PA RT Y P O L I T I C S

V O L 1 5 . N o . 4 pp. 487–520

Copyright © 2009 SAGE Publications www.sagepublications.com

Los Angeles London New Delhi Singapore Washington DC

PARTY NON-SYSTEMS A Conceptual Innovation Omar Sanchez ABSTRACT

This article makes a case for expansion of the conceptual framework for the classification of party universe types. In particular, it introduces the concept of ‘party non-systems’, defined as those party universes characterized by a fundamental absence of inter-temporal continuity in the identity of the main parties. At the heart of this concept is the explicit differentiation between intra- and extra-systemic volatility. Party nonsystems are characterized by persistently high transfers of votes away from the main parties towards new and small parties (i.e. high extrasystemic volatility), an ever-changing constellation of parties without a stable ‘core’. It is argued that the difference between non-systems and all other party universe types is not only one of degree (in level of institutionalization), but also one of kind. This conceptual innovation is then applied to a number of Latin American cases (Guatemala, Ecuador, Peru and Bolivia) at the low end of the institutionalization continuum to highlight important cross-country and inter-temporal differences in the nature of (core) party competition.

KEY WORDS
electoral volatility
institutionalization
Latin America
party systems
stability/instability

Introduction Institutionalization is the most important dimension along which party systems in the developing world differ. Variations in fragmentation or polarization are relatively unimportant criteria for comparing party systems whenever essential differences in the degree of institutionalization exist among them. Ever since Mainwaring and Scully’s (1995) landmark collaborative study, it has been more explicitly recognized that a low level of party system institutionalization poses serious problems for democratic governance. With the present study, my aim is to contribute to the still embryonic conceptualization of 1354-0688[DOI: 10.1177/1354068809334566]

PA RT Y P O L I T I C S 1 5 ( 4 )

party system under-institutionalization (Schedler, 1995). I contend that some party constellations, characterized by a fundamental inter-temporal discontinuity in the identity of the main parties in the polity, do not deserve the label of party ‘systems’. I introduce the concept of ‘party non-systems’ to join existing party universe categories and operationalize it in more concrete terms. The definition provided is then applied to a number of Latin American cases at the low end of the institutionalization continuum (Guatemala, Ecuador, Bolivia and Peru) in order to highlight important qualitative differences in the nature of inter-party competition among them – and across time in each country. In the process, I show the descriptive and analytical value of the concept of party non-systems. I thus question the tacit scholarly assumption2 that party systems exist throughout Latin America (Abal Medina and Cavarozzi, 2002; Achard, 2004; Alcantara Saez and Freidenberg, 2003; Artiga-Gonzalez, 2000; BID, 2002; Booth and Seligson, 1995; Cerdas Cruz, 1996; Goodman, 1992; Mainwaring and Scully, 1995; McDonald and Ruhl, 1989) or, indeed, throughout the developing world.

Defining Party Non-Systems A party system, properly understood, is more than simply the sum of its parts (Janda, 1993: 179). Standard definitions have conceptualized it as a set of political organizations with an identifiable or organized pattern of interactions (Mair, 1997) or ‘patterns of competition and cooperation among the different parties’ (Ware, 1996: 7). In his seminal study on party systems, Sartori (1976: 244–8) made the distinction between consolidated and ‘fluid’ systems, but abstained from theorizing about the latter. It can be argued that constructing this dichotomy requires establishing arbitrary cut points, and that institutionalization can more profitably be analysed as a continuum. Sartori’s dichotomy between consolidated systems and fluid systems has rightly been characterized as simplistic because it only posits two broad categories and does not recognize the many shades of grey and variety within each. The choice, however, need not be one between Sartori’s oversimplified scheme or Mainwaring and Scully’s (1995) continuum approach. Reintroducing, refining and expanding the idea of party system thresholds while concurrently retaining gradation is both possible and desirable.1 This author concurs with Coppedge (1998: 173) when he writes that the existence of a collection of parties is only a precondition of a party system, and thus ‘not all Latin American democracies have party systems’ (and neither do all developing country democracies). Yet, the conceptual theoretical implications of these and other similar observations have not been addressed. I argue that there are a limited number of countries at one extreme of the institutionalization continuum for which the term ‘non-system’ is apt and that its use is of descriptive and analytical value. The adoption of this new concept should 488

S A N C H E Z : PA RT Y N O N - S Y S T E M S

afford comparativists an increase in analytic differentiation by better capturing diverse forms of party universes. Mainwaring and Torcal (2006: 205) propose three criteria for identifying a party system. Few scholars would quarrel with this conventional definition: (1) A system implies at least two parties; (2) There must be some regularity in the distribution of electoral support, even when over time some parties rise and others decline; and (3) There must be continuity in the main components that form the system, that is, there must be evidence pointing to the institutionalization of (some) political parties. Operationalizing the second criterion inevitably involves a substantial degree of arbitrariness. The third criterion, however, is easier to evaluate and can be profitably used as a line of demarcation between (inchoate) party systems and non-systems. What precisely constitutes an institutionalized political party is undeniably an intricate puzzle that remains under-theorized. The few works that address this question since Huntington (1968) show that there is no normative agreement as to what the constitutive elements of institutionalization should be (Janda, 1993; Levitsky, 1998; Panebianco, 1988; Randall and Svasand, 2002).3 I take electoral stability as an operational definition of political party institutionalization because it is the best readily measurable proxy.4 Thus, a first approximation when seeking to ascertain whether a constellation of parties constitutes a party system would be to pose this question: is there a group of political formations (i.e. more than one) that has maintained a politically relevant electoral presence through time? If the historical record shows that all past and presently existing parties have faded into or are rapidly headed for electoral insignificance or outright extinction, without exception, then that unstable universe of ‘shooting star’ parties should be considered a non-system. I propose the following operationalization of the concept: if the identity of the top (two or three or four) party vote-getters (regardless of their electoral ordering) is not the same across more than two elections, then that party universe is best described as a ‘non-system’. As the definition implies, at least three elections need to pass before a party universe can be, retrospectively, labelled a ‘non-system’. It also implies that there needs to be evidence of more than one party showing signs of institutionalization (measured by stability in its vote-share). This rather straightforward definition has the twin benefits of parsimony and easy application. While this is one reasonable operationalization of the nonsystem notion, others are surely possible. At the root of the concept, as will be elaborated, is the difference between volatility at the core of the party universe and volatility at its periphery. Those party constellations where the identity of the main parties is constantly changing (that is, where there are no systemic parties that structure political society through time) can rightfully be considered party non-systems. Valid operationalizations of the concept must be strictly faithful to this essential idea. 489

PA RT Y P O L I T I C S 1 5 ( 4 )

How can party non-systems be characterized? A first approximation to this question is to establish what they are not. Party systems have been classified and described by the properties that define the interaction among party units. Arend Lijphart (1984), for instance, famously proposed the following: minimum winning coalitions, governmental durability, effective number of parties and the number of politically relevant thematic or cleavage dimensions. Other scholars working on party systems have elaborated and added their own list of properties. Non-systems, however, cannot be easily pigeonholed into pre-arranged categories precisely because of their constant fluidity and formless, amorphous nature. In a party non-system, all too many elections bring forth significant changes in the above properties, from the effective number of parties to overall ideological polarization. A party non-system’s very essence is that it defies stability in its collective characteristics. It therefore defies standard taxonomies in the realms of polarization and fragmentation (as well as other dimensions). In a context of permanent flux, describing the ‘typical’ features of a party universe by reference to averages in fragmentation, polarization or ideological tendency is an exercise of questionable value. The term ‘non-system’ succinctly captures this verity. Nonetheless, these party constellations can be characterized in ways other than a negative definition. Beyond their non-systemness, they do share traits. Chiefly, non-systems share much in the nature and texture of their component units: ‘parties’ that fit the traditional minimalist definition of what a party is (‘an organization that fields candidates for public office’), but nothing more substantive. In short, party non-systems are overwhelmingly populated by pre- or proto-parties (Olson, 1998), electoral vehicles par excellence with little claim to veritable societal roots or the representation of groups. In such contexts, party labels count less than personalities. Secondly, non-systems are the ultimate expression of party universes shaped ‘from above’ – that is, shaped by political elites independently of existing ethnic, cultural or socioeconomic cleavages in society. The main political battle lines (the issues emphasized) are arbitrarily chosen and redrawn by political entrepreneurs from election to election. Another related trait non-systems share is a very high level of volatility in (party) supply. Overall electoral volatility is a relational quality: it reflects the influence of both (party) supply and (voter) demand. Supply-side volatility is structural, in that voters of the vanished parties are forced to behave differently at the next election and new parties reduce vote support for the ‘old’ parties (Rose and Munro, 2003: ch. 5). Non-systems are characterized by a structural disequilibrium caused by floating party supply. (Yet, important demand-side sources of volatility are also present.)

Navigating the Waters of Conceptual Innovation I engage here in conceptual innovation following Sartori’s time-tested strategy for achieving analytical differentiation and avoiding conceptual stretching. 490

S A N C H E Z : PA RT Y N O N - S Y S T E M S

His ladder of abstraction (or ladder of generality) is based on the relationship between the meaning of concepts and the range of cases to which they apply. Concepts are ordered in terms of a taxonomic hierarchy at different levels of abstraction (Daly, 2003; Sartori, 1970, 1991). Improved classification can only further the way we ‘play the comparative game’, if it is done in a methodologically rigorous manner, taking root concept definitions seriously and avoiding common pitfalls in the strategy of systematic classification (parochialism, misclassification, conceptual stretching and degreeism). Here, I use Mainwaring and Torcal’s (2006) definition of ‘party system’ as the root concept and apply it rigorously in order to map party system (institutionalization) subtypes within the ladder of generality framework (see Figure 1). Political scientists currently use three broad qualitative categories to differentiate party systems in terms of their level of institutionalization: over-institutionalized, institutionalized and inchoate party systems. By establishing these categories, comparativists underscore that not all party systems are created equal: some party constellations are endowed

Party Universe OVERARCHING CONCEPT

Party Non-System DIFFERENT GENUS

Party System ROOT CONCEPT

Institutionalized Party System

Inchoate Party System

CLASSICAL SUBTYPE

DIMINISHED SUBTYPE

Over-Institutionalized Party System CLASSICAL SUBTYPE Figure 1. Ladder of generality: how party universe types relate to one another

