Illusions and Conceptual Flaws

Guillermo O'Donnell

Debate: Democratic Consolidation

The authors define democratic consolidation as "the achievement of substantial attitudinal support and behavioral compliance with the new democratic institutions and rules of the game they establish." ¹ Since "full consolidation" is an ultimately unattainable "ideal type" (p. 5 and rejoinder), it suffices that "all politically significant groups" (p. 7) provide such support and compliance. When this is the case, a democracy is "sufficiently consolidated." The authors locate the "dividing line" between "consolidated and unconsolidated democratic regimes" where "democratic regimes are sufficiently consolidated so as to survive and remain stable in the face of such serious challenges as major economic or international crises, or even serious outbreaks of terrorist violence" (p. 8, italics in original).

This approach has several flaws. It fails to distinguish between survival and stability as a definitional component of sufficient consolidation and as a consequence of the consolidation resulting from the attitudinal and behavioral acceptance of the regime (p. 3). It also fails to distinguish between democratic consolidation as a process and as a regime attribute achieved after "successful completion" of the former (p. 389). Lastly, if "all politically significant actors" accept a regime's [End Page 160] "rules of the game," it is not astounding news to assert that, other things being equal, this regime is consolidated--or likely to endure, or stable, or any equivalent idea. ²

The authors view consolidation as a several-stage process. It begins, and may overlap, with the transition from an authoritarian regime (pp. xii-xiii; we are not told how to distinguish processes pertaining to the one or the other). The democracy resulting from the transition may be simply unconsolidated, but it may also be "partially consolidated," or "substantially consolidated"; or it may be consolidated at the national but not the regional level--all this before eventually reaching "sufficient consolidation." ³ In turn, partial consolidation may mean that only some of the "partial regimes" ⁴ have consolidated (p. 410), or it may refer to a stage between nonconsolidation and sufficient consolidation. Nor need the latter stage be the end, for then may come "the stage of democratic persistence" (p. 413), which "represents the end product of a long democratization process" (p. xiii). Surprisingly, we read that at this stage "entirely new theoretical and practical concerns move to center stage, as the imperatives of transition and consolidation fade into the background of social inquiry" (p. 412). Presumably this applies in the cases of "Greece, Portugal, and Spain [which] are examples of consolidation and continuing [sic ] democratic persistence" (p. 413).

Still other possibilities exist: a consolidated or even a persistent democracy may deconsolidate and break down, or it may deconsolidate and reequilibrate, as in contemporary Italy (p. 15), or, in a different interpretation of this same case, see some of its partial regimes deconsolidate (p. 394)--in these cases we are not told if regimes are just reconsolidated or if they can return directly to the stage of persistence. Additionally, in categories even more casually drawn, a consolidated democracy may be "mired in its own contradictions and, hence [ sic ], become frozen" as a result of pacts (p. 406), or there may be "consolidated limited democracies" (p. 413), or even democracies that, puzzlingly, are "by and large consolidated" (p. 25).

I find all this confusing. How can we know when a democracy is substantially consolidated (i.e., no longer simply unconsolidated but not yet sufficiently consolidated), or substantially but "not yet" sufficiently consolidated, or partly consolidated, or when some of its partial regimes have consolidated or deconsolidated, or when it has moved from consolidation to persistence? Now I examine how the authors treat cases.

Cases and Indicators

Beyond their own assertions, the authors give no indication that would allow us to recognize when most of their "stages" have been reached. We do get some concrete references when they discuss Spain, but these lead to the further expansion of confusing and empirically [End Page 161] untraceable categories. We are told that in Spain before 1982-83, "substantial progress towards consolidation had been achieved" (p. 10), but that it "cannot be argued that full consolidation had been achieved by that time" (p. 11). ⁵ Then we read that Spanish democracy "has been substantially consolidated since about 1982 or 1983" (p. 21). At first sight, "substantial" would seem to be a synonym for "sufficient," yet it is not: the authors tell us that in Spain by 1982-83, consolidation had reached "completion at the national level" (p. 390), but this had not yet happened at the regional one, considering the antisystem opposition based in the Basque region (p. 11). ⁶ Further inklings that "substantial" and "sufficient" are not intended as synonyms come when we read that Greece (p. 31) and Italy were also cases of "substantial but incomplete levels of consolidation" (p. 22) before reaching a sufficient one. Furthermore, assuming that "incomplete" is equivalent to "partial" (otherwise the former would be still another "stage"), it seems that there is something more to substantial than to partial consolidation even if it still falls short of sufficient consolidation. Partial consolidation is not to be taken lightly, however, as it "constituted an essential resource that enabled the [Spanish] regime to survive [the] extreme test" of the 1981 coup attempt (p. 11). But surviving an "extreme test" is exactly the criterion that the authors propose for recognizing sufficient consolidation! How can we know, then, what is indicated by these "unequivocal empirical measures that have predictive value" (rejoinder)?

