average value of the predictor for each
country, and the second represents the dif- ference between the observed value for any given country-year and
the country aver- age. While the first item captures the cross-sectional variance of the predictor, the second
reflects variance within countries over time. Both components are included in the analysis. Coefficients for
country-average variables are not relevant to assess the causal impact of the time-varying covariates, but they
minimize omitted variable bias when purely cross-sectional predictors are included such as our measures of pre-
1978 levels of democracy.7 By contrast, the coefficient for the within-country component provides a reliable
estimate of causal effects. Because the item is cen- tered at the country level, the coefficient for this variable
replicates the estimate of a fixed-effects model.8
The analysis includes a measure of regime legacies (computed according to three alternative sources)
and several additional predictors, including rarely changing and time-varying covariates. Because information
for some items was not available for recent years, the sample extends to 2004.
Regime Legacies To capture regime legacies, we created a variable for the average level of democracy for
each country between 1900 and 1977. To avoid bias related to any particular index, we employ three alternative
measures of democracy: the Polity index, the Mainwaring, Brinks, and Pérez-Liñán regime classification, and
Smith's regime classification.9 Although widely used in political science, Polity scores are of questionable validity
for many Latin American countries for the first decades of the twentieth century. For example, against the
 12

comparative historiography on the subject, average Polity scores suggest that Honduras and Cuba were more
democratic than Chile and Uruguay in the first half of the century, and Costa Rica has an untainted score of 10
throughout the twentieth century despite a military dictatorship from 1917 to 1919 and civil war in 1948. 10 Given
these shortcomings, we rely more on the other two indicators than on Polity.
Scott Mainwaring, Daniel Brinks, and Aníbal Pérez-Liñán coded all Latin American countries as
democratic, semi-democratic, or authoritarian.11 Their scale is recoded here as democracy^ 1, semi-democracy
=0.5, and authoritarian =0, so that the average value for each historical period can be roughly interpreted as the
proportion of years that a country was democratic. Peter Smith classified Latin American regimes into four
categories: democratic, semi-democratic, oligarchic, and authoritarian.12 We again gave democracies a score
of 1, dictatorships a score of 0, and semi-democracies or oligar- chic regimes a score of 0.5
Structural Conditions Four independent variables are included to capture the impact of structural forces.
Following modernization theory, per capita GDP is included in our models because of the theoretical expectation
that a higher level of development might be favorable to a higher level of democracy.13 The natural logarithm (in
con- stant 2000 dollars) accounts for nonlinearity in the effects of per capita GDP. In their class approach to
democratization, Dietrich Rueschemeyer, Evelyn Huber Stephens, and John D. Stephens argue that the working
class is the pro-democratic actor par excellence, and that a large working class is favorable to democracy.14
Accordingly, a variable that reflects the share of the economically active population in manufacturing, mining,
construction, and transportation (from the World Bank's World Development Indicators) is included as a measure
of the relative size of the industrial working class.
A third structural variable reflects the dependence of regimes on natural resources. Several scholars
have argued that countries that depend on natural resources are likely to experience vicious cycles detrimental
to democracy.15 The average proportion of the gross national income represented by exports of fuel and
minerals in any given year (obtained from the World Development Indicators) captures this effect.
The final structural covariate reflects ethnolinguistic fractionalization. Some scholars argue that ethnically
divided countries are less likely to be democratic. Accordingly, an index is included that approaches a value of
zero when a country is highly homogenous and a value of one when the country is highly fractionalized.16
Because data are not available on a yearly basis, Anthony Annett and James Fearon and David Laitin computed
this index as a time-invariant indicator.
Economic Performance Poor economic performance may undermine democ- racy,17 so we include measures
of per capita income growth (as a proportion of GDP) and inflation.18 Because poor economic performance is
unlikely to affect regime conditions in the short run but might in the medium or long run, both variables are
measured as running averages beginning with the inception of the regime, for up to ten years.
Political Conditions Some scholars argue that presidential democracy is more problematic with multipartism.19
Therefore, a dummy variable for multiparty systems is used, operationalized as an effective number of parties of
three or greater. Pre- sumably the number of parties has a nonlinear effect - the difference between two and
three parties is more relevant for the argument than the difference between seven and eight parties. Thus, we
prefer a dichotomous indicator to a continuous measure of party system fragmentation.
Several studies have shown neighborhood political effects on political regimes,20 so we also control for
this factor. To measure neighborhood effects, the average Freedom House score of a country's immediate
neighbors during the previous year is used. In addition, because global international conditions varied over time,
our models include year fixed-effects.
Country-Level Variables Two independent variables in this set (the average level of democracy from 1900 to
1977 and ethnic fractionalization) are time-invariant. All other independent variables change over time. To
estimate hybrid fixed-effect models, we include the average value for each time-varying independent variable is
used as a stable country-level control. Estimates for the mean variables are not the focus of interest, but they
are necessary for proper model specification. The upper panel in Table 2 presents the estimates of fixed effects
based on the centered, time-varying predictors, while the lower panel presents the estimates for stable country-
level vari- ables (including the country averages as controls).
Evidence of Regime Legacies
 13