491

PA RT Y P O L I T I C S 1 5 ( 4 )

with higher levels of systemness than others. Systemness can be defined as the degree to which political parties, or a core of parties in a polity, have a structured, patterned interaction among them. How do the three mentioned party system types relate to the root concept and to one another? Institutionalized party systems clearly fit the category of a ‘classical’ subtype: they are full instances of the root definition, while they are differentiated vis-à-vis other subtypes of the concept. This term moves us down the ladder of generality because institutionalized party systems fit a narrower range of cases than the root concept. Over-institutionalized party systems move us further down the ladder of generality because they have more defining attributes (intrusive party penetration and politicization of civil society, deliberate exclusion of outsider parties, etc.) and fit an even narrower range of cases. Another strategy of conceptual innovation is that of creating ‘diminished’ subtypes (Collier and Levitsky, 1997). These are not full instances of the root definition. They specify specific attributes of the root concept that are missing. An ‘inchoate party system’ can be considered a diminished subtype in that it lacks the second defining trait of the root definition of ‘party system’. Party constellations exhibiting average volatility rates of, say, over 30 percentage points (an often used arbitrary cut-off point to separate consolidated and inchoate party universes) are less than full instances of a bona fide party system by virtue of the fact that there is little regularity in voting patterns. Yet, inchoate party universes remain part of the party systems genus because they exhibit other significant attributes of the root concept (basic continuity in the identity of the units, etc.). To draw an intuitive analogy, an ‘inchoate party system’ stands to a ‘party system’ as ‘illiberal democracy’ stands to ‘democracy’: one important attribute (civil liberties) is missing, but others (full suffrage and full contestation) are present. Where do party non-system constellations stand analytically in relation to their conceptual relatives? As I have defined it, a party non-system could conceivably be construed as a diminished party system subtype in that it lacks two of the three defining attributes of the root concept: minimal regularity in the distribution of electoral support (the second criterion) and continuity in the identity of the main units of the system (the third criterion). However, this would be tantamount to asserting that any constellation of two or more parties qualifies as a true system. This is in fact the loose manner in which the term ‘party system’ has been and continues to be used, uncritically applied to any and all polities where parties are found. The reality is that the term suffers from ‘conceptual stretching’, the distortion that occurs when a category developed for one set of cases is extended to additional cases which are sufficiently different that the category is no longer appropriate in its original form (Collier and Mahon, 1993). Explicit recognition of this distortion should prompt scholars to make use of the term ‘party universe’ or ‘party constellation’ as a more general category (than ‘party system’) that does not assume inter-temporal continuity – or any observable structured 492

S A N C H E Z : PA RT Y N O N - S Y S T E M S

relationship – among its constituent units.5 Non-systems constitute a different political species from the rest of party universes because they lack a core, which is what endows party constellations with systemic qualities. These party constellations transgress the land of systemness into the no-man’s land of a chaotic interaction among ever-changing units (parties). It is this latter trait that puts it in a different genus among party universes. Party nonsystems can be placed in the institutionalization continuum while concurrently standing separate from the ‘systemness’ genus. What needs to be more explicitly acknowledged is that party universes exhibiting very high levels of average volatility differ (often substantially) in the inter-temporal continuity in the identity of the main parties. Party universes exhibiting this difference are qualitatively distinct. Inchoate party systems exhibit high intra-systemic volatility, that is, fluctuations in vote preferences among the main parties in the polity. Non-systems exhibit, in addition, high extra-systemic volatility, defined as transference of votes away from the collective of main (core) parties in the polity (see Figure 2). Extra-systemic volatility will be low as long as small and new parties are unable to dent significantly the collective vote-share of the main parties at each electoral cycle. In short, high party birth and death rates may not necessarily affect the pre-eminent position of the main parties through time. If that is the case, that party universe deserves the label of a party system. Non-systems, on the other hand, are characterized by persistently high net

EXTRA-SYSTEMIC VOLATILITY

A

D

INTRA-SYSTEMIC VOLATILITY

B

E

C

F

PERIPHERAL PARTIES

MAIN (SYSTEMIC) PARTIES

EXTRA-SYSTEMIC VOLATILITY Arrows represent vote transfers Black arrows: net transfers away from main parties

Figure 2. Visual difference between intra- and extra-systemic volatility

493

PA RT Y P O L I T I C S 1 5 ( 4 )

outflows of votes from the main parties to small and new parties from election to election – as summarized in Table 1. In the chaotic and unstable world of party non-systems, small parties abruptly become large ones and vice versa, while new parties often achieve instant electoral prominence. Some students of party systems will predictably argue that the party universes identified here may not merit a special category because we can simply refer to their level of electoral volatility to place them in the institutionalization continuum. Establishing this new category, critics will contend, requires an arbitrary cut-off point between inchoate and non-systems. Two counter-arguments are in order. First, few concepts in comparative politics are immune from the charge of ambiguous or arbitrary boundaries; what matters is whether they enhance our ability to map out a complex political world – providing analytical differentiation while avoiding conceptual stretching. The threshold separating institutionalized and inchoate systems is hardly clear-cut or unambiguous. Yet, no one would claim that these categories are not useful. They are useful because they are effective ‘data containers’ that simplify political reality and make it more intelligible – without having to read the fine print. They make comparison possible by grouping together party universes that are reasonably similar. (In any case, the dividing line between inchoate and non-systems is comparatively speaking more clear-cut.) Second, those scholars who eschew nominal categories and defend the use of arithmetic measurement to capture the nature of the particular party universes here theorized fall prey to ‘degreeism’ or the mistreating of differences between cases as simply matters of degree rather than of kind (Daly, 2003: 99–100; Sartori, 1991: 20). As previously argued, the difference between inchoate and non-systems is also one of kind, not least because identifying the latter requires a shift to a different level of analysis (core party continuity) and measurement (extra-systemic volatility). Simply measuring overall electoral volatility is insufficient to determine whether a party constellation falls into one category or another. In a seminal article on conceptual innovation, Sartori cogently argued that the rules of concept formation are independent of, and cannot be derived from, quantification itself. It is worth quoting him at some length: In this messy controversy about quantification . . . we simply tend to forget that concept formation stands prior to quantification. The process of thinking inevitably begins with a qualitative (natural) language, no matter at which shore we shall eventually land. Correlatively, there is no ultimate way of bypassing the fact that human understanding requires cut-off points which basically correspond to the slices into which a natural or qualitative language happens to be divided. . . . Let it be stressed, therefore, that long before having data which can speak for themselves the fundamental articulation of language and of thinking is obtained logically – by cumulative conceptual refinement – not by measurement. Measurement of what? We cannot measure unless we know first what it is that we are measuring. Nor can the degrees of something tell us what a thing is. (Sartori, 1970: 1038; italics in the original)

494

S A N C H E Z : PA RT Y N O N - S Y S T E M S

Table 1. Difference between inchoate party systems and non-systems

Inchoate party systems Party non-systems

Intra-systemic volatility

Extra-systemic volatility

High High

Low-medium Persistently high

It follows from this that vote volatility acquires particular relevance as a practical indicator of party system institutionalization only after this latter concept is articulated and reified as important in the minds of political scientists, not before Huntington’s seminal definition of the concept (1968) and Mainwaring and Scully’s (1995) refinements of it. Similarly, the concept of non-systems can potentially elevate the importance of extra-systemic volatility as a variable worthy of explicit and independent measurement – insofar as it is a prime indicator of the degree to which a party universe possesses a core of main parties. Besides the conceptual rationale outlined above, there are at least three practical arguments for drawing the line here proposed to differentiate among under-institutionalized party constellations. First, the current practice of measuring only overall volatility hides important differences in extrasystemic vote-switching. Introducing the non-system denomination can succinctly capture this and other associated qualitative differences (outlined below) when analysing and comparing party systems. Second, while the continuous measurement of institutionalization provided by aggregate (volatility) measures allows for quantitative cross-national research, this approach fails to delineate any thresholds by which to identify inter-temporal party system transformation along the axis of institutionalization. While inchoate since its inception in 1978, there are clearly important differences, for instance, between the pre- and post-1990 Peruvian party universes. Our current conceptual lexicon is inadequate to the task of capturing this and similar transformations. Finally, the ‘non-system’ label can usefully and unambiguously identify party universes at one extreme end of the institutionalization continuum (see Table 2). It captures the quintessence of a chaotic party universe where parties are utterly unable to structure political society.

Routes Towards Non-System Status and Potential Causes So far I have not addressed the puzzle of how or why some polities succumb to chaotic party universes. The case studies that follow suggest at least three routes via which party non-systems may emerge. (Taking into consideration party universes beyond Latin America would certainly reveal additional pathways.) One such route is the birth of electoral democracy amidst a truly inauspicious political environment, one not conducive to political socialization. Where people have been comprehensively alienated from political life, 495

PA RT Y P O L I T I C S 1 5 ( 4 )

Table 2. Party system institutionalization continuum

Highest Overinstitutionalized Characteristics: 1. Main parties penetrate all aspects of society 2. Low electoral volatility 3. Inter-party collusion to limit competition

Lowest Institutionalized

Inchoate

Non-system

Characteristics: 1. Parties with . strong roots in . society 2. Low to medium . overall electoral . volatility

Characteristics: 1. Weak roots in . society 2. High to very . high overall . volatility 3. Low to medium . extra-systemic . volatility

Characteristics: 1. Parties with . virtually no . societal roots 2. Extremely high . overall volatility 3. Persistently high . extra-systemic . volatility

and there is little or no previous experience of democracy, building parties with loyal followers may well prove an extremely difficult, if not elusive, task. A second avenue towards party non-system status is the one charted by the collapse of a well-defined and well-entrenched party system. Such collapses invariably result from the discrediting of the old political order in the eyes of increasing numbers of citizens. A third pathway towards the abyss of non-systems also follows the drift into irrelevance of once-important parties, but is more gradual. The underlying forces fuelling such a transition may not be dissimilar from those responsible for party system collapse, but the transition away from the old party order is more drawn out – there is no single, identifiable transformational election. Note that I have referred to routes or avenues towards non-system status and avoided specifying concrete causal mechanisms. Suggesting causes must be done with caution. More party non-system cases (and from other world regions) than those identified in this article will need to be examined if we are to make valid generalizations about why they emerge at all. It is entirely conceivable, however, that generalizations will prove elusive. The advent of these chaotic party constellations may well obey historically contingent processes, and thus searching for a few identifiable common causes may be unproductive. In any case, we can probe the existing literature for clues. The factors that have been found to spur volatility in general obviously play some role in explaining part of the sky-high volatility underpinning nonsystems. Roberts and Wibbels (1999) find volatility in Latin America to be a function of short-term economic perturbations, the institutional fragilities of both democratic regimes and party systems, and fluid cleavage structures. But non-systems constitute extreme instances of structural disequilibria between parties and voters. Looking at sharp changes in GDP growth or inflation, or changes in electoral and other institutional rules, can yield only some explanatory leverage. These independent variables are important short-term 496

