-
Cleavage Structures and Party Systems in Latin AmericaAuthor(s):
Robert H. DixReviewed work(s):Source: Comparative Politics, Vol.
22, No. 1 (Oct., 1989), pp. 23-37Published by: Ph.D. Program in
Political Science of the City University of New YorkStable URL:
http://www.jstor.org/stable/422320 .Accessed: 01/10/2012 12:05
Your use of the JSTOR archive indicates your acceptance of the
Terms & Conditions of Use, available at
.http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars,
researchers, and students discover, use, and build upon a wide
range ofcontent in a trusted digital archive. We use information
technology and tools to increase productivity and facilitate new
formsof scholarship. For more information about JSTOR, please
contact [email protected].
.
Ph.D. Program in Political Science of the City University of New
York is collaborating with JSTOR to digitize,preserve and extend
access to Comparative Politics.
http://www.jstor.org
-
Cleavage Structures and Party Systems in Latin America
Robert H. Dix
In a seminal article in 1967 Seymour Martin Lipset and Stein
Rokkan posed a series of central questions for the comparative
study of party systems. The first set of questions concerned the
genesis of the system of cleavages within the national community,
including the timing of their appearance and their relative
salience and durability. A second group of questions focused on the
translation of cleavages into stable party systems, including the
question of why conflicting interests and ideologies in some cases
favored the emergence of broad aggregative coalitions, and in
others fragmentation. The final set of questions bore on the
behavior of voters within the various party systems. What were the
characteristics of those voters mobilized by the several parties,
and how did economic and social change translate into changes in
the strengths and strategies of the parties? The authors stressed
that all these and related questions were to be addressed
diachronically, that is, in historical perspective.'
While Lipset and Rokkan, as well as most of the many others who
have asked similar comparative questions, have focused almost
exclusively on the competitive party systems of Europe and the
Anglo-Saxon diaspora (the United States, Australia, Canada, and New
Zealand), it seems high time that questions like those raised for
industrialized countries now also be posed for Latin America,
particularly since Latin America constitutes the area of the world
that most closely approximates the developed West in culture,
levels of economic and social development,2 and experience with
competitive party systems. Only by examining such questions outside
the regions of the ancestral homes of political parties and party
systems can we expand our generalizations about the historical
development of political parties beyond the evidence of a
particular time and place. It is also at least highly plausible
that Latin America's experience with the construction of systems of
competitive party politics will prove more relevant to the future
trajectory of such politics in other parts of the so-called Third
World than will that of the developed West.
This article is an attempt to begin the systematic analysis of
that experience.3 Among the questions we pose will be the
following. Has the development of western party systems proven to
be the prototype for the evolution of competitive party systems in
Latin America? What are the kinds of parties and the patterns of
competition among parties in Latin America, and how have they
emerged over time? Have the West's past experiences with the onset
of mass politics and the politics of industrialization been more or
less replicated in contemporary Latin America? How might one
account for any differences? What follows is therefore meant
essentially as an exploratory exercise in delineating some broad
patterns of similarity and difference between the party systems of
Latin America and the developed West.
At the same time, our enterprise will be a good deal more modest
in scale and in
23
-
Comparative Politics October 1989
supportive detail than that undertaken by Lipset and Rokkan. In
part this is a function of the relative paucity or unevenness of
the kinds of reliable electoral data, opinion surveys, and single
country studies concerning Latin America in comparison to what is
available for the so-called western or industrialized countries.
Too, the electoral process in Latin America frequently suffers from
constraints that hamper analysis. Parties (sometimes major ones as,
at times, the American Popular Revolutionary Alliance, APRA, in
Peru) may be barred from presenting candidates, or fraud and other
controls may obscure fully accurate results, as in Mexico, not to
mention Paraguay. The democratic experience has also been briefer,
more recent, and more sporadic in the Latin American case and has
often been interrupted by periods of military and other
authoritarian rule that have effectively suspended competitive
politics altogether. Moreover, Latin America's parties may come and
go with startling rapidity and may form ever-changing alliances or
combinations of sometimes confusing complexity. Some have barely
deserved the designation "party" to begin with. Finally, and
perhaps most fundamental, parties by no means encompass the full
spectrum of groups competing for governmental power. In many Latin
American countries the armed forces or guerrilla insurgencies, on
occasion allied with one or another political party or even a
foreign country, employ armed force to compete for power,
necessarily making elections less definitive than has usually been
true of western Europe, North America, and Australasia.4
Nonetheless, it is our purpose to expand the comparative
horizons of the study of party systems by incorporating the Latin
American experience, particularly in regard to the development of
those systems over time and the impact on them of the onset of mass
politics.
Patterns of Party Development
At first glance the historic cleavage lines of Latin American
politics would appear roughly to parallel those of the European
past, albeit with notable time lags: the center versus the
periphery, the secularizing state versus the church, the landed
elite versus commercial and industrial interests, and finally, in
the wake of all the others, the class struggle of workers against
their employers.5 Thus throughout most of the nineteenth century,
and well into the twentieth in many cases, the political divisions
of Latin America tended predominantly to be those of conservatives
versus liberals, although they bore other names in some places and
almost everywhere showed a marked propensity for factionalism and
fragmentation, often centered around particular individuals,
families, or regions. The conservative parties tended to reflect
the interests and attitudes of those who favored strong central
government, protection of the Catholic church and its social and
economic prerogatives, and defense of the interests of traditional
landowners. Liberals, on the other hand, could usually be found
advocating federalism, disestablishment of the church, and the
defense of commercial interests, often including the advocacy of
free trade.6
One contrast to the European pattern was that the ethnic,
cultural, and interreligious dimensions of politics in much of the
West were largely absent in the southern Americas. Thus, while
center-periphery struggles led in some places (Argentina and
Colombia, for example) to almost endemic civil war for much of the
nineteenth century, they did not entail
24
-
Robert H. Dix
struggles between national and provincial or subnational
cultures with different languages or religious attitudes as they
did, say, in Norway, France, Spain, Belgium, and Canada. Nor was
the conflict over the church ever among different religions in the
Latin American case. As in southern Europe, the questions related
rather to church control over education, the registration of births
and deaths, and, not least, the ownership of land, with liberals
typically wanting to open up entailed church estates to the
operation of market forces.