Model 2.1 presents conventional fixed-effects estimates for reference. Models 2.2 to 2.4 present the
coefficients for the hybrid estimator using three different measures of regime legacies based on the Polity,
Mainwaring et al., and Smith indicators of democracy (shown at the bottom of the table). Even after controlling
for a large number of alternative explanations, an early history of democracy has a powerful impact on levels of
democracy among contemporary competitive regimes. The effect is powerful both statistically and substantively,
and the results for these variables are consistent across models.
In model 2.2, an increase of 1 point in the 21 -point Polity scale for 1900-1977 yields an increase of 0.21
points in the 13-point Freedom House scale for the post-1977 period. In model 2.3, a unit increase in the
Mainwaring et al. classification for 1900-1977 (that is, a change in conditions from a country that was always
authori- tarian between 1900 and 1977 to one that was always democratic) predicts a substan- tial increase of
4.4 points on the inverted Freedom House scale for contemporary competitive regimes. In model 2.4, using
Smith's historical classification of political regimes, a country that was consistently democratic from 1900 to 1977
would have a predicted
Freedom House score
4.6 points higher for the
post-1977 period than

a country that was

consistently
authoritarian. The impact
of past democracy on the
current level of democracy is similar to the result obtained by Michael Bratton and Nicolas van de Walle for forty-
seven African countries; a past history of more elec- tions and more electoral participation was favorable to a
higher level of democracy in Africa in 1994.21
Other variables also have effects in the expected direction. The positive coefficients for the level of
economic development in the top panel of Table 2 represent fixed- effects estimates of the impact of per capita
GDP within countries. The positive coef- ficients for GDP in the bottom panel indicate that countries with a higher
average per capita income also had higher average levels of democracy.22 For Latin America over a longer time
period, a higher level of development has not reduced the likelihood of democratic breakdowns or increased the
likelihood of transitions,23 but it has favored a higher level of democracy among competitive regimes in the post-
1977 period. 38
The results in the upper panel of Table 2 indicate that an increase in the share of the labor force in
manufacturing makes a moderate but statistically positive contribu- tion to the level of democracy.24 Reliance
on fuel and mineral exports lowers the level of democracy, consistent with the idea of a natural resource curse.
Ethnic fractionali- zation, a time-invariant predictor (listed in the lower panel of Table 2), has an insig- nificant
effect on the level of democracy.
A higher rate of economic growth supports the expansion of democracy (the large coefficients reflect the
fact that growth was measured as a proportion of GDP), while inflation fails to achieve statistical significance.
The indicator for multi- partism is also insignificant, but because this variable seldom changes within the same
country over time, the positive and significant coefficient in the lower panel of Table 2 suggests that countries
 14