S A N C H E Z : PA RT Y N O N - S Y S T E M S

sources of volatility, but are insufficient to explain the more self-sustaining structural conditions embodied in party non-systems. If one pathway to non-system status is party system collapse, this begs the question of why such collapses occur. Collusive party systems are good candidates for collapse, cartel-type arrangements whereby a few party leaders collude to share spheres of state power and exclude outsiders. Such schemes inevitably conspire against the competitiveness and inclusiveness of electoral contests and erode effective representation and accountability. Responsiveness is also critically diminished, as political elites dance to their own tune, independently of citizen preferences. Failing to give voice to important societal demands, interests and viewpoints prompts such collusive systems to implode – as Morgan (2007) shows for Venezuela, for instance. Ultimately, such regimes contain the seeds of their own destruction due to irresoluble legitimacy deficits. Party system collapse can also ensue from the irruption of new issues long ignored by rulers into political centre stage, as a result of the increased political organization and activism of historically marginalized or neglected sectors of society (Van Cott, 2005). This can often occur rather swiftly and catch the stultified established parties unprepared when faced with an entirely new political landscape. Other reasons can certainly be identified for the breakdown of party system edifices (Coppedge, 2002; Dietz and Myers, 2007; Morlino and Tarchi, 1996; Seawright, 2004). Not all party systems are created equal. The conditions of origin surrounding the birth of (electoral) democracy inevitably shape future party system development. The nature, rootedness and characteristics of the parties contesting the first critical foundational election are some important ‘condition of origin’ elements. The less rooted in society those parties are, the more likely it is that they will be replaced by a second batch of electoral vehicles, setting in motion a self-sustaining dynamic of unstable inter-party competition. Party rootedness at the origin (of electoral democracy) will, in turn, depend upon the idiosyncrasies of a country’s political past. Any and all factors that contribute to the fragility of political society before the onset of competitive elections can predictably affect the course and direction of future party system competition. Theories of party formation assume the existence of civil and political society. This is not a warranted assumption everywhere. Countries burdened with authoritarian legacies where the monopolization of politics was prolonged and methodical are, ceteris paribus, stronger candidates for giving birth to chaotic party non-systems. The different party system trajectories within the post-communist world provide support for this hypothesis, with former one-party Soviet republics displaying higher vote switchover rates than the more pluralist Eastern bloc countries (Bielasiak, 2002: 198). The nature of the dictatorship matters in at least another way. Dictatorial regimes that create new parties to support themselves (‘support parties’), hold elections and foster the co-optation of opponents by allowing them to share the fruits of office effectively disable the old democratic parties. This legacy will generate very high party and voter 497

PA RT Y P O L I T I C S 1 5 ( 4 )

volatility once electoral democracy is established, as Geddes (2004) has shown. State weakness (or state deficiencies) constitutes another element factoring into acute crises of representation and by extension party universe deinstitutionalization. The concept of state deficiencies implies that the state fails to fulfil some of its basic governance, legal and security functions – but it is less profound than state failure or state collapse. In particular, the national executive branch, the judiciary, the police and the armed forces are the state institutions that have the greatest impact on voters’ confidence in parties and legislatures. Ceteris paribus, where the performance of the state is perceived to be most deficient (in ensuring personal security, protecting legal rights, controlling corruption, providing economic opportunity, improving access and quality of education, etc.), disenchantment with representative institutions is greatest (Mainwaring et al., 2006). An extremely feeble state (lacking resources, ‘Weberian’ bureaucracies, administrative competence, human capital) inexorably affects the performance of government (however well-prepared, honest and visionary the office-holders of the executive branch), in a real sense predisposing governing parties to serial failure. The resultant ‘endemic discontent’ will be reflected in intense electoral attrition for the party in power (Molina, 2001). Indeed, in a context of severe state weakness, incumbency is a veritable curse in disguise for any party, dooming future electoral success – as the case of Guatemala shows so starkly. The profound state–society dislocation and crisis of representation that can result from iterated dismal governmental performance erodes popular confidence in parties, and can ultimately contribute to engendering the chaotic party universes of the sort theorized here. There is an unmistakable elective affinity between competitive authoritarianism and party non-systems – as observed, for instance, in Venezuela and Peru – that bears explanation. Competitive authoritarian regimes are defined as those where formal democratic institutions and rules are widely viewed as the principal means of exercising political authority, but where incumbents violate those rules so often and to such an extent that the regime fails to meet minimum standards for democracy (Levitsky and Way, 2002). Such regimes, very common in a post-Cold War era marked by the proliferation of hybrid political systems, are constructed in piecemeal but rather systematic fashion by so-called ‘elected dictators’ (Schedler, 2006). These authoritarian-minded leaders, often taking advantage of the vacuum of power that allowed for their ascent, have as an overarching agenda the steady concentration of political power in their hands through a complex mix of bribery, co-optation and subtle forms of persecution of opponents. Day-today governmental actions and practices, which asphyxiate political pluralism, make it exceedingly difficult for opposition parties to flourish and establish effective links with society. While electoral outcomes are not entirely preordained, elections are heavily stacked against challengers. Moreover, the architects of competitive authoritarian polities are generally uninterested 498

S A N C H E Z : PA RT Y N O N - S Y S T E M S

in crafting a strong, institutionalized political party, not least because that would entail some devolution of their discretionary, personalized authority. Rather, they tend to create empty electoral vehicles that can be discarded (‘disposable parties’). The above political dynamics are prone to produce nonsystems composed of a ‘party of power’ (itself a mutating entity in label and composition) alongside a multitude of allied and opposition electoral vehicles and proto-parties. After the ‘elected autocrat’ is driven from office, the legacy of his rule – in the form of pronounced institutional deterioration and a weakened civil society – makes party system reconstruction (that is, a transition away from non-system status) a very difficult task. The upshot is that non-systems survive long after their ‘architects’ have left the political scene. Perceptive readers will notice that the party non-systems which emerge in the context of competitive authoritarian dynamics do not conform to some of the qualities associated with non-systems here described (in particular, those concerning the very open nature of competition, electoral markets populated by floating voters, accessibility to government and unpredictability of electoral outcomes). The characterization of non-systems offered in this article is predicated on the assumption that minimal standards of electoral democracy prevail, which competitive authoritarian regimes obviously contravene. The party system literature has long displayed what might be termed an ‘institutionalization bias’. The assumption is that, with time, as political elites and voters adjust to one another in a trial-and-error process of interaction, party systems in most neo-democracies will move towards rationalization and stability. Au contraire, there are theoretical and empirical reasons to believe that inchoate party systems and non-systems may well become more common in the future, rather than less. Writing about Latin America in the late 1990s, Frances Hagopian (1998: 106) presciently observed, ‘political representation is in the process of a profound transformation brought on by structurally rooted variables’. Indeed, political representation has in recent times witnessed secular, accelerated decline, manifested in marked and persistent electoral and partisan dealignment. It is an empirical reality that during the past decade or so many more party systems in Latin America have undergone decay or disintegration than institutionalization or consolidation (Abal Medina and Cavarozzi, 2002; Roberts, forthcoming; Sanchez, 2008a). Many long-standing traditional parties have suffered swift declines in their electoral clout, while political outsiders atop electoral vehicles have scored remarkable gains in political power – capturing the presidency or substantial shares of seats in national legislatures (Corrales, 2005). In all probability, it is not coincidental that average regional electoral volatility has been climbing: higher in the 1990s in comparison to the 1980s, and higher still in the 2000s than in the two previous decades (Roberts, forthcoming). Empirically speaking, most Third Wave developing country party systems have not become more stable over time (Mainwaring and Zoco, 2007). Moreover, wherever old party systems have crumbled, party reconstruction 499

PA RT Y P O L I T I C S 1 5 ( 4 )

has not followed. That is because, in the contemporary environment, politicians do not have the capacity or the incentive to build party organizations in the old-fashioned way. As Levitsky and Cameron (2003: 2) have written, ‘strong parties are products of particular historical, sociological and technological conditions that are weakly present in the contemporary period’. Some of the new-found conditions that conspire against party system institutionalization include: the increased influence of mass media technologies (in particular, television) which make party organization at the grassroots less necessary, if not altogether expendable, to curry favour with voters (Skidmore, 1993); the reduction in the functions and role of the state (Vellinga, 1998), diminishing the ability of parties to distribute pork or policy benefits and thus reducing their usefulness to organized interests; the growth of the informal sector (Sabatini and Farnsworth, 2006), meaning that there is a greater number of ‘floating voters’, citizens without attachments to organized politics; and others.

The Damaging Consequences of Party Non-Systems In order to assess the implications of party non-systems for democratic quality, this section proceeds deductively. That is, given the way non-systems have been defined and characterized, what does extant democratic theory predict about the effects of persistently high levels of overall and extrasystemic electoral volatility as well as extremely weak parties on the quality of democratic governance? Non-systems fatally undermine both representation and governability. These are two political public goods that political scientists frequently counterpoise. In the short run, the greater the scope of representation embodied in a party system (i.e. the more different groups or interests represented) the lower the degree of governability. In non-systems, however, the political clout of parties is so small and their representational deficit so acute that governability is seriously compromised as well – rendering null the famed trade-off. The high level of personalism and low degree of party discipline typical of non-systems damages both representation and governability (Coppedge, 2001). In these party universes, societal groups tend to work around parties rather than through them, in voicing their demands, preferring direct interaction with the state. Party politics proceeds in a realm independent of the realm of societal dynamics and is unable to structure and channel social conflict. As a consequence, inter-party accords and legislative output are usually devoid of much societal support. Non-system environments are inimical to the creation and nurturing of a professional political class, especially one that minimally fulfils the function of representation. As Garreton et al. (2003: 79) write, in effect painting the outlines of party nonsystem polities: 500

S A N C H E Z : PA RT Y N O N - S Y S T E M S

In countries where proper parties really do not exist, it is very probable that issues and social groups will not have representation. Alternatively, they will be appropriated and spuriously represented by personalities with political ambitions coming from the media, the entertainment industry, sports, the military, or some other ‘anti-party’ constituency or by particularistic lobbies with social influence.

