Explaining the Chile–Uruguay Divergence in Democratic Inclusion: Left Parties and the Political Articulation Hypothesis

Abstract This article develops a political articulation-based explanation of divergence in democratic inclusion between two champions of liberal democracy in Latin America: Chile and Uruguay. Political articulation scholars depart from the traditional reflection models of political parties as mere expression of preexisting social cleavages, highlighting the relative autonomy of parties’ practices and their strategic role in structuring state-society relations. My work extends this current trend in comparative-historical sociology to the Latin American Left turn after the demise of the market-fundamentalist Washington Consensus, empirically identifying a set of strategies that boost Left parties’ capacity to articulate on a specifically class basis. These strategies, I argue, are endowed with causal efficacy, driving democratic variation beyond the restrictions and opportunities of the institutional environment. Combining process tracing account of historical sequences with my own analyses of labor statistics, protest events, and party linkages and manifestos, I show that differences in Left parties’ ability to build linkages with labor and advance its institutional representation as a class actor are at the root of the divergence in political inclusion between these two countries. This finding has substantial implications for contemporary democratic theory: after neoliberalism, strongly organized mass parties of the Left may not be a necessary condition for a given democracy’s stability and consolidation, but they may be a sufficient condition for a particular democracy’s realization of the normative ideal of political equality.


Introduction
Equally strong capitalist democracies may differ widely in the political inclusion of the powerless. Take the cases of Chile and Uruguay, two Latin American countries with long democratic histories. Mass protest movements against the pervasive commodification of social rights have recurrently erupted in Chile for more than a decade. The violent "social outburst" (estallido social) of 2019 has only confirmed the so-called awakening of civil society against what ordinary people view as an undifferentiated ruling elite, including the established center-Left Concertación coalition that governed during most of the postdemocratization period (Roberts 2016;Somma 2017). Scholars concur that the market-oriented policies instituted during Pinochet's dictatorship are not the ultimate cause of popular unrest. Beyond economic demands, the legitimation crisis of Chilean democracy speaks to citizens' deep-seated experiences of political inequality in the context of socially uprooted representative institutions (Luna and Altman 2011;Somma and Medel 2017)an explosive situation that might even lead to the collapse of the party system (Somma et al. 2020).
Meanwhile, Uruguay, another regional champion of democratic stability that experienced dictatorship and neoliberalism at roughly the same time is typically depicted as a success story of welfare state resilience and expansion. Unlike Chileans, Uruguayan citizens massively participate in elections and express significantly higher regard for democratic institutions, including political parties. Labor, feminist, student, and other social movements frequently organize strikes and mass demonstrations. But their actions take place peacefully, as part of the routine conversations of a democratic polity. Importantly, accounts of the inclusiveness of Uruguayan democracy stress the central role the Frente Amplio (FA)-a mass center-Left, labor-based party reminiscent of European social democracy-has come to play as the largest party in the country, allowing relatively powerless actors from civil society to raise their voices and demands within institutionalized political processes (Bidegain and Tricot 2017;Lanzaro 2011;Lissidini 2016;Padrón and Wachendorfer 2017).
Through the paired comparison of Chile and Uruguay, this article examines Left parties' contribution to forging more or less inclusive democracies, arguing that, notwithstanding notable historical similarities, the Uruguayan Left has gone further than its Chilean counterpart in realizing the ideals of political equality on which democracy rests (Fishman 2011;Markoff 1997). Latin American scholars have provided at best incomplete explanations for why the Left has performed so differently in amplifying ordinary citizens' voices in these apparently similar capitalist democracies. Early typologies of the so-called Latin American pink tide simply lumped the Chilean and Uruguayan center-Left governments together as the same subtype of postneoliberal turn. As evidence of divergence accumulated, scholars began highlighting differing Left parties' linkages with civil society (Handlin and Collier 2011;Luna 2014;Luna and Filgueira 2009;Pribble 2013). Yet, these accounts tended to view parties' character as path-dependent on the institutional environments crystalized during transitions from authoritarian rule, thereby understating the historical agency of party practices and, in particular, bottom-up processes of party formation and mobilization. Critically, the literature has paid scant attention to parties' role in strengthening the societal power and autonomy of organized labor as a representative class actor within capitalist democracies-allegedly, the historical mission of social democratic parties (Esping-Andersen 2017; Korpi 2018).
My central claim is that the root of divergence in democratic inclusion between these two otherwise comparable polities is whether, after redemocratization, Uruguay's and Chile's institutionalized Left parties developed as mass-based organizations that strengthened the power of the labor movement and its allies, thus contributing to the making of more or less robust democratic practices. I draw from Fishman's (2011: 236) concept of democratic practice to highlight "the ways in which political actors-including ordinary citizens, groups that are organized or spontaneous, and institutional office holders-make use of the rights and possibilities for action provided by democracy and deal with others who are similarly engaged." More or less inclusive democratic practices "have the capacity to substantially enhance [or inhibit] the goal of full political equality among citizens" (Fishman 2011: 236). Reconsidering the Latin American Left turns and their aftermaths through a new analytical lens, I show that Left parties' practices constitute a critical determinant of democratic development, the presence (or absence) of which sets in motion analytically distinct patterns of popular interest intermediation and political inclusion.