with multiparty systems were likely to establish higher levels of democracy. The only variable presenting an
unexpected effect is the influence of democratic neighbors. Increases in the democracy scores of neighbors
seem to foster a decline in a given country's predicted level of democracy. Although democratic neighbors help
promote transitions to democracy, countries follow independent and often contra- dictory trajectories after their
transitions take place.
The estimates in Table 2 suggest that regime legacies are one of the main factors that explain the quality
of democracy in the post- 1977 period. An authoritarian past did not prevent Latin American countries from
developing competitive political regimes in the post-1977 period, nor did it lead to full breakdowns of these com-
petitive regimes. But an authoritarian past did tend to limit the quality of democracy. By contrast, countries with
a past democratic heritage had a significant advantage in building a high-quality democracy in contemporary
Latin America. This is true even when, as in Chile (1973-1990) and Uruguay (1973-1984), military dictatorships
attempted to radically stamp out the democratic past.
Why Regime Legacies Have a Long-Term Impact
In short, a high level of democracy in the past predicts a high post-transition level of democracy in the
post- 1977 period. The statistical results, however, do not explain the causal mechanism that lies behind the
impact of regime heritage on the contemporary level of democracy. Our finding about the enduring impact of
early democratization has some similarities to arguments about path dependence in social science.25 Margaret
Levi defines path dependence as meaning that "once a country or region has started down a track, the costs of
reversal are very high." Events in one historical moment greatly alter the distribution of possible and probable
outcomes into the medium- and/or long-term future.26
The statistical results here tell a similar story. The early history of political regimes affects the current
level of democracy even controlling for a wide range of other variables. Two countries similar on all of the other
independent variables would have different predicted levels of democracy today if one had a past con- siderably
more democratic than the other. However, a general claim about path dependence does not indicate how regime
legacies are reproduced over time.
For Latin America, two problems undermine strong claims about path depen- dence. First, authoritarian
disruptions in countries with long democratic traditions, such as the ones occurring in Chile (1973-1990) and
Uruguay (1973-1984), defy any idea of "increasing returns" under democratic rule. Likewise, in the post-1977
period, several countries (Bolivia, the Dominican Republic, El Salvador, Guatemala, Mexico, Nicaragua, and
Paraguay) shifted from almost uninterrupted histories of authoritarian rule to somewhat durable competitive
regimes. Path dependence cannot explain so many radical departures from the past. Second, even if it is
assumed that disruptions of long-established democratic regimes (again, Chile and Uruguay) were driven by
exogenous causes (for example, the Cold War), fairly lengthy dictatorships should have depleted the pool of
democratic leaders, not only because authoritarian rulers repressed them but also because aging individuals
retired from politics or died over time.27 Any explanation of regime legacies can be logically sound and histori-
cally credible only if we identify an intertemporal bridging mechanism that accounts for how legacies of an early
but distant era carried their influence into the present despite an interlude of authoritarian rule. Without such a
mechanism, the concept of path dependence is too vague to explain regime legacies
Institutional Mechanisms
In some cases, the intertemporal bridging mechanism is given by the survival of indi- vidual leaders.
Patricio Aylwin and Julio Maria Sanguinetti were up-and-coming party leaders in Chile and Uruguay, respectively,
before military coups imposed repressive dictatorships in 1973, and they returned as inaugural presidents after
the transition took place in each country.28 But the survival of individuals cannot fully explain the long-lasting
legacies of the first wave of democratization depicted in Table 1.
Although different explanations are possible, some of which are addressed below, we hypothesize that
an early history of democracy favored the building of formal institutions such as party systems, courts, and other
agents of intrastate accountability that are favorable to a higher level of democracy in the contemporary period.
Our empirical test of this proposition focuses on two institutions: political parties and the legal system. If our
understanding of regime legacies is correct, well-established parties will favor a higher level of democracy only
if these parties functioned under democratic regimes. A party that was institutionalized under democ- racy should
 15

be an asset for a higher level of democracy. Most parties created under democracy generate interests (among
politicians, citizens, and organized groups) and identities connected to democracy; they socialize new
generations in those interests and identities, and preserve them over time. In contrast, an institutionalized gov-
erning party that sustained an authoritarian regime (such as the PRI in Mexico or the Colorados in Paraguay)
will not. Parties are not the only mechanism for sustain- ing a high-quality democracy, but they are one of the
main institutional carriers of democratic norms and practices.
A strong judicial system is important for protecting and promoting citizen rights, for generating
accountability, and for bolstering the quality of democracy. Presumably, under democracy, a stable court system
with long judicial careers fosters a stronger, more independent judiciary. If presidents tether the judicial system
to their own preferences and often dismiss judges, the court system will not act as a mechanism of accountability,
and it is less likely to be a staunch defender of citizen rights. Accordingly, a stable court system will be a pillar of
democracy if judges are selected under democratically elected presidents and are allowed to serve for a long
time. Even if those judges are severely constrained during authoritarian periods, their views of the law and their
rulings will reemerge once democracy is established. To the extent that legal practices and procedures are
reproduced over time, new judges will be socialized to uphold such democratic principles. However, the role of
courts as bridg- ing institutions of accountability can be undermined in two ways. If courts are reshuffled often
and judges have short or unstable careers, they will not preserve long-term democratic legacies. In addition, if
early judges were appointed under authoritarianism, stable courts will likely reproduce legal practices and
procedures, but ones inimical to creating a high level of democracy.29
To capture the institutionalization of competitive parties during the twentieth century, we created a new
indicator of democratic party system institutionalization. We collected data on a) the percentage of seats in the
lower chamber by party, b) the year when each party was created, and c) the number of years each party had
lived under a competitive political regime since 1900. This variable reflects the average age of the parties in
Congress, weighted by their seat share and by their experience with democracy since 1900. We count only years
that the party existed under a com- petitive (democratic or semidemocratic) regime. The formula for the index is