A universe populated by parties with feeble or non-existent social bases will exhibit acute deficits of governability. Because parties’ societal penetration is virtually absent, societal conflict is not transferred to the electoral arena, where compromise becomes more likely. Therefore, elections in non-systems will predictably fail to unlock or assuage societal stalemates. These party constellations can be expected to showcase acute praetorian traits. Resort to extra-institutional behaviour (including political violence) to ‘settle’ disputes or achieve political objectives will predictably be more common than in any other party universe. Party non-systems also fatally damage democratic responsiveness, or the ability of the democratic process to form and implement policies that the citizens want (Powell, 2005). The first linkage of responsiveness, which links citizens’ preferences to voting behaviour, is subverted by the constant flux in the party supply offered to voters. Some degree of stability in the party offerings is necessary for voters to vote rationally in using elections to support their policy preferences (Riker and Ordeshook, 1973). The high party birth and death rates, mergers and splinters that characterize nonsystems mean voters are burdened with an information overload when assessing their electoral choices. Non-systems are characterized by a structural disequibrium between party supply and demand, to use Rose and Munro’s (2003) apt term. Because the supply of parties never stabilizes, the demands of voters cannot show consistency. The surplus of antagonists in non-systems makes for an irrational form of inter-party competition. It is irrational in the sense that successive electoral contests are ineffective in stabilizing and streamlining the party system. Irrational also in the sense that voters are not afforded the opportunity to voice their preferences consistently through time, given the ever-changing party menu. In addition, non-systems comport a spurious mode of political competition, which stems from their excessive openness. Non-systems are the most open and unpredictable of all party universes: alternation in government and innovation in government formulas are the rule, while accessibility to government is open to any political entrepreneur (i.e. new parties can and do achieve instant success). In addition, because in non-systems virtually no single political party has developed durable roots in society, there are effectively no captive markets of voters (that is, citizens loyal to particular parties). Non-systems are electoral markets where virtually every citizen is effectively a ‘floating voter’ available for political mobilization. Where competition is wholly open, there is little that is predictable in either the patterns of competition or the character of competing forces. In non-systems, elections 501

PA RT Y P O L I T I C S 1 5 ( 4 )

represent exercises at which voters have a chance to punish and eject incumbents; however, they do not represent contests among competing societal interests, projects or world-views. Therein rests the inauthentic nature of competition in non-systems. The electorate is deprived of the capability to choose among identifiable and substantively meaningful alternatives insofar as parties are empty shells devoid of much content. Non-systems are also more susceptible to democratic breakdown. Electoral victory can more easily be attained by anti-democratic candidates and political formations, and thus extant democratic rules and institutions are more vulnerable to purposeful erosion or demolishment. Because they fail in the ‘gate-keeping’ function, non-system polities approximate a world of amateur politicians, many of whom have no experience with, and dubious commitment towards, democratic politics. The more weakly institutionalized a party universe, the greater the space for anti-party politicians to come to power and steer regimes in an authoritarian direction. This is not to suggest that this sequence is preordained. In an era when overtly straying away from democratic governance carries important costs, party non-system polities can persist in a low-level equilibrium of ‘diminished democracy’, rather than regress into forms of authoritarianism. Party non-systems also have unmistakable implications for democratic accountability. Voting based on the personality characteristics of candidates (i.e. personalism) is nowhere more prevalent than in non-systems, while programmatic/ideological or traditional/affective party–citizen linkages are extremely feeble. Accountability through individuals, as occurs in nonsystems, is much weaker than through bona fide parties – not least because their public policy positions are less visible. In party non-systems, political leaders are generally not accountable to their parties, which serve as their instruments (and often they have created). Politicians in such settings are in many ways effectively autonomous. This will contribute to the insidious incidence of betrayals of electoral mandates (Stokes, 1997). Another type of horizontal accountability, that which is exercised chiefly by the legislature in its role as a check-and-balance upon the executive, also is severely constrained. Party and legislative discipline is a crucial component of accountability (Carey and Reynolds, 2007). Such discipline will predictably be in short supply, given the large autonomy legislators enjoy. Legislators representing proto-parties have little to gain from protecting the party label, and often much to gain from breaking party discipline, ‘selling’ one’s vote to the highest bidder, and even switching parties in-between elections. Vertical accountability, that which flows from citizens to parties and political elites, is also seriously compromised in non-systems. Because parties are fluid and ephemeral, voters are unable to create a clear profile of them, and therefore parties are more difficult to appraise. The speed with which parties rise and fall, in addition to the frequency of party schisms, mergers and name changes, makes it a Herculean effort for citizens to order the political universe in intelligible ways – including coming to terms with parties’ positions. 502

S A N C H E Z : PA RT Y N O N - S Y S T E M S

Non-system polities are also prey to short-termism and its damaging consequences for political and economic life. In ever-changing party universes, actors display very short time horizons (Schedler, 1995). Because politico-electoral outcomes are highly unpredictable, political actors cannot develop expectations about the future, even in the medium term. The range of plausible political outcomes is damagingly wide: both Bermuda elections (where established parties disappear or become irrelevant) and Pandora elections (electoral coups d’état by political outsiders) are the norm rather than the exception. In a universe of programmatically vacuous proto-parties, party labels are not useful informational tools to estimate future policy choices – even after the winner is known. Given such high uncertainty, actors are forced to reduce their time horizons to the immediate future and are unable to expand their planning horizons. Political incentives are structured in a way that heavily discounts the future. A priori, the great turnover in power and the lack of systemic parties characteristic of non-systems can be expected to have a negative effect on the quality and stability of government policy (IADB, 2005: ch. 6). Electoral and policy volatility are related. The absence of a few parties that structure the political system through time means that non-systems are congenitally unable to produce anything approaching politicas de estado (vital policy stances that, out of an inter-party consensus, remain in place through time and are purposely placed above party politics). Non-systems conspire against inter-temporal bargains and against coalitional politics. Consolidation of policy innovations and sustained implementation of economic reform are bound to prove elusive in a party universe without true systemic parties.

Guatemala: A Paradigmatic Case of a Non-System Understanding the nature of Guatemalan party politics is scarcely possible without reference to the country’s recent history. The suffocation of political space and the extent of political violence during the four decades under military rule (counter-insurgency state) is probably without parallel in the region (Jonas, 1991). The legacy of that period has been dreadful for the subsequent attempt at building democratic governance: a highly underdeveloped civil society, deep-seated citizen distrust of the state and political institutions and a marked authoritarian political culture (Seligson et al., 2002). Another important feature impinging upon party system development was the ‘original sin’ surrounding the birth of electoral democracy in Guatemala. It was born in the midst of armed conflict, coupled with associated violations of human rights, political crimes, censuring of freedom of speech, the explicit proscription of leftist parties and a military that had not yet come under civilian control. These circumstances did not augur well for the development of bona fide political parties or an institutionalized party system. 503

PA RT Y P O L I T I C S 1 5 ( 4 )

What is notable about the political trajectory of this Central American nation since its transition away from authoritarianism is the ever-changing character of its party universe and, in particular, the absence of anything resembling a stable core of parties – as Table 3 shows. Without exception, all of the parties that were prominent at some juncture have failed to establish themselves as systemic political actors. All have faded into political irrelevance or outright disappearance, which makes Guatemala an example par excellence of a party non-system. Let us review the rise and fall of this carousel of shooting-star parties. The 1985 foundational election was won by a party with significant roots in the country’s history, the Democracia Cristiana Guatemalteca (DCG). The very fact that the DCG could claim an ideology (centrist reformism) and party relatives (the Christian Democratic family of parties throughout Latin America and Europe) differentiated it from virtually all of the other political formations in the country. Indeed, the bulk of participants in that crucial first election can best be characterized as proto-parties rather than bona fide political parties with an ideology, programmatic agenda, roots in society and internal organization. In a clear case of path dependency, that scenario has been self-replicated in subsequent elections. The DCG’s performance in office was lacklustre, with party leaders accepting the army’s priorities and avoiding substantive political and economic issues (Jonas, 1995). In the following 1990 elections, its support had been halved in terms of the parliamentary vote (see Table 3). Its precipitous decline continued unabated, becoming a secondary organization by 1995 (12 percent of the vote), and Table 3. Guatemala: top four parties in first-round presidential (P) and congressional (C) elections (in order of votes gathered)

1985

1990

1995

1999

2003

2007

UCN, MAS, DCG, PAN

PAN, FRG, AN, FDNG

FRG, PAN, DIA-UNRG, PLP

GANA, UNE, FRG, PAN

UNE, PP, GANA, FRG

Same order as above

Same order as above

FRG, PAN, ANN, DCG

GANA, FRG, UNE, GANA, UNE, PAN PP, FRG

(P) DCG-UCN, PDCN-PR, PLN-PID

(C) Same order as above

Source: Tribunal Supremo Electoral. *Based on May 2007 polls for the September 2007 presidential elections. DCG, Democracia Cristiana Guatemalteca (Guatemalan Christian Democracy); MAS, Movimiento de Accion Solidaria (Solidarity Action Movement); MLN, Movimiento de Liberacion Nacional (National Liberation Movement); PDLN, Partido Democratico de Liberacion Nacional; UCN, Union de Centro Nacional (National Centre Union); PAN, Partido de Avanzada Nacional (Party of National Advancement); AN (alliance composed of UCN, DCG and PSD); FRG, Frente Revolucionario Guatemalteco (Guatemalan Revolutionary Front); PLP, Partido Libertador Progresista (Progressive Liberator Party); URNG, Unidad Revolucionaria Nacional Guatemalteca; DIA, Partido DIA; UNE, Unidad Nacional de la Esperanza (Unity of National Hope); GANA, electoral alliance composed of: PP, Partido Patriota; MR, Partido Movimiento Reformador; PSN, Partido Solidaridad Nacional.