The authors' other criterion for sufficient consolidation, the absence of an important antisystem party or social movement, underlies the surprising argument that Italian democracy remained at "substantial but incomplete levels of consolidation" (p. 22) for about 30 years, in contrast to the speedy achievement of sufficient consolidation, and even democratic persistence, by the other three countries studied. ⁷

We are left without indications for recognizing most of the consolidation stages the authors envisage. Yet the authors' rejoinder denies that the first four criteria I transcribed from their book (p. 12) should be considered as indicators of (sufficient) consolidation. In their book they correctly assert that these criteria "should not be confused with the concept of consolidation itself"; they say on the same page that "the passing of a severe test . . . may constitute evidence that a regime is consolidated" (p. 12), adding that this and other kinds of information "should never be relied upon as the sole measure of a lack of [sufficient] consolidation" (p. 13, italics added). If one is looking for indicators of "sufficient consolidation," these seem reasonably good ones (although not the only ones), particularly considering the convenience of using a multi-indicator approach to map so complex and multidimensional a concept as "consolidation," whether sufficient or not. Although the authors' book proposes crisis stability as a criterion for (sufficient) consolidation, in their rejoinder they weaken their case by asserting that [End Page 162] they accept only "the absence of a politically significant antisystem party or social movement" as an adequate indicator of such a "stage." These two criteria may sometimes overlap, but they are different and may operate independently; using both gives the authors a stronger case. Let us see how these criteria operate.

In relation to Southern Europe, the authors are consistent when they consider the lack of antisystem actors, or regime survival or stability in the face of major crises, as prime indicators of sufficient consolidation. Yet their book explicitly intends to be broadly comparative, so we should look at how these criteria function in other areas of the world. Their picture is dismal. Only Uruguay and Costa Rica are considered (sufficiently, I gather) consolidated democracies. "Chile's transition seems to be well on its way to successful completion"; the rest of Latin America is "still struggling with transition problems of varying, and often major, magnitude and intensity"--but the situation in the postcommunist countries is "even worse" (p. 389). Elsewhere only South Korea "seems to have made considerable progress towards consolidation," while the other East Asian countries "still find themselves in various stages of their uncertain democratic transitions" (p. 390).

What are the grounds for these sweeping statements? Let us remember the criteria proposed by the authors: regime survival without destabilization, and no important antisystem actors. If these are the criteria, the authors' universe of sufficiently consolidated democracies would have to be broadened to include, among the new Latin American democracies, at least Argentina, Bolivia, and Brazil. These countries survived, without becoming politically destabilized or generating important antisystem actors, far more severe social and economic crises than Southern Europe has. The authors' error regarding some postcommunist countries is even more blatant. The Czech Republic, Hungary, Poland, and Slovenia have seen no significant antisystem party or social movement, have survived " simultaneous transition to democracy and to a market economy" (p. 390, italics in original; see also p. 396), including severe social and economic crises, and have even moved toward increasingly stable party and electoral systems. It is surprising that the authors do not apply their own criteria and conclude that these countries have worked the miracle of near-instantaneous "consolidation."

I must conclude that the authors do not use their own criteria consistently, or that these "reliable empirical indicators" (rejoinder) lack the analytical and empirical edge they claim.

Teleology

By a teleological concept I mean one which posits, explicitly or implicitly, that a given entity inherently tends to move from lower (or immature or incomplete) to higher (or more mature, or complete) stages, [End Page 163] up to an end point that marks the full development of its potentialities. Characteristically, the stages are understood from their end point: a seed is "basically" a potential tree, or a given democracy is, as the authors put it, "not yet" sufficiently consolidated. Entities are defined negatively, characterized not by their specific attributes but by what they lack in relation to the paradigmatic end point of their presumed trajectory. Negative definitions generate residual categories: cases are classified together on the basis of their sharing the lack of attributes that the more developed specimens of the same genus supposedly have. Furthermore, such a view usually entails conceiving of progress as movement through various "stages" along the posited "trajectory" toward "successful completion," that is, "`full consolidation' at the extreme end of the continuum" ⁸ (rejoinder). Consequently, other kinds of changes are deemed deviations from the proper trajectory, and factors that apparently hinder upward changes are seen merely as "obstacles" to such changes.