where DPIit is the democratic party institutionalization score for country i in year
t; Sj is the share of seats of the j-th party in the lower house; and a]t is an age function that reflects the number
of years that the j-th party has existed under a competitive regime between the time of its founding (or 1900 if
the party was founded in the nineteenth century) and year t. For example, if party j has existed for a hundred
years but the country experienced democracy or semidemocracy only for two decades during this period, a}X =
20. Age is presumably a nonlinear indicator of party institution- alization; the gap between a hypothetical party
system that is just one year old and another that is thirty years old is much greater than the gap between a party
system that is seventy-one years old and another that is a hundred years old. Therefore, we took the square root
of the regime-age function.30 The resulting indicator can be interpreted as weighted measure of the length of
the democratic experience of the political parties in the lower chamber, given the composition of the chamber in
any given year between 1978 and 2004.
We constructed a similar index to assess democratic institutionalization in the judiciary by focusing on
each country's Supreme Court. We expanded the dataset from Anibal Pérez-Liñán and Andrea Castagnola to
identify when different Supreme Court justices of eighteen Latin American countries served, and whether they
were appointed under a competitive regime or a dictator

where DCIit is the democratic court institutionalization score for country i; nt

is the number of judges sitting in the supreme court in year t; and d]X is a duration function that reflects the
number of years that a justice appointed by a competitive regime has been in office. If a justice j was appointed
by an authoritarian regime, the duration term is treated as zero. This index displays greater values when the
 16

average justice in the supreme court was appointed by a democratic regime and served for a long period. The
value of this index is particularly relevant after an episode of redemocra- tization because it shows whether
justices appointed before the authoritarian takeover preserved their seats until democracy was restored.32
Although this empirical indicator for democratic court institutionalization is limited to the supreme court, it
probably serves as a proxy for a broader measure of the degree to which the justice system is appointed by
democratic leaders and to which justices enjoy stability under democ- racy. We again take the square root of the
duration term to preserve a common metric for both indices of democratic institutionalization.
Empirical Evidence
We test the impact of these institutional mechanisms by adding our indicators of democratic party and court
institutionalization to the empirical models of regime lega- cies presented in Table 2. If our hypothesis is correct
and regime legacies are pre- served by institutional carriers, the new predictors should be statistically significant
in the hypothesized direction, and their inclusion should reduce the coefficients for measures of past democracy.
At least two alternative hypotheses may help explain why regime legacies from the first three-quarters of
the twentieth century affect contemporary levels of democ- racy. First, in earlier periods of democracy, elites may
have learned that democracy was not harmful to their interests, and hence they more readily tolerate a high level
of democracy in the contemporary period. Many scholars have emphasized elite willing- ness to accept
democracy as crucial to its viability.33 However, several facts weaken the plausibility of this alternative
hypothesis. In Chile, during the presideSalvador Allende (1970-1973), the right and Christian Democrats learned
that democracy could be harmful to their interests and supported a military coup. In a similar way, authoritarian
attitudes among the political elite do not consistently endure. Leaders of the COPEI party in Venezuela supported
a coup in 1948, but ten years later they embraced democracy.34 The main conservative party in Brazil between
1946 and 1964, the UDN, frequently conspired against democracy, but by the end of the 1990s conservative
parties and individuals had fully accepted democracy.35
A second alternative explanation is that a more democratic and tolerant mass political culture emerged
in the early democratizing countries. In this perspective, a supportive democratic political culture fosters a higher
level of democracy.36 However, there are reasons to doubt that mass support for democracy was relatively
steady in most countries during the twentieth century. Political polarization in the 1960s and the 1970s affected
not only the elites, but also common citizens.
There is no definitive way to test these alternative hypotheses given the absence of comparative surveys
for elites and mass publics going back to the early twentieth century. However, if a stronger democratic culture
among elites (or masses) is the main mechanism through which regime legacies affect current levels of
democracy, contemporary support for democracy among elites (masses) should be stronger in countries with
past histories of democracy, and it should have a positive effect on democratization in the contemporary period.
We use information from similar questionnaire items employed by the Survey of Latin American
Parliamentary Elites (PELA, its Spanish acronym) and by national public opinion surveys conducted by
Latinobarómetro.37 Both projects asked respondents whether they agreed with the statement, "Democracy is
prefer- able to any other form of government" (as opposed to a statement indicating con- ditional support for
authoritarian rule). The indicator of elite support for democracy is the average percentage of legislators who
agreed with the statement in each country between 1995 and 2005. The indicator of mass attitudes is the average
percentage of respondents who agreed with the statement in each country between 1995 and 2006. To address
gaps in the data, as well as lack of information for 1978-1994, national averages for both variables are treated
as country-level char- acteristics. Averaging at the country level is sensible because these control variables seek
to capture a latent cultural trait developed before 1978, not the effect of specific fluctuations in the current era.
Table 3 includes the new covariates designed to capture specific causal mecha- nisms in addition to the
general measure of regime legacies for 1900-1977. Model 3.1 reports the results of a conventional fixed effects
model for reference, and thus omits the country-level variables. Models 3.2 through 3.4 present the result of
Allison's hybrid estimator using the three different measures of democracy for 1900-1977.
The indicators of democratic party and court institutionalization present positive and significant effects on
current levels of democracy in all models. By contrast, the proxies for mass and elite political culture fail to
 17