504

S A N C H E Z : PA RT Y N O N - S Y S T E M S

failing to field a candidate for the 1999 presidential election as a result of internal party strife. In recent years, the DCG has struggled to survive as a political entity. The other major political formation of the first few years of the democratic era was the Union de Centro Nacional (UNC). This was a quintessential personal vehicle created by Jorge Carpio Nicolle, a member of the Guatemalan elite. After becoming the second most important party in the foundational 1985 contest, it won the first round of the 1990 presidential elections. Carpio Nicolle’s assassination in 1993 also prompted the death of his ‘party’. The UNC was only able to garner 2.5 percent of the vote in the 1995 parliamentary elections and did not even field a presidential candidate that year. It has since disappeared from the political map. In a clear manifestation of citizen disdain for political parties, the 1990 presidential contest was won by a political outsider, the evangelical maverick Jorge Serrano and his Movimiento de Accion Solidaria (MAS). Serrano’s absence of support in parliament led to tense relations and general stalemate in executive–legislative relations, underscoring one of the chief problems associated with weakly institutionalized party systems. Predictably, the MAS evaporated from the political scene after Serrano attempted a failed self-coup in 1993. The second half of the 1990s saw the rise of the Partido de Avanzada Nacional (PAN) and the Frente Republicano Guatemalteco (FRG), both nominally parties of the right, albeit largely non-ideological. The PAN was created by a small group of professionals and modernizing entrepreneurs with the intent of capturing the mayoralty of Guatemala City. It was winning the capital city that allowed the PAN to create an unassailable bastion of power, construct an effective political machine and develop a loyal voting clientele. The party’s original nucleus revolved around the figure of Alvaro Arzu Irigoyen, hailing from a prominent oligarchic family. Arzu became mayor of the capital city and then edged the FRG’s Alfonso Portillo in the second round of the 1995 presidential election. The PAN’s performance in office was lacklustre but not disastrous, and Arzu can be given much credit for the signing of the December 1996 Peace Accords. Yet, the PAN could not avoid the fate of predecessor parties in power. It was soundly defeated in the 1999 presidential election. At that time, control of the party was wrestled away from Arzu. The party’s electoral fortunes quickly languished. An important contingent of PAN party cadres left what they regarded as a sinking political boat. A clearer archetype of a party-church created in the image of its creator, the Guatemalan Republican Front was founded in 1988 by former military dictator General Efrain Rios-Montt. Propelled by the pro-poor populist oratory of Alfonso Portillo, the FRG won the 1999 presidential election with the largest mandate ever granted by the Guatemalan electorate (47.7 percent). However, the FRG’s star faded fast during and after its presidential stint, following the fate of predecessor governing parties. The FRG administration was marked by rampant corruption coupled with a grossly 505

PA RT Y P O L I T I C S 1 5 ( 4 )

ineffective governmental performance. The party has since been in disarray and its future doomed. Even the clientelistic linkage networks it built painstakingly during its years in office have not slowed down its electoral and political descent into irrelevance. The 2003 presidential election was won by a former PANista, Oscar Berger. Sky-high extra-systemic electoral volatility remained undiminished, as four newly created parties captured an enormous 42 percent of the vote. The GANA tripartite coalition of parties was assembled together to serve as Berger’s electoral vehicle. In line with previous (post-1985) democratic experience, the new GANA government suffered a massive decrease in public support after only a few months in office. The 2007 general elections brought about yet another overhaul of the party universe, including a drastic decrease in support for the ruling GANA, the meteoric rise of general Otto Perez Molina’s Partido Patriota and the relegation to irrelevance of the once-prominent parties PAN and FRG. After more than two decades of electoral democracy, no governing party has been re-elected to office, in what can be called the ‘golden rule’ of Guatemalan politics. Since the onset of competitive elections, Guatemala has displayed one of the very highest levels of overall volatility in the Latin American region and the developing world. Ominously, there has been no discernible trend towards lower levels of overall or extra-systemic volatility through time – which would allow the party universe to coalesce around a core of main parties. The process of ‘creative destruction’ of parties at every electoral cycle is self-sustaining and there are few good theoretical reasons to believe it may slow down or come to a halt. The perpetuation of this phenomenon partly reflects extremely low levels in other standard measures of party institutionalization: party rootedness in society and depth of internal party organization (Sanchez, 2008b). Generally speaking, Guatemalan parties’ claim to societal representation is close to null (ASIES, 2002; Sieder et al., 2002). Civil society groups (including organized business) and popular organizations regularly circumvent parties in their interaction with the state. This elicits the question: who do Guatemalan parties represent? It is only a slight simplification to affirm that, in essence, they solely represent their leaders and their financiers. At the centre of the enterprise of party creation lies private capital, which finances candidates in search of public favours and pursuit of rentseeking agendas. For citizens, party labels count little. But neither do they mean much for politicians. A telling anecdote is that five of the last six Guatemalan presidents have belonged to two or more parties.

Ecuador, Bolivia and Peru: Inchoate Systems or Non-Systems? Three of the clearest examples of under-institutionalized party systems in Latin America (Ecuador, Bolivia and Peru) are analysed here to further 506

S A N C H E Z : PA RT Y N O N - S Y S T E M S

illuminate what differentiates inchoate party systems from non-systems. The fact that these countries display some of the highest levels of electoral volatility in the entire region is prima facie evidence of acute party system under-institutionalization (see Table 4). Yet this standard aggregate measure hides as much information as it reveals. Is the nature of inter-party competition in these polities similar? The following section shows that it is not in one critical respect: the degree to which party politics (and by extension national political life) is anchored and structured around a group of core parties through time. Comparing extra-systemic volatility levels in these four countries reveals important differences. Guatemala and Peru are countries where there has been a constant turnover in the identity of the main parties in the polity, as shown by their high average extra-systemic volatility (see Table 5). They therefore qualify as non-systems. By contrast, Ecuador and Bolivia show much lower levels of extra-systemic volatility. The difference is particularly dramatic if we factor out the recent transformational elections in both countries (2002 and 2006 in Ecuador and 2005 in Bolivia). Only time will reveal whether those elections signify a movement towards a new, reconstituted party system or, rather, towards party non-system status. Ecuador Ecuadorean parties have been derided for their clientelistic character, personalistic appeals, pervasive corruption, non-ideological nature and marked incapacity to represent society (Conaghan, 1995). Fairly or not, they have been commonly blamed for many of the political and economic problems. Ecuador also displays one of the most volatile and fragmented party universes

Table 4. Electoral volatility in Latin America (selected countries)

Time-span Costa Rica Chile Argentina Paraguay Colombia El Salvador Panama Bolivia Venezuela Ecuador Guatemala Peru

Mean volatility*

1978–98 1989–97 1983–99 1989–98 1978–98 1985–2000 1994–99 1980–97 1978–2000 1989–98 1985–99 1980–2000

11.81 11.92 18.04 20.39 21.44 22.40 23.96 31.47 33.01 36.60 48.66 49.66

Source: Payne et al. (2003: 132). *Mean volatility of parliamentary and presidential elections.

507

PA RT Y P O L I T I C S 1 5 ( 4 )

Table 5. Extra-systemic volatility of top four parties (parliamentary elections) (selected countries)

Guatemala

Ecuador1

Bolivia

Peru

25.1 (1985–90) 35.2 (1990–95) –0.6 (1995–99) 63.8 (1999–03) 33.2 (2003–07)

35.6 12.8 6.2 9.4 2.3 52.4 11.6

6.3 13.8 –3.5 11.4 20.0

25.3 67.7 23.4 67.3 56.0

(1979–84) (1984–88) (1988–92) (1992–96) (1996–98) (1998–02) (2002–06)

(1985–89) (1989–93) (1993–97) (1997–02) (2002–06)

Average: 31.4 Average: 18.6 Average: 9.6 (Party non-system) Av. pre-2002: 13.2 Av. pre-2002: 7.0

(1980–85) (1985–90) (1990–95) (1995–01) (2001–06)

Average: 47.9 (Party non-system)

Source: Author’s calculations based on electoral data from Dieter Nohlen (ed.) Elections in the Americas: A Data Handbook, vols. 1 and 2 (2005). Note: Extra-systemic volatility, as I have chosen to operationalize it here, measures the difference between the collective share (percentage) of votes gathered by the top four vote-getters at a given election and their share at the subsequent election. A positive sign indicates the main parties in the polity as a collective have lost votes to small parties and new parties; a negative sign indicates the opposite. 1. If the top four chosen are the ones that have structured the system (ID, PSC, DP and PRE) levels of extra-systemic volatility are negative for much of this period, indicating party system institutionalization: 35.6 (1979–84); –11.0 (1984–88); –6.4 (1988–92); –1.1 (1992–96); –9.0 (1996–98); 54.6 (1998–02); –12.2 (2002–6).

in Latin America. Yet, this overall picture of party system weakness hides a considerable degree of party rootedness (Freidenberg and Alcantara, 2001). Not least, it hides underlying continuity as regards the main parties that compose the Ecuadorean political system during the post-1979 democratic era. From the mid-1980s the party system became anchored around four main parties that have succeeded each other in government – the SocialChristian Party (PSC), the Rodolsista Party (PRE), the Democratic Left (IU) and Popular Democracy (DP). In fact, the key, structuring role of these parties in the political system became more evident through time (see Table 6). Until 2002, there were signs of increased party system institutionalization (Pachano, 2004). Their combined share of the legislative vote steadily rose from the mid-1980s to the late 1990s. The presidency alternated among these four parties from 1981 until 2002. During that time-span the identity of the three or four top parties was invariably a combination of the aforementioned party quartet (see Table). While overall volatility was very high since the 1979 transition to democracy, such volatility was mainly of the benign intrasystemic kind: votes shifted among ID, PRE, PSC and PD, but not to other electoral options. Ecuadorean politics would have predictably been more unstable (and its praetorian qualities even more accentuated) in the absence of a core of main parties. Perhaps there is no better evidence of the role of this quartet of 508

S A N C H E Z : PA RT Y N O N - S Y S T E M S

Table 6. Ecuador: top four parties in first-round presidential (P) and congressional (C) elections (in order of votes gathered)

1978 1984 1988 1992 1996 1998 2002 2006

(P)

(C)

PUR, CFP, PSC, PLR ID, PSC, PUR, CFP ID, PRE, PSC, DP PUR, PSC, PRE, ID PSC, PRE, NP, DP DP, PRE, ID, Pachacutik PSP/Pachacutik, PRIAN, RP, ID PRIAN, Alianza PAIS, PSP, ID

CFP, ID, PSC, PLRE ID, PSC, PD, CFP ID, PRE, PSC, DP-UDC PSC, PUR, PRE, ID PSC, PRE, MUPP-NP, DP-UDC PSC, DP-UDC, PRE, ID PSC, PRE, ID, PRIAN PRIAN, PSP, PSC, ID

Note: Parties that have structured the party universe are listed in bold face. Key: CFP (Concentracion de Fuerzas Populares); PLRE (Partido Liberal Radical Ecuatoriano); PUR (Partido Unidad Republicana); MUPP-NP (Movimiento Unidad Plurinacional Pachacutik-Nuevo Pais); PRIAN (Partido Renovador Institucional Accion Nacional); PSP (Partido Sociedad Patriotica 21 Enero).