However, there is a more important consideration for the
understanding of contemporary Latin American party systems and
their contrasts with western patterns. For if western party systems
evolved more or less incrementally, with parties based on newly
salient cleavages' being added to the existing system, in time
shunting aside parties founded on previously prominent cleavages,
reducing them to minor party status, or interacting with them in
complex ways, this has been the case only exceptionally in Latin
America.
Chile's party system, until its effective suspension by the
military coup of September 1973, did substantially follow the
classic continental European pattern.7 In Argentina, too, the
current governing party, the Radical Civic Union (UCR), traces its
roots to the 1890s. Ecuador and Panama have also exhibited some
evolutionary continuity, albeit much more tentatively. However, in
these countries the fragmentation and even virtual disappearance of
the traditional parties and the volatility of newer ones have
tended to blur the patterns, characteristic of Chile and Argentina,
whereby new parties were added to the system in response to newly
mobilized classes. Effectively, their current party alignments
constitute new party systems.
Yet the great majority of Latin America's party systems do not
fall into the kind of evolutionary pattern typical of the West.
Thus, in a pattern that might be dubbed "discontinuous," the
parties and party systems of perhaps a dozen Latin American
countries have emerged more or less de novo, usually after a
revolution or a long period of dictatorial rule, with few
perceptible links to the prerevolutionary or predictatorial past.8
Most of the traditional conservative and liberal parties simply
ceased to exist, leaving no visible progeny.
True, in a few instances one can find some traces of linkage.
Thus in the Brazilian case the tiny Republican Party of the
post-1946 republic could trace its lineage to the dominant
Republicans of the Old Republic (1889-1930), and some of the rural
political bosses of an earlier era became pillars of the later
so-called Social Democratic Party (PSD).9 Yet the parties, as well
as the party system, of the pre-1930 period were essentially
destroyed by the advent of Getulio Vargas to power in 1930. When
democracy was restored in 1945, the new party system bore little
resemblance to the old.
Rather, then, than the European model of party development
suggested by Lipset and Rokkan, whereby the principal differences
among contemporary party systems can be traced to distinctive
configurations of early cleavages (center-periphery, church-state,
and landowners-commercial/industrial interests), variations among
many of Latin America's party systems reflect divergent responses
to the expanded political mobilization of the last several
decades.
Just as striking, though fewer in number, are those "continuous"
Latin American party systems (Colombia, Honduras, Paraguay, and
Uruguay) that simply have not evolved or changed much at all over
time, despite their countries' marked increases in social and
political mobilization and the emergence of new social classes.
Liberals and conservatives
25
-
Comparative Politics October 1989
-or factions and splinters thereof--still predominate, as they
have since the last century, and while third parties of some
significance have from time to time emerged, they have so far
scarcely shaken the party alignments inherited from history. Not
least, party identification tends to be intense and to have deep
roots in the distant past. Thus Paul Lewis portrays a Paraguay
where "party identification is practically universal," where
"membership in one of them is almost always a lifetime commitment,"
and where to switch party allegiances connotes virtual treason to
one's friends and family. 10
The trajectory of party system development in Latin America has
therefore differed from that of the developed West in part because
the former has largely lacked the cultural, ethnic, and
interreligious cleavages that have characterized the latter, but
most of all because the vagaries of political history have in the
Latin American case all but eradicated those parties that took form
prior to the onset of mass politics (or, in several other cases,
paradoxically forestalled the emergence of new parties altogether
by freezing in place the historic party pattern). Only in a handful
of instances, notably Chile and, less clearly, Argentina, has party
development even roughly followed the western model.
The Advent of Mass Politics
If both the nature and significance of the patterns of cleavage
that prevailed prior to the onset of mass politics mark the
development of Latin America's party systems as different from
those of the industrialized western countries, the coming of
industrialization and universal suffrage likewise had quite
divergent impacts.
In the western case (albeit with the notable exceptions of the
United States and Canada) the admission of the middle and working
classes to effective political participation in the second half of
the nineteenth and early decades of the twentieth century saw the
attendant formation of parties that have been variously termed
"parties of integration," "class-mass parties," or "socialist
working-class parties." Socialist and Communist parties were
typical of the new style of party.11 Ideologically such class-mass
parties tended to be Marxist, or at least to adhere to programs
that spelled out quite explicitly the desirability of a future
where the state owned the means of production. The focus was on the
class struggle, and the appeals of such parties were primarily to
the organized industrial working class.
The coming of universal suffrage and high levels of political
mobilization in Latin America, on the other hand, some decades
later than in the European case, did not eventuate in the kind of
class-mass parties familiar from the industrializing period in the
West, but in something it seems fair to call a Latin American
version (or rather several versions) of the "catch-all" party.
Indeed, such catch-all parties, broadly conceived, have continued
to be the predominant type of party in the Latin America of the
1980s. In most contemporary Latin American party systems, single
class parties (whether working class or bourgeois) have tended to
be relatively peripheral, or mere adjuncts to party systems that
instead revolve around an axis of one or more multiclass
parties.
The catch-all party is one that eschews dogmatic ideology in the
interests of pragmatism and rhetorical appeals to "the people,"
"the nation," "progress," "development," or the like, that
electorally seeks (and receives) the support of a broad spectrum of
voters that extends the party's reach well beyond that of one
social class or religious denomination, and
26
-
Robert H. Dix
that develops ties to a variety of interest groups instead of
exclusively relying on the organizational and mobilizational assets
of one (such as labor unions).