achieve conventional levels of sig- nificance. Given the limited nature of these proxies, we do not infer much
from

results, but the evidence supports our hypothesis about institutional carriers irrespec- tive of whether cultural
mechanisms have a long term impact on democracy.
The coefficients for the measure of regime legacies decline by 12 percent in Model 3.2 (compared to
2.2), by 29 percent in Model 3.3 (compared to 2.3), and by 22 percent in Model 3.4 (compared to 2.4). Moreover,
the coefficient for the pre- vious history of democracy is not significant in the last two models. The reduction in
the size of the effects hints that the indicators of democratic institutionalization may capture part of the process
that creates regime legacies. The coefficient remains positive and significant in Model 3.2, so other unspecified
mechanisms may operate as well. After we control for the new variables, the coefficients for fuel and mineral
exports and the size of the industrial labor force lose statistical significance, qualifying some of the results in
Table 2.
Regime Legacies, Path Dependence, and Democratization
Notwithstanding some important similarities between path dependence and regime legacy arguments,
we distinguish between the two. Although the existing literature does not agree on how broadly or narrowly it
defines path dependence, the concept should be bounded in such a way as to imply greater stability and linearity
than is the case with Latin American political regimes. The stunning and unpredicted trans- formations of many
political regimes in Latin America are inconsistent with the more bounded concept of path dependence advocated
by Paul Pierson, Margaret Levi, and Douglass North. In the more bounded conception, path dependence means
that switching courses is costly and relatively unlikely. In contrast to this emphasis on the low probability of
dramatic shifts in course, many Latin American countries have radically broken from their past political regimes
in the post-1977 period. Switching from authoritarianism to democracy and vice versa has been common in Latin
America. Long-standing democracies fell in Uruguay and Chile in 1973, and several countries (for example, El
Salvador, Guatemala, Nicaragua, and Paraguay) that had almost always had authoritarian regimes until the third
wave have had durable competitive regimes in recent decades.
 18

It is best to limit the notion of path dependence to contexts that create greater inertia and stability, and
less radical departures from the past than many political regimes in Latin American have experienced. In post-
1977 Latin America, regime legacies have significantly affected the level of democracy, but many countries have
established and preserved competitive political regimes despite an authoritarian past. The history of political
regimes in Latin America is inconsistent with the argument that path-dependent arguments are relatively
deterministic. A regime legacy argument is a less-bounded, less-deterministic version of path dependence, with
less emphasis on the improbability of profound shifts in path and greater emphasis on the probability of
recovering an early democratic trajectory.
A regime heritage argument must explain why past legacies shape the current level of democracy.
Moreover, it must explain how distant experiences with com- petitive politics have such effects even in countries
where an extended period of authoritarianism separated two democratic eras. Our analysis suggests lasting
effects through political parties and legal institutions.
In well-established democratic regimes, parties and heads of government who are usually recruited
through parties are the most powerful actors. They often have a strong interest in restoring or preserving
democracy, and they are often the most important actors pushing for an expansion of democracy. Parties
socialize their mem- bers in particular values, policy preferences, and tactics that are preserved across
generations with a certain probability. Yet parties are not static; ideas may evolve over time, organizations may
be converted to perform new functions, and extra- ordinary events may reconstitute the institutional landscape.38
The justice system also operates as an institutional carrier of regime legacies. Throughout the twentieth
century, presidents were unlikely to reshuffle supreme courts in countries with strong democratic traditions, such
as Chile, Costa Rica, and Uruguay. Today, according to the World Bank Governance Indicators, these three
countries have by far the best perceptions for rule of law and control of corruption in the region, both of which
are important to building high-quality democracy. Con- versely, justice systems that historically were bulwarks of
authoritarian regimes have been obstacles to creating high-level democracies in the third wave. Given a past of
either unstable and shackled judiciaries or judiciaries appointed under authoritarian rule, it has been difficult to
build a justice system that helps generate strong demo- cratic rule of law, promotes and protects citizen rights,
and serves as an effective mechanism of democratic accountability.
Dictatorships in Chile (1973-1990) and in Uruguay (1973-1984) crushed the capacity of the court system
to stand up for democracy. Even so, new democratic regimes after 1990 and 1984, respectively, drew on the
tradition of a solid court system. In part because many judges from the earlier democratic regimes remained in
the court system, the current democratic regimes in both countries relatively quickly rebuilt a judicial system that
could sustain a democratic rule of law.39
In sum, parties develop interests, norms, and preferences that typically favor some continuity in regime
legacies. They have a reservoir of inherited interests, normative principles, policy preferences, and operational
rules - an institutional "common sense" - that provides a historical underpinning to their strategic considerations.
Like- wise, the probability that courts will sustain higher levels of democracy is greater if justices were appointed
under democratic presidents and are not highly vulnerable to dismissal. For both reasons, the cumulative
experience of past generations affects the level of democracy in contemporary political regimes. The level of
Latin American democracies after 1977 has offered a prime example of this causal mechanism.