parties in structuring Ecuadorean politics than the prevalence of (hidden) coalitional dynamics in the legislature. Inter-party ‘ghost coalitions’, based on the exchange of ‘pork’ and patronage for congressional support, have greased the wheels of decision-making (Mejia-Acosta, 2006), allowing a degree of governability. These informal elite power-sharing agreements rested upon an implicit inter-temporal ‘pact’ among the four main parties. The forging of this informal institution would have proved elusive but for the electoral consolidation of this ‘party quartet’. The 2002 and 2006 electoral contests could well prove to be ‘critical elections’, i.e. votes which produce a ‘sharp and durable realignment between parties’ (Key, 1955). But, alternatively, they may signal the onset of structural and persistent dealignment, or, to put it another way, of non-system dynamics. The combined vote-share of the traditional parties dropped to 53 percent in the 2002 presidential elections. The presidential victory of a political outsider in the figure of Lucio Gutierrez, the first time one of the four main parties had not won the presidential contest since the transition to democracy, led to predictions that the end of the old party system was near. The October 2006 presidential election lent support to this view, with the ‘traditional’ four parties witnessing a marked continued decline in popular support. The electoral vehicles of banana magnate Alvaro Noboa (PRIAN), university economist Rafael Correa (Alianza PAIS) and military officer Gilmar Gutierrez (PSP) emerged as the big winners, displacing the party quartet from political centre stage. In summary, in spite of parties’ limited societal rootedness and the very high levels of volatility, the continuity in the Ecuadorean party universe in the post-1978 democratic era was unmistakable. While clearly inchoate, it unambiguously qualifies as a party system. However, in recent times the ‘traditional’ parties have become marginal, the party system has lost all 509

PA RT Y P O L I T I C S 1 5 ( 4 )

signs of cohesion and extra-systemic volatility has increased dramatically. An era of persistent vote-switching may become entrenched, with the PRIAN and the PSP unable to consolidate their positions as the new systemic parties. Another real possibility, given recent developments, is that enough voters may fall under the spell of President Rafael Correa and the country may drift into competitive authoritarianism, which would bode ill for the future crystallization of the party system. (The results of the 2007 elections for the Correa-sponsored Constituent Assembly point in this direction.) If either of these scenarios materializes, a party non-system can ensue. Bolivia From 1985 until 2002, Bolivia’s party system was firmly anchored around three main parties – the centre-right Movimiento Nacionalista Revolucionario (MNR), the right-wing Accion Democratica y Nacionalista (ADN) and the centre-left Movimiento de Izquierda Revolucionario (MIR) – each of which became the pillars of successive coalition governments. During this time, no other parties were able to dislodge them atop the party system (see Table 7). The charge of an ‘exclusive partytocracy’ was not unfairly levelled against this triumvirate of parties (Gamarra and Malloy, 1995). The combined vote-share of the Bolivian ‘party triumvirate’ was fairly steady from the mid-1980s to the late 1990s. To be sure, these parties were weakly institutionalized: none of them had stable followings along sectoral or class divisions, exhibited clear programmatic or ideological identities, articulated the interests of civil society, or otherwise developed strong roots in society (notwithstanding the MNR’s pivotal role in the country’s modern history). All three parties have exhibited high levels of personalization and, Table 7. Bolivia: top four parties in first-round presidential (P) and congressional (C) elections

(P)

(C)

1997

UDP, MNR ADN, PS-1 UDP, MNR ADN, PS-1 ADN, MNR, MIR, MNR-I MNR, ADN MIR, CONDEPA MNR, ADN-MIR, CONDEPA, UCS ADN, MNR, CONDEPA, MIR

2002 2005

MNR, NFR, MAS, MIR MAS, Podemos, FUN, MNR

UDP, MNR, ADN, PS-1 UDP, MNR, ADN, PS-1 ADN, MNR, MIR, MNR-I MNR, ADN, MIR, CONDEPA MNR, AP (ADN-MIR), CONDEPA, UCS ADN-NFR-PDC, MNR, CONDEPA, MIR MNR-NBL, MAS, NFR, MIR-FRI MAS, Podemos, FUN, MNR

1979 1980 1985 1989 1993

Note: Systemic parties appear in bold face. Key: UDP (Unidad Democratica y Popular, coalition of MIR and Communist Party); UCS (Unidad Civica Solidaridad); CONDEPA (Conciencia de Patria); MAS (Movimiento al Socialismo); Podemos (Poder Democratico y Social); FUN (Frente de Unidad Nacional).

510

S A N C H E Z : PA RT Y N O N - S Y S T E M S

notably, have been patrimonial at their core – vehicles to capture and circulate state patronage. In light of these characteristics, it is not surprising that electoral volatility in Bolivia’s neodemocracy has been high, even by regional standards. However, the more seldom remarked fact is that this volatility was intra-systemic in nature, meaning that the bulk of vote-shifting took place among the three main parties – and not transferred in significant quantities to outside political formations. From 1985 onwards, Bolivian party politics acquired a centripetal tendency around a hegemonic centre. This phase, which lasted until 1993, was characterized by a logic of party pacts. Most governing party coalitions involved either the ADN/MIR, on the one hand (such as the 1989 Acuerdo Patriotico), or the MNR allied with satellite parties, on the other. This dynamic of pacts and inter-party cooperation deepened and widened in 1993 with the Pacto por la Democracia forged by the ADN and the MNR, whereby these two parties agreed to adopt a common posture vis-à-vis democracy and structural adjustment. The sweeping economic and political reforms undertaken throughout the 1980s and 1990s were the result of explicit party ‘summits’ among the parties with the largest legislative blocs. Political polarization along the left–right spectrum declined and a hegemonic agenda around ‘democratic neoliberalism’ became consolidated (Mayorga, 2004). In short, a system of moderate multipartism emerged, characterized by low polarization, coalitional politics and a centripetal dynamic. Party system stability – amid much instability in other realms – came rightly to be considered one of the greatest achievements of Bolivia’s young democracy. Indeed, without the consolidation of a group of core, systemic parties, the observed phenomena of inter-party cooperation, explicit pacts and consensus-based reform programmes would have proved elusive. The irruption and rise of neopopulist parties in the 1990s (CONDEPA and UCS) did not change the prevailing political dynamic, as these parties joined the logic of inter-party pacts. The party system began to unravel around 2000 when new, potent social and ethnic demands rocked the political status quo (Barr, 2005). Riding the tidal wave of rejection of the traditional parties, Evo Morales’s Movimiento al Socialismo (Movement Toward Socialism or MAS) came to power by questioning representative democracy and advocating more direct forms of democratic governance. The last two general electoral contests (2002 and 2005) clearly constitute transformational elections in that they have ushered in the demise of the old party system (see Table 7). The old triumvirate has disappeared and the party landscape is entirely novel. Only time will tell whether the post-2002 party universe evolves into a new identifiable party system or drifts into a non-system. The MAS, as a decade-old grassroots organization with a clear social campesino (peasant) base and a defined social and economic agenda, has the potential to become institutionalized. But the completion of its transformation from social movement to a bona fide nationwide party is fragile and surely not preordained. 511

PA RT Y P O L I T I C S 1 5 ( 4 )

In summary, it is clear that, notwithstanding very high levels of overall volatility, Bolivia’s political universe from 1985 to 2002 displayed continuity in its main party pillars and unambiguously qualifies as an (inchoate) party system. However, in recent years, the vanishing of the traditional parties coupled with unmistakable signs of increased societal alienation towards party politics portends a potential transition towards non-system status. Peru From the elections for the Constituent Assembly of 1978 and throughout the 1980s, the Peruvian party system was anchored upon four main parties: on the right, Popular Action (AP) and the Popular Christian Party (PPC); on the left, the United Left (IU); and at the centre the populist American Popular Revolutionary Alliance (APRA) (Cotler, 1995). While Peru’s 1980s’ party system cannot be said to have been well institutionalized, all of these parties could trace their roots to country-specific groups and movements of the past, all had discernible ideologies, national structures and fairly clear social bases. The electoral hegemony of the ‘traditional’ parties was uncontested throughout the 1980s: from 1978 until the first round of the 1990 presidential election, they together captured more than 70 percent of the votes (and over 90 percent in most of them). While volatility was undeniably high, it was of the more benign intra-systemic sort – keeping alternative political groups and independent outsiders on the margins of party politics (see Table 8). Table 8. Peru: top four parties in presidential (P) and congressional (C) elections, 1978–2006 (in order of vote received)

1978*

1980

1985

1990

1995

2001

2006

AP, APRA, IU, PPC

APRA, IU, PPC, AP

FREDEMO, C90/NM, APRA, IU

C90/NM, UPP, APRA, CD-PP

PP, APRA, UN, FIM

UPP, APRA, UN, AF

(C)

1980

1985

1990

1995

2001**

2006

APRA, IU, PPC, Frenatraca

AP, APRA, PPC, UNIR

APRA, IU, CODE, AP

FREDEMO, APRA, C90/NM, IU

C90/NM, UPP, APRA, FIM

PP, APRA, UN, FIM

UPP, APRA, UN, AF

(P)

Note: Systemic parties appear in bold face. *Constituent Assembly elections. **Fraudulent presidential elections were held in 2000. After Fujimori’s resignation, new elections were scheduled for 2001. Key: APRA (Partido Aprista Peruano); AP (Accion Popular); PPC (Partido Popular Cristiano); C90/NM (Cambio 90/Nueva Mayoria); UNIR (Union de Izquierda Revolucionaria); UPP (Union por el Peru); PP (Peru Posible); CODE (Convergencia Democratica); FIM (Frente Independiente Moralizador); UPP (Union por el Peru); UN (Unidad Nacional); AF (Alianza por el Futuro, includes C90/NM).

512

S A N C H E Z : PA RT Y N O N - S Y S T E M S

The traditional parties became widely discredited by the late 1980s for a whole host of reasons, and voters dealt them a major blow at the 1990 general elections. But it was Alberto Fujimori’s successful 1992 self-coup that signified the collapse of the extant Peruvian party system. Fujimori’s power-grabbing agenda included the deliberate weakening of the ‘traditional parties’ in his quest to rule without constraints. It is in this light that one can best interpret a number of executive-driven institutional reforms6 designed to open up the political system to independent and anti-party candidates and organizations, and the establishment of clientelistic relations with those new actors (Tanaka, 2005). The decade-long era of Fujimorismo (1990–2000) represented a classic case of competitive authoritarianism: the president conquered and co-opted the main institutions of the state, established his political movement as the hegemonic actor, effectively erased checks and balances upon executive power and limited political competition to undemocratic levels. President Fujimori painstakingly built a thoroughly corrupt political regime, where his operatives co-opted, bought off, harassed, intimidated or persecuted political opponents in all relevant state institutions (Conaghan, 2005). The political environment became adverse for the survival (let alone institutionalization) of the existing political formations. Indeed, they had little prospect of building effective roots in society. It is thus not surprising that there was constant change in the identity of the main Peruvian ‘parties’ (de facto electoral vehicles) throughout the 1990s: in 1990, the right-wing FREDEMO, APRA and Fujimori’s taxi party Cambio 90 accumulated close to three-quarters of the vote to the chamber of deputies; in 1995, it was Cambio 90, UPP and APRA that totalled 72.6 percent of the vote together; in 2000, other taxi parties dominated the elections, in particular, Fujimori’s renamed Peru 2000, Alejandro Toledo’s Peru Posible and FIM. Fujimori decimated an already debilitated party system, while making no attempt at building a veritable political party of his own. Relying on personalistic linkages to his mass base (Roberts, 2006), rather than establishing an institutionalized relationship, he created as many as four different (disposable) electoral vehicles to contest different elections (Change 90 in 1990, New Majority in 1995, Let’s go Neighbours in 1998 and Peru 2000 in his illegal second re-election bid). Unexpectedly, the 2001 presidential election saw the resurrection of politicians associated with the 1980s’ party system, including Alan Garcia (APRA), Flores Lourdes Nano (PPC, UN) and Diez Canseco (IU). These candidates together won more than half of the first-round presidential vote, and two of Peru’s traditional parties (APRA and AP) won almost a quarter of votes in the legislative election (Kenney, 2003). In 2006, Alan Garcia and APRA made an astonishing political comeback after years in the political wilderness by winning the presidential contest. Yet, a careful examination reveals that this does not amount to a resurgence of the traditional parties per se as much as the conjunctural resurgence of particular leaders. It is clear that the legacy of a decade of competitive authoritarianism under 513