In short, whereas the catch-all party came into prominence in
Europe in the postindustrial era of development and in the wake of
a politics substantially structured by parties with their principal
roots in, and appeals to, one or another social class (or religious
denomination), the Latin American catch-all party has surged to the
front as the preeminent party form during the industrializing stage
of development and in lieu of (prior to?) the emergence of class
parties of the European stripe.'2
Apart from broad characteristics that together mark them as
catch-all,'3 there are, to be sure, distinctions among Latin
American catch-all parties as well. Some are essentially
personalistic instruments of caudillos, often but not necessarily
military in background. Examples include Argentina's Peronists,
Peru's National Odrifsta Union (UNO), the political vehicle of
former military dictator General Manuel Odria, and Ecuador's
National Velasquista Federation (FNV), the party that served as the
political instrument of the late Jose Maria Velasco Ibarra. Still
others tend to be more structured and enduring, with a more
consistent democratic vocation. The programs and ideologies of such
parties, while sometimes couched initially in dogmatic terms, very
quickly become highly pragmatic in an effort to attract broad,
multiclass support and confront the real problems of governing.14
Examples are numerous, but include Peru's APRA, Venezuela's
Democratic Action (AD), Costa Rica's Party of National Liberation
(PLN), and the Dominican Revolutionary Party (PRD). Mexico's
Institutional Revolutionary Party (PRI) and Bolivia's National
Revolutionary Movement (MNR) (at any rate prior to its
fragmentation) would also fit this category of catch-all party. Yet
a third type of Latin American catch-all party tends to be squarely
based on the middle class and to be led by upper middle class
professionals. In fact, it may begin as a "bourgeois" party. But
when such parties successfully reach out to peasants, workers, or
slum dwellers, as Chile's Christian Democrats did in the 1960s, or
to a broader electorate, as Argentina's Radicals have in the 1980s
under the leadership of Raul Alfonsin or as Peru's Popular Action
Party (AP) once did under former president Fernando Belatinde
Terry, they go beyond a single-class constituency to become genuine
catch-all parties.
Notwithstanding the differences among them, all merit the
designation catch-all in that they are pragmatic or eclectic in
program and ideology, multiclass in their support, and oriented to
broad-based electoral appeals that go beyond the mobilization of a
committed constituency. In contradistinction to the pattern of
western party development, catch-all parties, rather than the
class-mass party, have generally been the immediate successors to
traditional elite-centered parties and politics in Latin
America.
The Evidence
Electoral returns from a variety of countries--some of it based
on ecological evidence, some on survey data-as well as evidence on
the social composition of party leadership confirm the general
argument.
In Venezuela, two parties, the social democratic AD and the
social Christian COPEI, have together won at least 85 percent of
the vote since 1973. Yet demographic variables do
27
-
Comparative Politics October 1989
not account for the voting intentions of the Venezuelan
electorate. If one were to select at random one hundred Herrera
(COPEI) supporters and place them in a room, and then repeat this
procedure with Pinerua (AD) supporters, the two groups would look
very similar."5 Polls in two states (one carried by AD, the other
by COPEI) prior to the 1978 election showed each party to be
favored by a significant proportion of each economic stratum: no
less than 25 percent and no more than 45 percent for either party
in every one of four social categories (upper, middle, working, and
poor).'6 Clearly, both qualify as catch-all parties.
Similarly, a 1976 survey in Costa Rica found that voting for the
National Liberation Party (PLN), that country's dominant party, was
only slightly related to such variables as rural-urban residence,
housing conditions (a surrogate for social class), and age. Party
identification tied to events of the brief 1948 civil war appeared
to be a far better predictor of the PLN vote.7
Even such parties as Argentina's Peronists (today formally known
as the Justicialist Party), while perhaps appealing
disproportionately to the working class, are, when looked at more
closely, in reality quite heterogeneous and genuinely multiclass.
Thus between 26 and 49 percent of the lower middle class (in the
federal capital) voted for Peronist candidates in the five
elections between 1960 and 1973. In fact, the proportion of the
upper middle class, as well as of the upper class, voting for Peron
ranged as high as 31 and 30 percent respectively in September 1973.
The Radicals, too, though getting somewhat more support from the
middle class (both lower and upper) than elsewhere, ranged rather
evenly across the social spectrum, with 22 percent of the working
class vote and 28 percent of the upper class vote in the September
1973 election.'8
The leadership of catch-all parties confirms their socially
eclectic nature. Some, such as Argentina's Peronists, Colombia's
ANAPO, Peru's National Odrifsta Union, and the short-lived movement
that adhered to the political banner of former Chilean dictator
Carlos Ibdifiez del Campo in the 1950s, were not only led by former
military officers but, at least in their formative stages, had a
number of active or retired military officers in other prominent
leadership positions.19
Landowners and businessmen have also made up significant
proportions of the national leadership of many catch-all parties.
For the Peronists, 12 percent of their legislators in 1946 and 20
percent in 1963 were landowners,20 while industrialists were among
the prominent confidants and advisers of Peron's first regime,
despite its allegedly working-class base.21 While Venezuela's AD
drew disproportionate electoral strength from rural areas as well
as from organized labor, especially during the 1960s, its founders
and top leadership cadres have largely been comprised of lawyers,
educators, medical doctors, and professional politicians.22
Bolivia's MNR in its heyday showed a similar pattern.23
Some scholars, like Gliucio Ary Dillon Soares, have argued that
the politics of development are the politics of class and ideology
and that Brazil during the years 1946-64, prior to a military coup
that restructured the party system, was a case in point. Thus in
the election of 1960 there was a clear and consistent tendency for
voter preferences for the National Democratic Union (UDN) to
decline and for the Brazilian Labor Party (PTB) to increase as one
"descended" the occupational scale from persons with professional
and high level administrative jobs to unskilled manual workers. At
the same time, it should be noted that neither the PTB nor the UND
found majority favor in any social grouping in what was then a
genuinely multiparty system and that the support of each was in
reality quite
28
-
Robert H. Dix
heterogeneous. Thus among manual supervisory workers the UDN
held 32.5 percent of the party preferences, and the PTB only 28.5
percent. Indeed, the UDN had more than 18 percent support among
"unskilled manual" workers and more than 20 percent among the
"skilled manual" category, while the PTB had over 10 percent of the
electoral allegiance of the top two social categories.24 In fact,
rather than a class party in any usual sense, and despite its name,
the PTB was essentially a vehicle by which such elitist politicians
as Getulio Vargas and Joao Goulart sought to control organized
labor to the government's advantage. The most heterogeneous among
the major parties, the PTB, attracted wealthy landowners, middle
level government employees, artisans, members of the new urban
upper class, and leftist intellectuals, as well as an important
rural following in the states of Amazonas and Rio Grande do
Sul.25
Following the end of military rule in 1985, the Party of the
Brazilian Democratic Movement (PMDB) has at least temporarily
become the overwhelmingly preponderant (and clearly multiclass)
party in Brazil, winning more than half the seats in the chamber of
representatives in the election of 1986. Meanwhile, the heirs to
the tradition of the old PTB have divided into two parties, while a
third, more authentically Workers' Party (PT), has formed as
well.26 None has so far prospered electorally.