5. Key Terms and Points: (Review of your Fundamentals in Political Science)

FIDP
Social Science
Nation-State
Totalitarian
Democratic
Democratization
Government- it is a political institution.
Governance
 19

Political Behavior
Ideology
Policy
Communism

a. Type of States ([Link]

• State- vary based on who holds power, who elects the empowered, and how authority is maintained.
-Political Community in a specific geographical location under a government, control the affair (notes)
- States are not necessarily the same as nations. New state spaces are redefining borders, and they may not be ruled
by national governments.
• Authoritarianism- governments differ in who holds power and in how much control they assume over those that they
govern, but all are marked by the fact that the empowered are unelected individuals. Ex. Monarch
• Monarchy- form of government in which supreme power is absolutely or nominally lodged with an individual, who is the
head of state, often for life or until abdication. The person who heads a monarchy is called a monarch. Hereditary rule is
often a common characteristic, but elective monarchies are also considered monarchies (e.g., The Pope) and some states
have hereditary rulers, but are considered republics (e.g., the Dutch Republic). Currently, 44 nations in the world have
monarchs as heads of state.
• Democracy- Democracy is a form of government in which sovereignty is held by the majority of citizens within a country or
a state.
- form of government in which the right to govern is held by the majority of citizens within a country or a state. All
members of society have equal access to power and all members enjoy universally recognized freedoms and
liberties.
• oligarchy: A government run by only a few, often the wealthy
- form of government in which power rests with a small elite segment of society distinguished by royalty, wealth,
family, military, or religious hegemony. Ex. theocracy.
• Democratic presidential republic: A system of government where an executive branch is led by a president who head of state
and head of government; in such a system, this branch exists separately from the legislature, to which it is not responsible
and which it cannot, in normal circumstances, dismiss.
• theocracy: Government under the control of a Church or state-sponsored religion. “by the grace of God.”
- a state is governed by immediate divine guidance or by officials who are regarded as divinely guided such a god or
deity.
• dictatorship- govern without consent of the people. “a system that does not adhere to democracy”
- -concerns the source of the governing power (where the power comes from, the people or a single leader)
- A type of government where absolute sovereignty is allotted to an individual or a small clique.
• totalitarianism- the power to govern extends to all aspects of life in nearly every aspect of public and private life.
-concerns the scope of the governing power (what is the government and how extensive is its power).
-A system of government in which the people have virtually no authority and the state wields absolute
control, for example, a dictatorship
• diarchy: A form of government where power is shared between two joint authorities.
• hereditary rule: Hereditary rule is a form of government in which all the rulers come from the same family, and the crown is
passed down from one member to another member of the family.
• elite: A special group or social class of people which have a superior intellectual, social or economic status as, the elite of
society.
• Tyranny- use of violence
• De jure- In de jure oligarchies, an elite group is given power by the law. (Oligarchies)
• De facto- In de facto oligarchies, those with more resources are able to gain political power, despite laws that ostensibly treat
all citizens equally. (Oligarchy)
• corporatocracy—a system in which power effectively rests with a small, elite group of inside individuals, sometimes from a
small group of educational institutions, or influential economic entities or devices, such as banks, commercial entities,
 20

lobbyists that act in complicity with, or at the whim of the oligarchy, often with little or no regard for constitutionally
protected prerogative. W influence to elected officials.
• representative democracy: A policy under the rule of people acting on the behalf of and, to a lesser extent, in the interests of
the voting blocks by which they were elected.
• separation of powers: A theoretical model for governance, common in democratic states, which features the division of
sovereign power into at least three (but sometimes up to six) organs of state in order to forestall tyranny, by preventing the
acquisition of a monopoly of power by a monarch or oligarchy; also, such an arrangement.
• tyranny of the majority: A situation in which a government or other authority democratically supported by a majority of its
subjects makes policies or takes actions benefiting that majority, without regard for the rights or welfare of the rest of its
subjects.
• Political borders: A border is a geographical boundary of political entities or legal jurisdictions.
• European Union: A supranational organization created in the 1950s to bring the nations of Europe into closer economic and
political connection. At the beginning of 2007, 27 member nations
• Global city: A global city (also called world city or sometimes alpha city or world center) is a city generally considered to be an
important node in the global economic system.