By evincing the causal efficacy of party-led democratic inclusion, this article likewise contributes to the new comparative-historical sociology of political parties. A growing stream of scholarship is now looking at variation in what parties are and do from a deeply historical and international perspective, taking seriously the methodological premise that "parties are not stable or developmentally unidirectional" (Mudge and Chen 2014: 323). Building on this premise, I offer a political articulation-based explanation of the Chile-Uruguay divergence. The political articulation school posits that parties' autonomous practices matter, in the sense that the intermediation capacities developed by strongly organized, tightly coupled, externally mobilized political parties 1 critically determine how both capitalist democracies and civil societies develop (de Leon et al. 2009(de Leon et al. , 2015Desai 2002;Eidlin 2016Eidlin , 2018Mudge 2018;Riley 2010;Riley and Fernández 2014;Ziblatt 2017). This emphasis on parties' historicity makes the articulation framework distinct from reflection models that conceive of party politics as a mere expression of preexisting social cleavages and identities. Still, beyond this broad theoretical proposition, this article empirically identifies a set of strategies that boost the Left's capacity to articulate on a specifically class basis: safeguarding the organizational autonomy of its grassroots, mobilizing them between elections, boosting the bargaining power of the labor movement, elevating its institutional status as a unified class representative above special interests, and cultivating relationships with experts recruited from socially progressive circles. Thus, rather than simply confirming the importance of party-driven articulation in general, I show that the class-based articulation strategies of Left parties yield a specific outcome: inclusive democracy.
This inquiry into how the articulation work of Uruguayan and Chilean Lefts gave shape to markedly different democratic regimes also speaks to a classic problem in historical (Left) party research. Left parties' role in broadening the institutional representation of popular sectors in capitalist democracies is well established in the historical literature (see e.g., Eley 2002). It is also a salient theme in Mudge's (2018) multicase study of Left reinventions in Europe and the United States throughout the twentieth century. Classic as the issue may be, there is much to be learned about the relationship between the Left and democratic inclusion (or lack 1 According to Ziblatt (2017: 43), this type of party organization combines "professionalism at the top, local mobilizing structures at the bottom, and control by the center," all characteristics that endow mass parties with the organizational skills to "manage the introduction of new lines of conflict into battles over political power" (ibid.). thereof) in the current historical moment and in places beyond the Global North, where the technocratic, promarket drift of Social Democracy has deprived working class and other subaltern constituencies of key vehicles of political expression. Conventional wisdom states that the erosion of Left parties' capacities to speak to (and for) their traditional constituencies has fueled the electoral growth of right-wing populism and therefore represents a direct threat to the legitimacy of Western democracy. But how universal is this process? What can we learn from Latin America?
Cutting-edge comparative work on some governing parties of the Latin American pink tide reveals that certain organizational characteristics function as effective buffers against oligarchic tendencies. Specifically, this research shows that when parties develop mechanisms that safeguard the autonomy of their core constituencies and hold party leaders accountable to grassroot members, they are better placed to promote the inclusion of nonelite groups not only as voters but also, crucially, as policy makers between election cycles (Anria 2018;Bentancur et al. 2019;Rosenblatt 2018). Openness to vertical accountability, however, features only in some, but not all, party organizations. Seldom has it been present in parties of the populist type, which build political power through a combination of charismatic leadership, particularistic forms of clientelism, and social movement cooptation (e.g., Peronism in Argentina or the United Socialist Party of Venezuela under Chavez). Likewise, the inclusion of party supporters and allied organizations in policy formation is no longer found in Left parties that, albeit originated in organic processes of grassroots participation and mobilization, have eventually institutionalized rules that shield the party elite from the political pressure of their mass constituencies (e.g., the Brazilian Workers Party). Informal avenues for vertical accountability to grassroots and autonomous social movements is nonetheless the norm in the Movement Toward Socialism (MAS) in Bolivia (Anria 2018). But the MAS features as a rather unique movement party (Della Porta et al. 2017) that articulates indigenous and class identities (peasants and urban informal workers), not as the classic labor-based party of the social democratic Left. This article focuses on the latter type, comparing the developmental trajectories of two institutionalized parties of the Latin American Left that come closest to European Social Democracy.
Thus, in explicating the relationship between Left party's organizational development and democratic practice, I extend the party articulation framework to a new family of outcomes: varieties of Left turns and political inclusion in postneoliberal Latin America. In addition to widening the explanandum of the articulation approach, I also seek to strengthen its explanans. To situate Left parties' articulation strategies within a plausible explanation of divergence in the inclusiveness of democratic practices, I document the institutional embeddedness of grassroots involvement and popular interest aggregation. Borrowing from the formal logic of set theoretic methods, I contend that the class-based articulation strategies observed in this pair comparison are causally efficacious when Left parties manage to overcome certain institutional constraints and skillfully exploit the political opportunities of democratizing scenarios. I therefore examine the institutional conditions that enable or inhibit the development of Left parties as mass-based, labor-mobilizing organizations, integrating structure and agency into a single political articulation account.