Parties such as Argentina's Peronists and Brazil's PTB have
surely had something of the class party about them, notably in the
disproportionate support received from the working class. Yet the
heterogeneity of such parties-both in their electoral support and
in their leadership--and the failure of the majority of the working
class to support them clearly mark them as something other than a
typical European working-class party. Often they can be better
categorized as catch-all parties in Kirchheimer's meaning of the
term.
Indeed, evidence concerning the social base and leadership of
party after party-with Peru's APRA, Chile's Christian Democrats,
and Mexico's PRI only the most prominent among them-could be
invoked by way of further demonstrating that throughout most of
Latin America the catch-all party has tended to preempt the
class-mass party as the predominant party form.
Latin America's Non-Catch-All Parties
There are of course a number of non-catch-all parties in Latin
America in the 1980s. They include the Communist parties of the
hemisphere, as well as a myriad of Marxist variants and splinter
groups that retain a highly ideological content and direct their
appeals especially to workers, although sometimes to peasants or
nonproletarian slum dwellers as well. The Communists are the single
party in Cuba and, prior to the 1973 coup that brought the military
to power, were an important presence in Chile, regularly winning
between 11 and 16 percent of the vote in congressional elections in
the years between 1961 and 1973.27 Some might also wish to classify
Nicaragua's Sandinistas as a Marxist party dominant in its system,
although its membership and appeals are broader than those of most
class-mass parties. Elsewhere, although Communist parties
exist--legally or not--in every country, they are generally
marginal to electoral politics.
There is also a scattering of "bourgeois" parties that appeal
largely to business or middle class constituencies, some strictly
regional or provincial parties (Argentina has had a
29
-
Comparative Politics October 1989
number, for example), occasional parties representing various
narrowly defined issues or causes, and a host of parties centering
on the ambitions of an individual or clique that make little real
effort to develop a broad-based appeal. Parties such as these
generally have been either quite small or ephemeral or both.
Parties of rather greater significance that also do not fall
strictly into the catch-all category are for the most part of one
or two types. The first is the traditional "vertical" party that
continues to dominate such political systems as the Colombian,
Uruguayan, Honduran, and Paraguayan. Yet in key respects they
resemble catch-all parties, or provide a functional surrogate for
them. They are nonideological and pragmatic, and they successfully
mobilize the support of a broad array of groups and social classes,
from landowners and industrialists to shopkeepers, peasants, and
workers. They are perhaps less than catch-all, on the other hand,
in that at election time they rely more on the mobilization of
committed constituencies linked to the party by clientelistic ties
or by a kind of inherited loyalty than on searching out new
supporters among the uncommitted and undecided.
A second variety of party that falls outside the catch-all
designation even while exhibiting certain of its attributes
includes certain parties or coalitions of the left. Some of these
are breakaways from the Communist party (for example, the Movement
toward Socialism, MAS, in Venezuela); others, like the Broad Front
that has participated in Uruguayan elections both before and after
the recent period of military rule (1973-84), contain the
Communists as one element but include as well such parties as the
Christian Democrats. Such parties or electoral fronts are a good
deal more ideological than the typical catch-all party and tend to
appeal to a more restricted social base. Yet they often garner
proportionately more support from the middle and even upper classes
than they do from workers,28 while populistic appeals often take
precedence over the mobilization of union or other class
constituencies. In fact, Chile's Socialist Party, presumptively one
of Latin America's clearest examples of a class-mass party, has at
various times in its history taken on many of the attributes of
populism.29 These parties, too, then, although not strictly
catch-all parties, share certain of their qualities.
If not all significant Latin American parties fall under the
catch-all rubric, neither are catch-all parties equally dominant in
all party systems. The traditional or so-called continuous party
systems have had little place for them.30 And the pre-1973 Chilean
party system (as well as its presumptive postmilitary successor)
saw a catch-all party like the Christian Democrats sharing the
political stage with significant parties of both working-class and
bourgeois orientation.
Yet it remains the case that the pragmatic, multiclass party is
overwhelmingly the predominant party type in the Latin America of
the 1980s. Indeed, in the great majority of Latin American
countries one or more such parties garner the larger proportion of
the vote. Around them, more often than not, the whole system of
parties revolves. Moreover, as we have noted, even the traditional
vertical parties, as well as a variety of putative class parties or
coalitions, manifest distinct catch-all aspects or tendencies. Thus
it seems clear that the preponderant party form in contemporary
Latin America comes considerably closer to reflecting the
characteristics of the catch-all party than to replicating the more
ideological, class-centered mass parties of the West's initial
decades of mass mobilization. If the mass politics of an
industrializing Europe tended to be exclusivist in class and group
terms, in Latin America they have instead been mainly
inclusivist.
30
-
Robert H. Dix
Some Explanations
Why is it, though, that in Latin America inclusivist parties
have proven to be the archetypical parties of the era of
industrialization and mass mobilization instead of the
class-exclusivist parties more typical of the western experience?
Why, in short, has the class-mass party failed, at least so far, to
play much of a role in the politics of Latin America?