Thus democracies are classified not as positively defined types X or Y, but as (sufficiently) consolidated, or unconsolidated, or partially consolidated, or substantially but not sufficiently consolidated, and other references to what they lack in relation to the ideal type. The location of each case in the presumed "trajectory"--its "stage"--determines how it is conceptualized. Of course, if such a trajectory does not exist, or cannot be mapped with even minimal precision, analyses based on this view will be theoretically flawed and most probably misleading. ⁹

We have seen that teleological thinking abounds in the authors' book. Furthermore, they properly, if perhaps unwittingly, emphasize their use of "the concept of trajectory [which] is meant to capture and highlight the particular combination and interplay of freedom and constraint at each successive stage of the democratization process" (p. xvi), adding that they "regard continued movement towards the ideal type of democratic consolidation . . . as very significant" (p. 9), a movement that leads to the "successful completion" of (sufficient) consolidation (pp. 389, 405). The rejoinder strongly reiterates these views.

I argued that teleological views entail negative definitions of lower-end cases that wind up in analytically useless residual categories; this book is an extreme instance of this flaw. The authors describe "unconsolidated" democracies as cases where "important and powerful elites and their supporters deny the legitimacy of the existing regime and may seek to overthrow it," where "few political actors are prepared to stake their futures on the workings of democratic institutions . . . . [in part] because they also perceive rival political parties as conditional in their support for democracy," and where "mass mobilizations in the streets take the place of bargaining among representative elites," as a consequence of which a cycle "may be set in motion that progressively polarizes relations among groups and raises the overall level of violence within the polity" (p 10). But here the authors describe situations that [End Page 164] simply do not fit a definition of democracy--neither the authors' nor the one I proposed in my essay. The cases that meet the authors' somber description are not consolidated democracies because, to begin with, they are not democracies. They correspond to Samuel P. Huntington's "mass praetorian" regimes, or they may be liberalized authoritarianisms or electoralist regimes--but they are not democracies. ¹⁰ Even if they have reasonably competitive elections and a fair count of the votes, they do not qualify as democracies, or polyarchies, because they lack other definitional attributes of this type of regime.

For a work that claims to have comparative import, the authors' book features a remarkable number of unwarranted generalizations about many of the new (and, in fact, also the old) democracies outside the Northwestern quarter of the globe. Many of these countries have serious problems derived from particularism and inequality (both of which are far from alien to Southern Europe). Yet even though several of these new democracies--including at least the Czech Republic, Hungary, Poland, Slovenia, Argentina, Bolivia, Brazil, Ecuador, Panama, Benin, and Malawi--show no signs of mass praetorianism or sheer electoralism, the authors lump them with nondemocratic countries in the residual, negatively defined category of "unconsolidated democracies." This is not a good way to forge the theoretical and comparative tools that the authors proclaim as one their goals.

Southern Europe and Latin America

In their rejoinder, the authors argue a truism. Obviously, before deciding if a democracy is consolidated, or if it belongs to type X, Y, or Z, one must define what one means by democracy. Against the opinion of the authors, we happen to agree on this matter. They use Juan Linz's definition, which, as they note, is similar to Robert Dahl's definition of polyarchy, which I adopted in my essay for reasons and with some additions explained there. ¹¹

Having cleared up this point, I now turn to the rejoinder's arguments about "Latin America," which actually refer only to Chile and Brazil. The authors see Chile as a case of "incomplete democratization"; they seem to use this argument to refute me, but, as should be clear from my essay, I agree with them. They further mention "the tremendous differences in levels of public support for democratic regimes" in Southern Europe versus Latin America. ¹² But public opinion data from several Latin American countries do not show this. The rejoinder cites only Brazilian surveys in which a proportion of respondents that "never exceeded 48 percent" agreed that "democracy is preferable to any other form of government." This they contrast with 78 and 90 percent agreement figures from Spain and Greece, respectively. Yet they fail to mention that 80 percent in Uruguay, 76.6 in Argentina, and 52.2 in [End Page 165] Chile agreed with this same statement. ¹³ Again, the authors do not give even one clear and consistent indicator of (sufficient) consolidation, to say nothing of its various "stages."

The rejoinder also contains a misinformed allusion to "mainstream Brazilian political leaders [who] have made flagrantly semi-loyal statements." The only such leader turns out to be Leonel Brizola, who at the time had already fared dismally in elections, lost control of his party, and become a completely marginal figure. Furthermore, Brazil underwent a series of very "severe tests" ¹⁴ without politically destabilizing or generating any important antisystem party or social movement. According to the authors' own criteria, Brazil should be considered at least "substantially" consolidated.