PA RT Y P O L I T I C S 1 5 ( 4 )

Fujimorismo (1990–2000) has left deep scars that will continue to undermine party system institutionalization for the foreseeable future. The lesson political entrepreneurs drew in the 1990s was that the building of bona fide parties with an organizational base was dispensable in the quest for influence and power. In a real sense, democracy was restored with the 2001 general elections and during Toledo’s presidency, but the party system has not been re-established. Given the acute weakness of social and political actors since 1990, continued extreme volatility and change in the identity of the top parties can be expected. Judged by the criteria delineated in this article, Peru’s party universe during the 1980s qualifies as a party system, but during the past two decades it clearly fulfils the criteria of a non-system – and, like Guatemala’s, one with many self-perpetuating qualities.

Conclusions The conceptual innovation advanced here is that of a new party universe type: party non-systems. The main theoretical claim made in this study is that constellations of parties do not make a ‘system’, properly understood, when there is a fundamental lack of inter-temporal continuity in the identity of the main parties of the polity. Distinguishing between inchoate party systems and non-systems can profitably add to the descriptive and analytical toolbox of comparativists. The adoption of the concept of party non-systems allows for differentiation between party universes where volatility is largely intra-systemic and those where it is also substantially extra-systemic – with consequences for the quality and performance of democracy. Such a distinction has yet to be conceptualized or operationalized. Using Sartori’s ladder of generality scheme, I have argued that non-systems differ from all other party universes not only in degree (of institutionalization) but in kind, as they transgress the land of ‘systemness’ into a chaotic terrain of fundamental discontinuity marked by the absence of core, systemic parties. As explained, there are both empirical and theoretical grounds to think such party universes may become more, rather than less, common in the foreseeable future. This article attempts to order and classify underdeveloped party systems’ polities according to the relative mix in the type of vote-switching they exhibit – rather than focusing simply on overall electoral volatility as an indicator of institutionalization. The locus of continuity and discontinuity matters. There are strong theoretical reasons to believe that extra-systemic electoral volatility is more malign than its intra-systemic cousin. When it is sustained, extra-systemic volatility spawns highly deleterious effects: national politics is unanchored and coalitional dynamics virtually unattainable; uncertainty of outcomes is excessive; inter-party competition has a spurious quality; accountability, representation and responsiveness are all fatally damaged; and policy stability proves elusive. Important differences in main party continuity are found when comparing party systems generally grouped 514

S A N C H E Z : PA RT Y N O N - S Y S T E M S

simply as ‘inchoate’. In this important respect, the Guatemalan and (post1990) Peruvian party universes, which qualify as non-systems, are found to be dissimilar to those of Bolivia or Ecuador, for instance. Moreover, important inter-temporal changes in (core) party continuity are detected within polities. Our current conceptual lexicon is inadequate to the task of identifying such changes through time. I have contended that the ‘inchoate’ label is excessively broad, allowing significant intra-category variation and an attendant loss of connotative precision. As the study of Third World party systems expands and the degree of institutionalization comes to be considered the most important dimension along which such systems differ, the need to establish more finelygrained distinctions along this dimension becomes more urgent.

Notes I thank Jana Morgan, Scott Mainwaring and three anonymous referees for comments on previous versions of this article. 1 The chief advantage of establishing relevant categories is that it permits greater connotative precision and analytical differentiation, while it also allows for the identification of relevant party system transformations. Party system fragmentation can be analysed both as a continuum (using the Laakso–Taagepera index) that can be measured to decimal points, as well as in terms of discrete categories (ranging from dominant party systems to multiparty systems). The coexistence of these two approaches is not incongruent. Each has its scholarly uses. Institutionalization as a relevant analytical variable can similarly benefit from a dual approach. 2 The few exceptions to this generalized assumption include Sartori (1976), Coppedge (1998) and Mainwaring (1999). Mainwaring (1999: 24) explicitly suggests Peru as a possible case of party constellation that does not qualify as a party system. 3 Perhaps the two most common phenomena associated with party institutionalization are ‘value infusion’ (organizations are valued intrinsically, beyond their ability to achieve certain objectives) and ‘behavioral routinization’ (internal party rules and norms become entrenched). Levitsky (1998) cogently argues that these two phenomena do not necessarily vary together. 4 Other ways in which institutionalization has been measured include party age, counts of splits and mergers or on scales built from measures of age, legislative stability and leadership change. Simpler measures have included minimum durability (as in how many elections a party contests) (Rose and Mackie, 1988). 5 All conceptual innovations involve some change or restructuring of the existing conceptual lexicon. This is indeed the way in which the term ‘party universe’ is used throughout this article, as the overarching term that encompasses all party constellations – including non-systems. Correspondingly, the term ‘party system’, as used in this article, is slightly less inclusive. 6 Some of these included: the introduction of methods of direct and participative democracy, the creation of spaces alternative to formal representative democracy and electoral rules that favoured the representation of minorities.

515

PA RT Y P O L I T I C S 1 5 ( 4 )

References Abal Medina, Juan Manuel and Marcelo Cavarozzi (2002) El Asedio a la Política: los Partidos Latinoamericanos en la Era Neoliberal. Rosario: Homo Sapiens. Achard, Diego (2004) ‘Un desafío a la democracia: los partidos políticos en Centroamérica, Panamà y República Dominicana’, in Diego Achard and Luis E. González, con la colaboración de Edelberto Torres-Rivas y otros. San José, C.R.: BID, IDEA y OEA. Alcantara Saez, Manuel and Flavia Freidenberg (2003) Partidos Politicos de America Latina: Centroamérica, México y República Dominicana, vol. 2. Fondo de Cultura Economica. Artiga-Gonzalez, Alvaro (2000) La Politica y los Sistemas de partidos en Centroamerica. San Salvador: FundaUngo. ASIES (2002) ‘Diagnóstico del Funcionamiento del Sistema de Partidos Políticos en Guatemala Revista’, ASIES, no. 1. Barr, Robert R. (2005) ‘Bolivia: Another Uncompleted Revolution’, Latin American Politics & Society 47: 69–90. BID (Banco Inter-Americano de Desarrollo) (2002) Democracies in Development: Politics and Reform in Latin America. Washington, DC: Johns Hopkins University Press. Bielasiak, Jack (2002) ‘The Institutionalization of Electoral and Party Systems in Postcommunist States’, Comparative Politics (January): 189–210. Booth, John and Mitchell Seligson (eds) (1995) Elections and Democracy in Central America, 2nd edn. Chapel Hill, NC: University of North Carolina Press. Carey, John and Andrew Reynolds (2007) ‘Parties and Accountable Government in New Democracies’, Party Politics 13: 255–74. Cerdas Cruz, Rodolfo (1996) ‘Political Parties and Party Systems’, in Rachel Sieder (ed.) Central America: Fragile Transition. London: Institute of Latin American Studies, University of London. Collier, David and Steven Levitsky (1997) ‘Democracy with Adjectives: Conceptual Innovation in Comparative Research’, World Politics 49: 430–51. Collier, David and James E. Mahon (1993) ‘Conceptual Stretching Revisited: Adapting Categories in Comparative Analysis’, American Political Science Review 87 (December): 845–55. Conaghan, Catherine (1995) ‘Politicians Against Parties: Discord and Disconnection in Ecuador’s Party System’, in Scott Mainwaring and Timothy Scully (eds) Building Democratic Institutions: Party Systems in Latin America. Stanford, CA: Stanford University Press. Conaghan, Catherine (2005) Fujimori’s Peru: Deception in the Public Sphere. Pittsburgh, PA: Pittsburg University Press. Coppedge, Michael (1998) ‘The Evolution of Latin American Party Systems’, in Scott Mainwaring and Arturo Valenzuela (eds) Politics, Society and Democracy: Latin America. Boulder, CO: Westview Press. Coppedge, Michael (2001) ‘Party Systems, Governability and the Quality of Democracy in Latin America’. Paper prepared for presentation at the conference on ‘Representation and Democratic Politics in Latin America’ Buenos Aires, Argentina, 7–8 June 2001. Coppedge, Michael (2002) ‘Explaining Democratic Deterioration in Venezuela