It might seem plausible to suggest that there has been an
incentive to form broad-based multiclass parties in Latin America's
presidential systems in order to maximize the chances of capturing
the all-important office of chief executive. There may be something
to this-a comparison of congressional and presidential elections
does show some tendency for presidential candidacies to be fewer
than the number of competing congressional party slates in the same
election. The difference is not usually great, however-the
incentive to coalesce is seldom strong enough to prevent multiple
candidacies for the office of president."3 Moreover, such
coalescence as does occur is usually temporary, joining distinct
parties for electoral purposes only, and does not as a rule lead to
the long-term merger of parties. Political explanations must in
this case, it seems, give way to sociological ones.
However, explanations such as relative affluence and the absence
of feudalism and rigid status barriers prior to the advent of
industrialization--among the explanations used to account for the
absence of class parties in the United States32-can not very well
be used to account for the relative weakness of such parties in
Latin America. Neither affluence nor relative social equality can
be said to have been hallmarks of the Latin American condition.
Certainly, at least a number of the Latin American countries are
urban, industrial, and literate enough to have sizable, articulate
working classes potentially capable of forming the basis of
class-centered political parties. To take only the "upper middle
income" countries among them by World Bank criteria (Chile, Brazil,
Panama, Uruguay, Mexico, Argentina, and Venezuela), they range in
the mid 1980s from 69 to 85 percent urban (except for Panama, which
is only 50 percent urban). The comparable range for today's
"industrial market economies" is 56 percent (Austria) to 92 percent
(Great Britain).33 Contemporary Latin America is therefore
virtually as urban as the now-developed West, and certainly more so
than the West was at a comparable stage of industrialization.
Data on industrial employment for 1980, meanwhile, show that
between 25 and 34 percent of the labor force in the same Latin
American countries is employed in industry (again with Panama as an
exception at 18 percent). The comparable range for today's
"industrial market economies" is 29 to 44 percent, a meaningful,
yet hardly drastic, difference.34 Thus, while Latin America's work
force is less "industrial" than those of the "industrial
economies," the proportion of industrial workers in at least a
number of the Latin American countries is surely high enough to
sustain class-mass parties of the kind characteristic of earlier
stages of the West's industrialization.
The nature and timing of social mobilization in Latin America
has nonetheless been quite different from that of Europe. Whereas
in the industrialized West those employed in the secondary sector,
in industry and related occupations, succeeded agricultural and
other primary sector employment as numerically predominant, in
Latin America the tertiary or service sector has done so. Secondary
sector employment has never predominated in Latin
31
-
Comparative Politics October 1989
America, and presumably never will; as agricultural employment
declines, the service sector has expanded more rapidly. The massive
inmigration to Latin America's cities in recent decades has to some
extent found employment in industry, of course, but domestic
services, (often temporary) construction work, and petty
entrepreneurship have absorbed the bulk. The more advanced
countries of today's Latin America, still far from fully
industrial, yet with the agricultural sector a rapidly diminishing
proportion of the work force, have seen their service sectors reach
proportions approaching those of the postindustrial West.35
Individuals so employed are much less susceptible to union
organization and class-oriented political appeals in the western
sense than are industrial workers. Indeed, evidence from a number
of Latin American (and other) countries shows that the urban
migrants, so often residents of the shantytowns that ring the
burgeoning cities, tend to see their present and future in terms of
individual, rather than class or group, mobility, thereby adhering
to what Alejandro Portes has called a "migrant ethic."36 Their
demands tend to center on acquiring a bit of land on which to
construct a dwelling and on such amenities as sewers and
transportation for their barrios, rather than on grievances against
a factory boss, much less against the capitalist system itself.
Such is not the kind of social situation in which class solidarity
thrives.
The union movement, moreover, even where it has been quite large
and robust, as in Argentina and Venezuela, has been socially very
heterogeneous, containing high proportions of white collar workers.
The latter are not necessarily less militant-the contrary is often
the case-but the diversity of perspectives and of class outlooks
of, say, teachers, government employees, and metallurgical workers
has a tendency to dilute the kind of tightly knit working-class
subcultures that once flourished in the industrial countries. Thus
in Argentina as of 1970 only four of the largest seventeen unions
were industrial (metallurgical, textile, garment, and automotive
workers), one was comprised of railroad workers, and two of workers
in construction. All of the others were either white collar unions
(including two of the three largest, teachers and commercial
employees) or workers and employees of government or services of
various kinds.37
In some countries the campesinos have been politically mobilized
as well, almost in tandem with the industrial proletariat, in
societies where agricultural employment, though declining, is still
important. Indeed, a number of the catch-all parties--Venezuela's
AD, Mexico's PRI, and Bolivia's MNR among them-have had peasants as
a principal organizational and electoral support base, despite
their close ties to industrial and mining labor and the urban
middle class.
Taken together, these facts make clear that the industrial
working class has comprised but one element, and not necessarily
the most important, available for mass mobilization in the newly
industrializing countries of Latin America. Generally underemployed
urban migrants, an expanding class of white collar workers, and
campesinos have also constituted groups available for mobilization
and electoral appeals. Any party that seeks to win a plurality, let
alone a majority, of the vote has had to encompass them, or a
substantial part of them, as well, yet their demands and outlooks
have seldom been those of a Marxian industrial proletariat.
Latin America's would-be industrialists and the modernizers
among middle class professionals were meanwhile impelled by their
own weakness in competition with traditional landed and commercial
elites and by their common interests in development to
32
-
Robert H. Dix
seek political allies among such groups, thereby adding another
ingredient to at least some of the political coalitions we have
termed catch-all parties. As noted earlier, such individuals are
frequently found among the leaders of such parties in Latin
America.
Still other factors might be adduced to help account for the
weakness of class-mass parties in Latin America and the
corresponding prevalence of the multiclass catch-all party during
Latin America's industrializing phase, among them the fact that the
working class tended to be granted the suffrage relatively early
and "from above," by elites seeking political allies or pursuing a
strategy of cooption, thus precluding a prolonged
consciousness-raising struggle for political participation.38
Yet the fundamental determinant of the difference in the types
of parties thrown up by the twin processes of industrialization and
the introduction of mass suffrage in Latin America and the West,
respectively, would seem to lie in the heterogeneous composition of
the "masses" in Latin America, in contrast to the high salience of
class conscious industrial workers in the West, and the consequent
incentive for political leaders to form broad-based parties to
encompass them.