b. Forms of Government ([Link]

• government- means by which state policy is enforced, as well as the mechanism for determining the policy of the state.
: The body with the power to make and/or enforce laws to control a country, land area, people or organization.
• State-: A political division of a federation retaining a degree of autonomy, for example one of the fifty United States. See also
Province.
• Aristarchy- ruled by the "best" people.
• Aristocracy- rule by elite citizens, noble birth
• Meritocracy- rule by the meritorious; a system of governance where groups are selected on the basis of people's ability,
knowledge in a given area, and contributions to society.
- Meritocracy, in an administrative sense, is a system of government or administration wherein
appointments and responsibilities are objectively assigned to individuals based other upon their "merits"
and achievements.
• Technocracy- refers to rule by the educated; a system of governance where people who are skilled or proficient govern in
their respective areas of expertise in technology would be in control of all decision making.
• Junta- The ruling council of a military dictatorship.
• Autocracy- ruled by one person who has all the power over the people in a country. Examples include authoritarian,
totalitarian and fascist governments.
• Democracy- direct, representative, plain
- all of the people in a country can vote during elections for representatives or political parties that they
prefer
• Monarchy- ruled by a king or a queen who inherits their position from their family, which is often called the “royal family”.
• Oligarchy- ruled by a small group of powerful and/or influential people. These people may spread power equally or not
equally.
• Authoritarianism- social and economic institutions exist free from governmental control
- form of social organization characterized by submission to authority as well as the administration of said
authority. highly concentrated and centralized power maintained by political repression and the exclusion
of potential challengers.
• Totalitarianism- an extreme version of authoritarianism. “"indefinite political tenure"
• Dictatorship- autocratic form of absolute rule by leadership unrestricted by law, constitutions, or other social and political
factors within the state.
• Plutocracy- defines a society or a system ruled and dominated by the small minority of the wealthiest citizens. not rooted in
an established political philosophy and has no formal advocates.
• oligarchy: A government run by only a few, often the wealthy
 21

• Democratic presidential republic: A system of government where an executive branch is led by a president who serves as
both head of state and head of government; in such a system, this branch exists separately from the legislature, to which it is
not responsible and which it cannot, in normal circumstances, dismiss.
• theocracy: Government under the control of a Church or state-sponsored religion
• diarchy: A form of government where power is shared between two joint authorities.
• hereditary rule: Hereditary rule is a form of government in which all the rulers come from the same family, and
the crown is passed down from one member to another member of the family.
• totalitarianism: A system of government in which the people have virtually no authority and the state wields
absolute control, for example, a dictatorship.
• dictatorship: A type of government where absolute sovereignty is allotted to an individual or a small clique.
• representative democracy: A policy under the rule of people acting on the behalf of and, to a lesser extent, in
the interests of the voting blocks by which they were elected.
• separation of powers: A theoretical model for governance, common in democratic states, which features the
division of sovereign power into at least three (but sometimes up to six) organs of state in order to forestall
tyranny, by preventing the acquisition of a monopoly of power by a monarch or oligarchy; also, such an
arrangement.
• tyranny of the majority: A situation in which a government or other authority democratically supported by a
majority of its subjects makes policies or takes actions benefiting that majority, without regard for the rights or
welfare of the rest of its subjects.
• Electoral Democracies: This map depicts electoral democracies around the world, as judged by Freedom House
in 2006. Different colors indicate different forms of democracy. Orange countries are parliamentary republics.
Green, yellow, and blue are presidential republics with less (green) or more (blue) presidential power. Red are
parliamentary constitutional monarchies in which the monarch does not personally exercise power.
• European Union: A supranational organization created in the 1950s to bring the nations of Europe into closer
economic and political connection. At the beginning of 2007, 27 member nations were Austria, Belgium, Bulgaria,
Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia,
Lithuania, Luxembourg, Malta, The Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden,
United Kingdom.
• Global city: A global city (also called world city or sometimes alpha city or world center) is a city generally
considered to be an important node in the global economic system.
• Political borders: A border is a geographical boundary of political entities or legal jurisdictions.

c. Democracy ([Link]

• egalitarian: Characterized by social equality and equal rights for all people.
• civil society: All of the institutions, voluntary organizations, and corporate bodies that are less than the state but greater than
the family.
• Parliament: A democratic government's [Link]
• Demarchy: Demarchy (or lottocracy) is a form of government in which the state is governed by randomly selected decision
makers who have been selected by sortition (lot) from a broadly inclusive pool of eligible citizens.
• Occupy movement: The Occupy movement is an international protest movement against social and economic inequality; its
primary goal is to to make economic structure and power relations in society more fair.
• deliberative democracy: Deliberative democracy or discursive democracy is a form of democracy in which deliberation is
central to decision making. Deliberative democracy differs from traditional democratic theory in that authentic deliberation,
not mere voting, is the primary source of legitimacy for the lawmaking processes.
• Enlightenment: A 17th and 18th-century philosophical movement in European history; the Age of Enlightenment
or Age of Reason emphasizing rationalism.
• constitutional monarchy: A monarchy in which the monarch's power is limited by a written constitution.