On the subject of clientelism and, more generally, particularism, as important informal institutions in many new polyarchies, the rejoinder makes two points that I not only readily admit but elaborated in my essay. These are that all democracies "include to varying degrees informal structures of the kind O'Donnell discusses," and that widespread particularistic practices "are antithetical to the quality of democracy." But the authors go too far in claiming that these practices are "incompatible with the unhindered exercise of suffrage." This is simply empirically untrue. Furthermore, if the authors are right in drawing this conclusion because particularism "involves and perpetuates unequal treatment of individuals or groups" and clientelism "entails systematic and persistent power imbalances within society, polity, and economy," then they are implicitly making a radical critique of any democracy: Can they ignore the pervasive consequences of class, status, bureaucratic power, and so on that everywhere also perpetuate unequal treatment and generate systematic and persistent imbalances?

The authors also argue that "over the long term, resentment over these exploitative relationships . . . can help to undermine the stability of the regime." Yet in their own account, the "undermining of stability" may lead not to breakdown but to "reequilibration" at a higher point in the "consolidation trajectory." As for the relative likelihood of breakdown in informally institutionalized versus formally institutionalized polyarchies, in my essay I declared myself agnostic pending better research. At least, I noted there that particularistic relationships are very important and pervasive in such long-lasting polyarchies as India and Japan and, indeed, Italy. With the exception of their already-quoted reference to Spain, the authors downplay in their book and ignore in their rejoinder a large literature showing the major role various kinds of particularistic phenomena play in the actual workings of the Southern European polyarchies. ¹⁵ I do not know how far these phenomena approximate those cases to the Latin American ones on which my essay focused. We will never know the answer, however, if we join the authors in dismissing these phenomena as pathologies that afflict cases [End Page 166] that have not reached the fortunate status of "sufficient consolidation," and which they do not deem worth noting as "consolidated" cases.

Guillermo O'Donnell is Hellen Kellogg Professor of Government and International Studies at the University of Notre Dame. His essay "Illusions About Consolidation" appeared in the April 1996 issue of the Journal of Democracy.

Notes

1. Richard Gunther, P. Nikiforos Diamandouros, and Hans-Jürgen Puhle,eds., The Politics of Democratic Consolidation: Southern Europe in Comparative Perspective (Baltimore: Johns Hopkins University Press, 1995), 3. Except when otherwise noted, all page references in parentheses are to this volume.

2. Assuming that a settlement can be reached regarding the eventually thorny issue of who are, or should be, the "politically significant actors."

3. The authors are inconsistent in the use of their terminology. Sincethey regard "full consolidation" as "the unachievable end point of our ideal-type continuum" (rejoinder), they deal with "sufficient consolidation" as the higher point in the consolidation "trajectory." On many occasions, however, they merely refer to "consolidation" or to a case as "consolidated," when in the context it seems that they should have added the qualifier "sufficient." In these cases, I add this term in parentheses or brackets.

4. See Philippe Schmitter, "Organized Interests and Democratic Consolidation in Southern Europe," in Gunther, Diamandouros, and Puhle, eds., Politics of Democratic Consolidation, 284-315.

5. Since "full" consolidation is an unattainable "ideal type," I gather that the text actually refers to "sufficient" consolidation.

6. The possibility of such a national-regional consolidation split, andits practical and theoretical consequences, are not elaborated.

7. The authors assert that "democracy in Italy was by and large consolidated" when "by the end of the 1970s, the PCI [Italian Communist Party] had demonstrated its reliability as a loyal democratic competitor" (p. 25). Until then, "concern over the PCI's commitment to democracy and its perceived antisystem stance constituted the major obstacles to consolidation" (p. 391). One could argue no less plausibly that since the Svolta de Salerno in 1944, the PCI had stubbornly demonstrated democratic loyalty (see Gianfranco Pasquino, "The Demise of the Fascist Regime and Italy's Transition to Democracy: 1943-1948," in Guillermo O'Donnell, Philippe Schmitter, and Laurence Whitehead, eds., Transitions from Authoritarian Rule: Southern Europe [Baltimore: John Hopkins University Press, 1986], 45-70), and that it was in the cynical interest of other political actors to raise the "specter of communism" to hinder the electoral chances of the former. The position of the authors in this matter seems influenced by another criterion of sufficient consolidation that they introduce entirely ad hoc. In their discussion of Portugal they refer to provisions of the 1976 Constitution, including those "guaranteeing the irreversibility of the nationalizations carried out during the revolutionary period," as the "last obstacle" to (sufficient) consolidation "removed through the May 1989 constitutional reform" (p. 28; see also p. 390). The rationale for this argument is that such clauses removed "important substantive policy issues from the everyday give and take of democratic politics" (p. 28). Here the authors give us a narrow and ahistorical implicit definition of democracy, according to which only issues of private property can be removed from political "give and take," and which has no place for a constitution that gives the state an important role in the economy. Their arguments about Greece are also indicative. In arguing "for using 1977 as the date signaling the end [sic] of consolidation," the authors point out as decisive the Greek Socialists' abandonment of their "erstwhile Third World orientations towards Greek foreign policy" (p. 30). Apparently democracies cannot exist, much less consolidate, if all politically significant actors do not share "the requisite conditions of moderation and restraint" (p. 391; see also p. 402) that the authors prefer--in this view India's and Israel's democracies cannot possibly have existed. These criteria are too ad hoc and subject to too many comparative variations across time and space to merit further discussion.