516

S A N C H E Z : PA RT Y N O N - S Y S T E M S

Through Nested Inference’. Paper presented at the Congreso Latinoamericano de Ciencia Politica, Salamanca, Spain, 9–11 July 2002. Corrales, Javier (2005) ‘Expresidents and Newcomers Running for President . . . and Winning: Recent Elections in Latin America’. Paper presented at the Annual Conference of the Society for Latin American Studies, Universiteit Leiden, Leiden, The Netherlands. Cotler, Julio (1995) ‘Political Parties and the Problem of Democratic Consolidation in Peru’, in Mainwaring and Scully (eds) Building Democratic Institutions: Party Systems in Latin America. Stanford, CA: Stanford University Press. Daly, Siobhan (2003) ‘The Ladder of Abstraction: A Framework for the Systematic Classification of Democratic Regime Types’, Politics (Political Studies Association) 23: 96–108. Dietz, Henry and David Myers (2007) ‘From Thaw to Deluge: Party System Collapse in Venezuela and Peru’, Latin American Politics and Society 49: 59–86. Freidenberg, Flavia and Manuel Alcantara (2001) Los Duenos del Poder: Partidos Politicos en Ecuador 1978–2000. Quito: FLACSO. Gamarra, Eduardo and James Malloy (1995) ‘The Patrimonial Dynamics of Party Politics in Bolivia’, in Mainwaring and Scully (eds) Building Democratic Institutions: Party Systems in Latin America. Stanford, CA: Stanford University Press. Garreton, Manuel Antonio, Marcelo Cavarozzi, Peter Cleaves, Gary Gereffi and Jonathan Hartlyn (2003) Latin America in the 21st Century: Toward a New Sociopolitical Matrix. Miami: North-South Center Press. Geddes, Barbara (2004) ‘The Development of Party Systems in Latin America’. Paper prepared for presentation at the Western Political Science Association Meeting, Portland, March 2004. Goodman, Louis (1992) ‘Political Parties and Political Systems of Central America’, in Louis Goodman, William Leogrande and Johanna Mendelson (eds) Political Parties and Democracy in Central America. Boulder, CO: Westview Press. Hagopian, Frances (1998) ‘Democracy and Political Representation in Latin America in the 1990s: Pause, Reorganization, or Decline?’ in Felipe Aguero and Jeffrey Stark (eds) Fault Lines of Democracy in Post-Transition Latin America. Miami, FL: North-South Center Press. Huntington, Samuel (1968) Political Order in Changing Societies. New Haven, CT: Yale University Press. IADB (Inter-American Development Bank) (2005) The Politics of Policies: Economic and Social Progress in Latin America, 2006 Report. Washington, DC: IABD. Janda, Kenneth (1993) ‘Comparative Political Parties: Research and Theory’, in A. Finifter (ed.) Political Science: State of the Discipline II. Washington, DC: American Political Science Association. Jonas, Susanne (1991) The Battle for Guatemala: Rebels, Death Squads and US Power. Boulder, CO: Westview Press. Jonas, Susanne (1995) ‘Electoral Problems and the Democratic Project in Guatemala’, in John Booth and Mitchell Seligson (eds) Elections and Democracy in Central America. Chapel Hill, NC: University of North Carolina Press. Kenney, Charles (2003) ‘The Death and Rebirth of a Party System. Peru 1978– 2001’, Comparative Political Studies 36: 10: 1210–39. Key, V. O. (1955) ‘A Theory of Critical Elections’, Journal of Politics 17: 3–18. Levitsky, Steven (1998) ‘Peronism and Institutionalization: the Case, the Concept, and the Case for Unpacking the Concept’, Party Politics 4: 77–92.

517

PA RT Y P O L I T I C S 1 5 ( 4 )

Levitsky, Steven and Lucan Way (2002) ‘The Rise of Competitive Authoritarianism’, Journal of Democracy 13: 51–65. Levitsky, Steven and Maxwell Cameron (2003) ‘Democracy without Parties? Political Parties and Regime Change in Fujimori’s Peru’, Latin American Politics and Society 45. Lijphart, Arend (1984) Democracies: Patterns of Majoritarian and Consensus Government in Twenty One Democracies. New Haven, CT: Yale University Press. Mainwaring, Scott (1999) Rethinking Party Systems in the Third Wave of Democratization: the case of Brazil. Stanford, CA: Stanford University Press. Mainwaring, Scott and Timothy Scully (1995) Building Democratic Institutions: Party Systems in Latin America. Stanford, CA: Stanford University Press. Mainwaring, Scott and Mariano Torcal (2006) ‘Party System Institutionalization and Party System Theory after the Third Wave of Democratization’, in William Crotty (ed.) Handbook of Political Parties. London: Sage. Mainwaring, Scott, Ana Maria Bejarano and Eduardo Pizarro Leongomez (eds) (2006) The Crisis of Democratic Representation in the Andes. Stanford, CA: Stanford University Press. Mainwaring, Scott and Zoco, E. (2007) ‘Political Sequences and the Stabilization of Party Competition: Electoral Volatility in Old and New Democracies’, Party Politics 13(2): 155–78. Mair, Peter (1997) Party System Change: Approaches and Interpretations. Oxford: Clarendon. Mayorga, Jose Antonio (2004) ‘La Crisis del Sistema de Partidos: Causas y Consecuencias. Caso Bolivia’, in Partidos Politicos en la Region Andina: entre la Crisis y el Cambio. Lima: International Institute for Democracy and Electoral Assistance (IDEA). McDonald, Ronald and J. Mark Ruhl (1989) Party Politics and Elections in Latin America. Boulder, CO: Westview Press. Mejia-Acosta, Andres (2006) ‘Crafting Legislative Ghost Coalitions in Ecuador’, in Gretchen Helmke and Steven Levitsky (eds) Informal Institutions and Democracy: Lessons from Latin America. Baltimore, MD: Johns Hopkins University Press. Molina, José (2001) ‘The Electoral Effect of Underdevelopment: Government Turnover and its Causes in Latin American and industrialized Countries. Electoral Studies 20: 427–46. Morgan, Jana (2007) ‘Partisanship During the Collapse of Venezuela’s Party System’, Latin American Research Review 42: 78–98. Morlino, Leonardo and Marco Tarchi (1996) ‘The Dissatisfied Society: the Roots of Political Change in Italy’, European Journal of Political Research 30: 5–30. Nohlen, Dieter (ed.) (2005) Elections in the Americas: A Data Handbook, vol. 1 (Central America) and vol. 2 (South America). Oxford: Oxford University Press. Olson, David (1998) ‘Party Formation and Party System Consolidation in the New Democracies of Central Europe’, Political Studies 46: 432–64. Pachano, Simon (2004) ‘El Territorio de los Partidos. Ecuador’, in Partidos Politicos en la Region Andina: entre la Crisis y el Cambio. Lima: International Institute for Democracy and Electoral Assistance (IDEA). Panebianco, Angelo (1988) Political Parties: Organization and Power. Cambridge: Cambridge University Press. Payne, Mark, Daniel Zovatto, Fernando Carillo Florez and Andres Allamand (2003)

518

S A N C H E Z : PA RT Y N O N - S Y S T E M S

La Democracia Importa: Democracia y Desarrollo en America Latina. Washington, DC: Banco Interamericano de Desarrollo. Powell, Bingham (2005) ‘The Chain of Responsiveness’, in Larry Diamond and Leonardo Morlino (eds) Assessing the Quality of Democracy. Baltimore, MD: Johns Hopkins University Press. Randall, Vicky and Lars Svasand (2002) ‘Party Institutionalization in New Democracies’, Party Politics 8: 5–29. Riker, William and Peter Ordeshook (1973) An Introduction to Positive Political Economy. Englewood Cliffs, NJ: Prentice-Hall. Roberts, Kenneth (2006) ‘Are Parties Necessary for Democracy? Lessons from the Peruvian Experience’, in Julio Carrión (ed.) The Fujimori Legacy. University Park: Pennsylvania State University Press. Roberts, Kenneth (forthcoming) Changing Course: Party System Change in Latin America’s Neoliberal Era. Cambridge: Cambridge University Press. Roberts, Kenneth and Eric Wibbels (1999) ‘Party Systems and Electoral Volatility in Latin America: A Test of Economic, Institutional and Structural Explanations’, American Political Science Review 93: 575–90. Rose, Richard and Thomas T. Mackie (1988) ‘Do Parties Persist or Fail? The Big Trade-off Facing Organizations’, in Kay Lawson and Peter H. Merkl (eds) When Parties Fail: Emerging Alternative Organizations, pp. 533–58. Princeton, NJ: Princeton University Press. Rose, Richard and Neil Munro (2003) Elections and Parties in New European Democracies. Washington, DC: CQ Press. Sabatini, Christopher and Eric Farnswoth (2006) ‘The Urgent Need for Labor Law Reform’, Journal of Democracy 17(October): 50–64. Sanchez, Omar (2008a) ‘Transformation and Decay: The De-Institutionalization of Party Systems in South America’, Third World Quarterly, 29(March): 315–37. Sanchez, Omar (2008b) ‘The Guatemalan Party Universe: A Case Study in Underinstitutionalization’, Latin American Politics and Society 50 (Spring). Sartori, Giovanni (1970) ‘Concept Misformation in Comparative Politics’, American Political Science Review LXIV(December): 1033–53. Sartori, Giovanni (1976) Parties and Party System: A Framework for Analysis. Cambridge: Cambridge University Press. Sartori, Giovanni (1984) Social Science Concepts: A Systematic Analysis. Beverly Hills, CA: Sage. Sartori, Giovanni (1991) ‘Comparing and Miscomparing’, Journal of Theoretical Politics 3: 243–57. Schedler, Andreas (1995) ‘Under and Overinstitutionalization: Some Ideal Typical Propositions Concerning New and Old Party Systems’. Working Paper no. 213. Notre Dame: Hellen Kellogg Institute for International Studies, University of Notre Dame. Schedler, Andreas (2006) Electoral Authoritarianism: The Dynamics of Unfree Competition. Boulder, CO: Lynne Rienner. Seawright, Jason (2004) ‘The Demand Side of Party System Collapse’. Workshop on the Analysis of Party Competition. Duke University, Department of Political Science, April 2/3. Seligson, Mitchell, Malcolm Young, Max Eduardo Lucas and Dinorah Azpuru (2002) La cultura política de los guatemaltecos en el nuevo siglo. Guatemala City: ASIES.

519

PA RT Y P O L I T I C S 1 5 ( 4 )

Sieder, Rachel, Megan Thomas, George Vickers and Jack Spence (2002) Who Governs? Guatemala Five Years After the Peace Accords. Cambridge, MA: Hemisphere Initiatives. Skidmore, Thomas (1993) Television, Politics and the Transition to Democracy in Latin America. Baltimore, MD: Johns Hopkins University Press. Stokes, Susan (2001) Mandates and Democracy: Neoliberalism by Surprise in Latin America. Cambridge: Cambridge University Press. Tanaka, Martin (2005) ‘Peru 1980–2000: Chronicle of a Death Foretold? Determinism, Political Decisions, and Open Outcomes’, in Frances Hagopian and Scott Mainwaring (eds) The Third Wave of Democratization in Latin America: Advances and Setbacks. Cambridge: Cambridge University Press. Van Cott, Donna Lee (2005) From Movements to Parties: The Evolution of Ethnic Politics. New York: Cambridge University Press. Vellinga, Menno (ed.) (1998) The Changing Role of the State in Latin America. Boulder, CO: Westview Press. Ware, Alan (1996) Political Parties and Party Systems. Oxford: Oxford University Press.

OMAR SANCHEZ is Assistant Professor at Texas State University at San Marcos. His areas of research are Latin American politics, party systems and the politics of economic reform. He has recently published articles in Latin American Politics and Society, Third World Quarterly, Democratization and Bulletin of Latin American Research. Address: Texas State University, Department of Political Science, ELA 265, 600 University Drive, San Marcos, Texas 78666, USA. [email: [email protected]] Paper submitted 6 February 2007; accepted for publication 14 August 2007.

520