• absolute monarchy: A state over which a sole monarch has absolute and unlimited power.
 22

• the rule of law: The rule of law is a legal maxim whereby governmental decisions are made by applying known
legal principles.
• Enlightenment: A 17th and 18th-century philosophical movement in European history; the Age of Enlightenment
or Age of Reason emphasizing rationalism.
• liberalism: Any political movement founded on the autonomy and personal freedom of the individual, progress
and reform, and government by law with the consent of the governed.
• Electoral Democracies: Countries highlighted in blue are designated "electoral democracies" in Freedom House's
2010 survey Freedom in the World.

• legislative power: Legislative power refers to the power of a legislature, or deliberative assembly to pass, amend
and repeal laws.
• electoral college: The Electoral College consists of individual state appointed electors who formally elect the
President and Vice President of the United States.
• suffrage: The right or chance to vote, express an opinion or participate in a decision.

• Equal Rights Amendment: The Equal Rights Amendment (ERA) was a proposed amendment to the United States
Constitution designed to guarantee equal rights for women.
• Nineteenth Amendment: The amendment to the United States Constitution, passed in 1920, that gave women
the right to vote.

• democracy: a system of rule by the people, especially as a form of government; either directly or through elected
representatives
• autocracy: A form of government in which unlimited power is held by a single individual.
• deliberative democracy: Deliberative democracy or discursive democracy is a form of democracy in which
deliberation is central to decision making. Deliberative democracy differs from traditional democratic theory in
that authentic deliberation, not mere voting, is the primary source of legitimacy for the lawmaking processes.
• direct democracy: Direct democracy (or pure democracy) is a form of government in which people vote on policy
initiatives directly, as opposed to a representative democracy in which people vote for representatives who then
vote on policy initiatives.

• Aristotle: Aristotle was one of the first theorists of democracy.

• Sphere of Public Authority: The Sphere of Public Authority is that of the state, the realm of the police, and the
ruling class.
• Public sphere: The public sphere is an area in social life where individuals can come together to freely discuss and
identify societal problems, and through that discussion influence political action. It mediates between the private
sphere and the Sphere of Public Authority.
• Third sector: The voluntary sector or community sector (also non-profit sector or "not-for-profit" sector) is the
sphere of social activity undertaken by organizations that are for non-profit and non-governmental. This sector is
also called the third sector, in reference to the public sector and the private sector. Civic sector is another term for
the sector, emphasizing the sector's relationship to civil society.
• Participatory democracy emphasized the broad participation of constituents in the direction and operation of
political systems.

• An absolute monarchy refers to when the monarch has no or few legal restraints in state and political matters.
• In a constitutional monarchy, the monarch retains a unique legal and ceremonial role.
• Liberal democracy requires universal suffrage, competitive politics, and the rule of law and is currently the
dominant world political ideology.
• Minimalist Democracy system of government in which citizens give teams of political leaders the right to rule in
periodic elections. citizens cannot and should not rule because, for example, on most issues, most of the time,
they have no clear views or their views are not well-founded
 23

• Direct Democracy on the other hand, holds that citizens should participate directly in making laws and policies,
and not do so through their representatives. Proponents of direct democracy offer varied reasons to support this
view, declaring that political activity can be valuable in itself, since it socializes and educates citizens, and popular
participation can check powerful elites. Most importantly, according to this theory, citizens do not really rule
themselves unless they directly decide laws and policies for themselves.
• Deliberative Democracy is based on the notion that democracy is government by discussion. Deliberative
democrats contend that laws and policies should be based upon reasons that all citizens can accept. The political
arena should be one in which leaders and citizens make arguments, listen, and change their minds.
• Radical Democracy is based on the idea that there are hierarchical and oppressive power relations that exist in
society. Democracy's role is to make visible and challenge those relations by allowing for difference, dissent, and
antagonisms in the decision-making processes.
• Civil society is the arena outside of the family, the state, and the market where people associate to advance
common interests.
• The public sphere is an area in social life where individuals can come together to freely discuss and identify
societal problems, and through that discussion influence political action. “discursive space” “a theater in pol
participation is enacted”