8. Notice, however, that according to the authors "the extreme end" is the"stage of democratic persistence"; the criteria and even the end points of this "trajectory" seem to be in permanent oscillation.

9. Against the authors' disclaimers in both their book and rejoinder, ateleological view does not imply that all cases reach maturity, or that they cannot "freeze" at an intermediate stage, or that they cannot degenerate after reaching maturity. The point is that these events are seen as deviations from the proper, somehow preexistent path (p. 398). Moreover, a teleological view does not preclude asserting that a given entity has somehow "leap-frogged" some developmental stages (pp. xiv-xx), or that different objects may travel "quite different routes" (p. 19) toward the end point, or that progress toward the end point should be "unilinear" (p. 20).

10. On "mass praetorian" regimes, see Samuel P. Huntington,Political Order in Changing Societies (New Haven: Yale University Press, 1968). On "electoralist" regimes, see Terry L. Karl, "Imposing Consent? Electoralism vs. Democratization in El Salvador," in Paul Drake and Eduardo Silva, eds., Elections and Democratization in Latin America, 1980-85 (San Diego: Center for Iberian and Latin American Studies, 1986), 9-36.

11. Because of space limitations, I refer the reader to this definition in my essay. Whatever its merits, this definition should suffice to show that only an extremely casual reading of my essay could have led the authors to assert that I regard "the holding of elections as the only institution relevant to the study of democratization," or to attribute to me a term ("informal polyarchies") that I never wrote and that bears no relation to the "informally institutionalized polyarchies" discussed in my essay.

12. Actually, in their book the authors are rather dismissive of thesignificance of this factor, and probably would have vigorously denied it if I had seen them using it as an indicator of consolidation. Furthermore, in their book they note that "pervasive cynicism at the mass level and extraordinarily low levels of mass-level involvement with politics reflect the incomplete status of efforts to resocialize the Spanish public" (p. 22). The reference is to Spain, but similar phenomena have been found in the other Southern European cases; see especially José María Maravall, Los resultados de la democracia: Un estudio del sur y el este de Europa (Madrid: Alianza Editorial, 1995), 257-302.

13. Data from Latino Barometer, 1995, transcribed by Juan Linz and Alfred Stepan, Problems of Democratic Transition and Consolidation: Southern Europe, South America, and Postcommunist Europe (Baltimore: Johns Hopkins University Press, 1996); I thank these authors for facilitating my access to these data. Should we conclude from these data that Uruguay and Argentina are "as consolidated" as Spain while Greece is even more "consolidated" than these three countries, or that Chile, the only newly democratized Latin American country that the authors see as "well on its way" to consolidation, may not be so? Obviously, this does not seem a promising kind of discussion.

14. In addition to several years of four-digit inflation, one should count at least: the premature death of a very popular elected president (Tancredo Neves), five years of an extremely inept presidency (José Sarney), the impeachment for corruption of another president (Fernando Collor de Mello), and the indictment on similar charges of several prominent legislators--all without rumor or fear of military intervention. It should be noted, too, that Argentina not only weathered even higher inflation rates than Brazil but, when faced with a military-coup threat similar to Spain's, saw all political and social leaders (also as in Spain) rally unequivocally to democracy, with mass demonstrations supporting their stand. It seems that to the authors, surviving "severe tests" indicates "substantial" or "sufficient" consolidation in Southern Europe, but only "unconsolidation" in the rest of the world.

15. For a recent assessment and bibliographical references on these andrelated matters, see Maurizio Ferrera, "Il modello sud-europeo di welfare state," Rivista italiana di scienza politica 26 (April 1996): 21-66.

[an error occurred while processing this directive].