Therefore the situation of Latin American countries as late
developers (or "late-late developers," as Albert Hirschman has
called them)," particularly the telescoping of their urbanizing and
industrializing processes and the alliance of diverse groups and
classes behind certain broad goals of development and nationalism,
seems best to account for the contrasts in the nature of western
and Latin American mass parties during the early to middle levels
of their respective eras of industrialization.
Conclusion
We have argued that, broadly speaking, the development of Latin
American party systems has diverged from that of the now-developed
West in two fundamental respects. First, whereas in the western
case early patterns of cleavage and party development cast the
basic mold for contemporary party systems, with the (albeit
important) partisan manifestations of the worker-employee cleavage
subsequently appended to them, in the Latin American case this has
been true only in a small minority of instances (notably in Chile
and Argentina). In most Latin American countries the party past has
been rendered largely irrelevant to the present by history. The
majority of Latin American party systems (here called
discontinuous) at best bear only traces of a past that precedes the
contemporary era of economic development, urbanization, and
universal suffrage. (On the other hand, in a few Latin American
countries with so-called continuous party systems the original form
of the party system has paradoxically become set in concrete, as it
were, seemingly resistant to the impact of rapid economic and
social change and broadened political participation.)
Our second principal argument has been that, when mass politics
did appear in Latin America, they tended to take the form of the
inclusive, multiclass party of rather eclectic, pragmatic ideology
and appeals. Even those systems where the parties of an earlier era
still hold sway may be said to be dominated by parties of a
multiclass appeal and ideological pragmatism, even while their
origins and structures may differ from the purer versions of
catch-all parties. The parties of integration (or class-mass
parties) so familiar in European politics during its era of
economic development and mass politics have, by and large, not
33
-
Comparative Politics October /989
appeared in Latin America. Our argument therefore supports Leon
Epstein's contention that "large-membership working-class parties
are a product occurring only at certain stages of social
development in certain nations."4
Whereas Epstein saw such conditions as receding in contemporary
Europe,41 it may be that in Latin America (as in the United States,
albeit for different reasons) they have never really arrived, and
may not do so in the future. In fact, it might be suggested that
the catch-all parties which Kirchheimer saw as having succeeded the
class-mass party in postwar Europe in tandem with the advent of
"postindustrial" society simply appeared in Latin America without
the intervening stage, for reasons rooted in the nature and timing
of Latin America's process of late development.
By the same token, there are manifest differences between
Europe's catch-all parties and the Latin American versions of such
parties, reflecting their distinct origins and styles and above all
the distinct functions they have performed in their respective
societies and time frames.
If most of the West's catch-all parties have evolved from the
socialist or religious parties of the prewar period, most of Latin
America's parties have been created de novo, the consequence of
traumatic breaks in their party systems (although the Christian
Democrats of Chile had their origin in the youth wing of the
Conservative party, while Argentina's Radicals, which emerged in
the late nineteenth century to challenge the dominant oligarchy of
the day, arguably only in the 1980s have evolved into a
full-fledged catch-all party).
Typically, they have centered around key personalities, not only
in their formative years, but often for a long time thereafter as
well. Victor Radl Haya de la Torre (APRA, Peru), R6mulo Betancourt
(AD, Venezuela), Jose (Pepe) Figueres (PLN, Costa Rica), and Juan
Peron (Justicialists or Peronists, Argentina) are but a few of
those founding fathers who remained leaders of their parties for
decades. Unusual in the western case, among Latin America's
catch-all parties-even the most institutionalized among
them-dominant personalities have frequently played major roles and
provided a significant cohering element to otherwise quite opposing
groups and classes.42 Rather than confronting the problems and
conditions of postindustrial societies, the policies and programs
of the Latin American catch-all parties have tended to be the
agents of economic development and to stress the mobilization of
"the people" above or across deep-seated cleavages in support of
broad national, even nationalist, goals. The Latin American
versions of this type of party have therefore tended to play a
mobilizational role with respect to the entry of new groups into
political life, a role more akin to the class-mass parties of
western experience. Indeed, even such non-catch-all parties or
coalitions as Uruguay's Broad Front and Chile's Socialists have at
times shown marked tendencies toward class and group inclusion
(that is, toward populism), rather than toward the exclusivism of
strictly class-based parties.
Most important, whereas Kirchheimer's catch-all parties were a
largely postindustrial phenomenon, reflecting the assuaging of
class tensions and the consolidation of the welfare state, along
with a certain "bourgeosification" of the working class, Latin
America's catch-all parties have emerged relatively early in the
industrializing process and in the initial stages of mass
politics.
We end by affirming the conclusion of Ergun Ozbudun, summing up
a volume that examined competitive elections in a variety of
developing countries: far from being confined to postindustrial
societies, "catch-all parties seem to be the norm, rather than the
exception,
34
-
Robert H. Dix
in Third World countries."43 The point is even more critical for
politics than for scholarly understanding of political institutions
in developing countries, at least if Ozbudun is also correct in
arguing that "the success of democratic politics in developing
societies is strongly associated with the presence of
broadly-based, heterogeneous, catch-all parties,"44 a point with
which we would agree but which we have not directly sought to
develop here. Even more than it is the politics of class and
ideology, the politics of development is the politics of
cross-class coalitions and programmatic pragmatism.
NOTES
I wish to thank my colleague John Ambler as well as two
anonymous readers for their perceptive substantive and editorial
suggestions on the draft of this article.
1. Seymour M. Lipset and Stein Rokkan, "Cleavage Structures,
Party Systems, and Voter Alignments: An Introduction," in Seymour
M. Lipset and Stein Rokkan, eds., Party Systems and Voter
Alignments (New York: The Free Press, 1967), pp. 1-64.'
2. Cf. World Bank, World Development Report 1987 (New York:
Oxford University Press, 1987), pp. 202-3. Latin America in the
context of this article encompasses the nineteen independent
countries of the western hemisphere with an Iberian heritage and
colonial background.
3. Of course, this is not literally the beginning of all
cross-national study of party systems in Latin America. However,
seldom have such studies sought explicitly and systematically to
address the questions raised here, and none of which I am aware has
endeavored, except perhaps in passing, to analyze the evolution of
Latin American party systems in juxtaposition to the experience of
the West. Among the previous comparative studies of Latin American
party systems is Ronald McDonald, Party Systems and Elections in
Latin America (Chicago: Markham, 1971), currently being revised and
updated for a new edition. A comprehensive and very useful
compilation of country-by-country, party-by-party descriptions is
Robert J. Alexander, ed., Political Parties in the Americas:
Canada, Latin America and the West Indies, 2 vols. (Westport:
Greenwood Press, 1982).
4. The classic statement of the tentativeness of Latin American
political systems is Charles Anderson, Politics and Economic Change
in Latin America (Princeton: Van Nostrand, 1967), chap. 4.
5. See Lipset and Rokkan, "Cleavage Structures." 6. Of course
liberals and conservatives were not always consistent in what they
advocated. Liberals, for example,
once in power, often became staunch promoters of central
authority, in practice at least, if not in doctrine. Moreover,
there were indeed conservative merchants and liberal landowners. In
fact, the very designations conservative and liberal at times
appeared to be mere labels, adopted by one or another caudillo in
order to enhance his image or legitimacy.
7. Conservatives and Liberals (with their various factions and
permutations) were by 1857 supplemented by the Radicals (much in
the French tradition of that designation), then by the Democrats, a
small "petit bourgeois" party (1887), the Communists (1921), the
Socialists (1933), and much later, in part as a breakaway of the
Conservatives' youth wing, the Christian Democrats, not to mention
a myriad of other, mostly ephemeral, parties of varying ideological
hues. The Conservatives and Liberals merged to help form the
National Party in 1966.
8. In a case like Brazil, such traumatic interruption has
occurred several times, first when the Liberals and Conservatives
of the empire disappeared with the advent of the republic in 1889,
then when the dictatorship of Getulio Vargas put an end to the Old
Republic in the 1930s, and again when the authors of a military
coup brought an end to the party system of the Second Republic in
1964 and effectively decreed a two-party system comprised of a
government party and an opposition. In an effort to divide its
opposition the military subsequently (1979) opened the system to a
variety of parties that have competed under civilian rule since
1985.
9. See Riordan Roett, Brazil: Politics in a Patrimonial Society,
rev. ed. (New York: Praeger, 1978), pp. 65 and 69. Despite the
name, the PSD was not a social democratic party in the European
sense.
10. Paul H. Lewis, Paraguay under Stroessner (Chapel Hill:
University of North Carolina Press, 1980), pp. 145-50. For a
comparable discussion concerning Colombia and its "hereditary
hatreds" that led to some 200,000 deaths in interpartisan violence
as late as 1946-1966, see Robert H. Dix, The Politics of Colombia
(New York: Praeger, 1987), pp. 92-94. For Uruguay see Juan Rial,
"The Uruguayan Elections of 1984: A Triumph of the Center," in Paul
W.
35
-
Comparative Politics October 1989
Drake and Eduardo Silva, eds., Elections and Democratization in
Latin America, 1980-1985 (San Diego: Center for Iberian and Latin
American Studies, University of California, San Diego, 1986), pp.
262--64.
11. For these terms see, respectively, Sigmund Neumann, Modern
Political Parties (Chicago: University of Chicago Press, 1956);
Otto Kirchheimer, "The Transformation of the Western European Party
Systems," in Joseph LaPalombara and Myron Weiner, eds., Political
Parties and Political Development (Princeton: Princeton University
Press, 1966), pp. 177-200; and Leon Epstein, Political Parties in
Western Democracies (New Brunswick: Transaction, 1980). Analogous
in some respects were parties like Germany's Catholic Zentrum,
where religion replaced social class as an integrating force.
12. The classic depiction of the catch-all party is found in
Kirchheimer. I do not mean to imply that there were no important
parties with a multiclass following in prewar Europe (for example,
Great Britain's Conservatives), nor that all of Europe's putative
class parties appealed almost exclusively to the industrial working
class (France's Socialists did not, for example). Rather,
tendencies and contrasts are at issue, the central question being
why Latin America has failed to develop sizable Communist parties
or such ideological, working-class-oriented parties as the prewar
Social Democrats of Sweden or Germany and the Labor parties of
Great Britain and Norway.
13. The use of the term catch-all is not critical here, as long
as the point is carried that Latin America's mass parties tend to
be cross-class and nonideological in nature. Some Latin
Americanists might prefer the term populist to refer to many
(although perhaps not all) such parties.
14. For an excellent example of this process with regard to
Peru's APRA, see Frederick B. Pike, The Modern History of Peru (New
York: Praeger, 1967), and Grant Hilliker, The Politics of Reform in
Peru: The Aprista and Other Mass Parties (Baltimore: The Johns
Hopkins University Press, 1971).
15. Robert E. O'Connor, "The Electorate," in Howard Penniman,
ed., Venezuela at the Polls (Washington, D.C.: American Enterprise
Institute, 1980), pp. 86-87.
16. Ibid., pp. 80-81. 17. Mitchell A. Seligson, "Costa Rica and
Jamaica," in Myron Weiner and Ergun Ozbudun, eds., Competitive
Elections in Developing Countries (Washington, D.C.: American
Enterprise Institute, 1987), p. 171. 18. Peter G. Snow, Political
Forces in Argentina, rev. ed. (New York: Praeger, 1979), pp. 36-39.
19. See Robert H. Dix, "Populism: Authoritarian and Democratic,"
Latin American Research Review, 20 (1985),
33. 20. Snow, p. 32. 21. Jose Luis de Imaz, Los Que Mandan
(Albany: State University Press of New York, 1970), p. 17. 22. John
W. Martz, Accidn Democrdtica (Princeton: Princeton University
Press, 1966), pp. 195ff. 23. Christopher Mitchell, The Legacy of
Populism in Bolivia (New York: Praeger, 1977), pp. 17-19 and 26-28.
24. Glducio Ary Dillon Soares, "The Politics of Uneven
Development," in Lipset and Rokkan, eds., p. 187. 25. Roett,
Brazil, p. 67. 26. Most of the PT's leaders are blue collar
workers, a rare occurrence in Latin America; see Margaret E.
Keck,
"Great Expectations: The Worker's Party in Brazil (1979-1985),"
paper prepared for the Thirteenth International Congress of the
Latin American Studies Association, Boston, October 1986.
27. Arturo Valenzuela, Origins and Characteristics of the
Chilean Party System: A Proposal for a Parliamentary Form of
Government, Working Paper (Washington, D.C.: The Wilson Center
Latin American Program, 1985), Table 1.
28. Thus O'Connor found in a two-province study of the 1978
Venezuelan election that, while 8 percent each of the upper and
middle strata supported the left, only 6 percent of the working
class and 3 percent of the poor did so (O'Connor, p. 81). Charles
G. Gillespie found a similar pattern for Uruguay; see his
"Activists and Floating Voters: The Unheeded Lessons of Uruguay's
1982 Primaries," in Drake and Silva, eds., p. 234.
29. See Paul W. Drake, Socialism and Populism in Chile, 1932-52
(Urbana: University of Illinois Press, 1978). In 1952, for example,
the majority wing of the divided Socialists, together with an
eclectic array of parties and political "movements," backed the
presidential candidacy of former military dictator (1927-31) Carlos
Ibfilez del Campo. 30. A conspicuous, if short-lived, exception to
this generalization flourished in Colombia in the late 1960s and
early
1970s. The National Popular Alliance (ANAPO), essentially the
political vehicle of former dictator General Gustavo Rojas Pinilla,
won 39 percent of the vote, and nearly the presidency, in 1970 in a
multicandidate election with a populistic appeal and strong support
from the urban masses, plus the support of a number of rural areas,
all of them disaffected from the then-reigning power-sharing
agreement between the Conservative and Liberal parties called the
National Front. Significantly, and attesting to the strength of
traditional party loyalties in Colombia, ANAPO at its peak
functioned not as a formally separate party (though it became one
in 1971), but as a combination of dissident
36
-
Robert H. Dix
factions of the two major parties. When in 1970 Rojas Pinilla
himself ran for president, he did so under the Conservative label;
see Robert H. Dix, "Political Oppositions under the National
Front," in R. Albert Berry, Ronald G. Hellman, and Mauricio
Solatin, eds., Politics of Compromise (New Brunswick: Transaction
Books, 1980), pp. 140-64. ANAPO has since virtually faded from
sight. 31. The cross-national variance is considerable in this
respect. Pre-1973 Chile was at one extreme: there were
typically three or four presidential candidates but many more
congressional slates, with many minor parties explicitly or tacitly
backing one of the major party candidacies. In Peru in 1980, on the
other hand, there were fifteen presidential candidacies and fifteen
congressional slates; cf. Sandra L. Woy-Hazelton, "The Return of
Partisan Politics in Peru," in Stephen M. Gorman, ed.,
Post-Revolutionary Peru (Boulder: Westview Press, 1982), p. 55.
More typical than either of these cases are Costa Rica and
Venezuela, where the two leading candidates as a rule garner 8-10
percent more of the vote than do their respective congressional
slates and there are usually two or three more parties seeking
representation in congress than there are presidential
candidates.
32. Cf. Epstein, Political Parties, chap. 6. 33. World Bank,
World Development Report 1987, p. 267. 34. Ibid., p. 265. 35. Ibid.
36. Alejandro Portes, "Urbanization and Politics in Latin America,"
Social Science Quarterly, 52 (December 1971),
697-720; see also Joan M. Nelson, Migrants, Urban Poverty, and
Instability in Developing Nations (Cambridge, Mass.: Harvard Center
for International Affairs, 1969).
37. Juan M. Villareal, "Changes in Argentine Society: The
Heritage of Dictatorship," in Monica Peralta Ramos and Carlos H.
Waisman, eds., From Military Rule to Liberal Democracy in Argentina
(Boulder: Westview Press, 1987), p. 95. 38. Cf. Epstein, concerning
the relevance of this factor in the case of the United States. 39.
Albert O. Hirschman, "Underdevelopment, Obstacles to the Perception
of Change, and Leadership," Daedalus,
97 (Summer 1968), 925-37. 40. Epstein, p. 132. 41. That the
decline of class politics in Europe may have been exaggerated is
suggested in ibid., pp. 368-74; see also
Steven Wolinetz, "The Transformation of Western European Party
Systems Revisited," West European Politics, 2 (January 1979), 4-28.
42. France's Charles de Gaulle was in a sense such a leader.
However, he considered himself above, or apart from,
political parties, even those that adhered to his cause or
invoked his name; he personally spent little time building or
leading his own party. 43. Ergun Ozbudun, "Institutionalizing
Competitive Elections in Developing Societies," in Myron Weiner
and
Ergun Ozbudun, eds., p. 405. 44. Ibid.
37
Article Contentsp. 23p. 24p. 25p. 26p. 27p. 28p. 29p. 30p. 31p.
32p. 33p. 34p. 35p. 36p. 37
Issue Table of ContentsComparative Politics, Vol. 22, No. 1
(Oct., 1989), pp. i-ii+1-122Front Matter [pp. i-38]Rural
Development Areas in Swaziland: The Politics of Integrated Rural
Development [pp. 1-22]Cleavage Structures and Party Systems in
Latin America [pp. 23-37]Patterns of Church Influence in Brazil's
Political Transition [pp. 39-61]The Adequacy of the Electoral
Motive in Explaining Legislative Attention to Monetary Policy: A
Comparative Study [pp. 63-82]Variations in Union Political Activity
in the United States, Britain, and Germany from the Nineteenth
Century [pp. 83-104]Review ArticleReview: Deregulation and the
State in Comparative Perspective: The Case of Telecommunications
[pp. 105-120]
Errata: Popular Democracy in the New Nicaragua: The Case of a
Rural Mass Organization [pp. 121-122]Back Matter