I proceed as follows. In the next section, I revisit the Latin American Left turns to expose the Uruguay-Chile divergence as an anomaly for which this literature cannot account. I then discuss and give credit to the political articulation hypothesis vis-à-vis rival accounts that emphasize the institutional legacies of authoritarian regimes. After a brief methodological note, I provide a summary narrative of what Chilean and Uruguayan Left parties have done (and not done) to organize antineoliberal struggles and effectively include the voices of relatively powerless actors. I close with a final discussion of the role of Left mass parties in advancing more inclusive democracies.
The Chile-Uruguay Anomaly: Divergence, Not Convergence As the twenty-first century began, market-oriented, export-led development lost credibility throughout Latin America, precipitating legitimation crises in newly democratic regimes. In a few countries, social unrest seriously destabilized the political system (Silva 2009). In many, left-wing/neopopulist parties won access to power with the mandate of shifting the course of top-down conservative modernization to reincorporate popular sectors in the political arena (Luna and Filgueira 2009;Roberts 2016). The so-called pink tide signaled the beginning of the post-Washington consensus era, or postneoliberalism.
Beyond commonalities, scholars concur that the wave of Left turns was heterogeneous in intensity and character. Inevitably, this diversity of Left governments was "rooted in distinct historical experiences and pathways to political power" (Levitsky and Roberts 2011: 4). The first regime typologies (see appendix for illustrations) distinguished neopopulist from moderate experiments led by institutionalized, social-democratic Left parties, based on the degree to which governments pursued a regression to state dirigisme (Castañeda 2006). Others looked at variation based on whether Left governments relied on established party organizations or newly created political movements, and whether authority was concentrated in charismatic leaders (Levitsky and Roberts 2011). Still others focused on how excluded popular sectors regained recognition and became reincorporated in political institutions (Silva 2017;Silva and Rossi 2018). Whether political parties served as main vehicles for reincorporation features as a key differentiating factor. Roberts's (2015) multicase study, for example, posits that institutionalized Left parties' capacity to effectively channel societal dissent into the political arena explains why some Latin American democracies successfully coped with the social tensions caused by neoliberal restructuring, while others underwent severe crises that led to the collapse of their party systems. Where the Left politicized market-driven social inequalities and generated a scenario of "contested liberalism," party systems survived.
The problem with Roberts's and other multicase typologies of postneoliberal regimes is that they generally lump Chile and Uruguay together (see appendix). In doing so, they fail to account for important differences within the subtype of Left turn to which these two countries are said to belong. Tacitly assumed is that because in both countries electorally competitive, programmatically moderate Left parties were available, they proved equally efficacious in articulating popular sectors' reincorporation demands, effectively serving as institutional mediators and, hence, preventing major legitimation crises.
The historical evidence does not support this claim. Left parties' mobilization of civil society to resist neoliberalism simply did not take place in postdictatorship Chile, which may be better characterized as a case of nonpolitical, market-based reincorporation of popular sectors within a "low-intensity democracy" (Luna and Filgueira 2009: 385). Through technocratic forms of policy making, the Concertación governments  were able to incrementally expand the universal coverage, transparency, and equity of pensions, healthcare, and family allowances. Doing so, however, came at the cost of maintaining the quasimarket schemes of social service delivery instituted by the Pinochet's dictatorship. This strategy further alienated the Left's traditional constituencies, namely, labor and social movements. Persisting market insecurities caused by widespread commodification of social rights in the context of a socially uprooted political system fueled popular disaffection with political institutions, including established Left parties (Luna and Altman 2011;Somma and Medel 2017).
In Uruguay, the FA reincorporated into democracy as an independent third party, strategically oriented toward mobilizing civil society, especially labor and its allies. Because of the effective opposition of a mobilized Left, postdictatorship Uruguay preserved some of the core institutions of its welfare state during more than a decade of right-wing government, and then advanced farther toward decommodified universalism when the Left won the presidency (2005-20) (Castiglioni 2003;Pribble 2013).
Hence in-depth analyses of one or both cases exhibit remarkable cross-country divergence in the linkage strategies of Chilean and Uruguayan Left parties (Handlin and Collier 2011;Luna 2014;Luna and Filgueira 2009;Pribble 2013). Likewise, the data based on country experts' reports (Kitschelt 2013) show that the Uruguayan FA's linkage strategy focuses much more on making programatic appeals than the Chilean Concertación, offering universalistic public policies to its traditional constituency in exchange for electoral suport (see also Luna 2014). Uruguay's center-Left party also ranks significantly higher than its Chilean counterpart in the index of programmatic structuration of economic issues, that is, the extent to which parties advocate income redistribution, state dirigisme, and public provision of pensions, healthcare, and education. Kitschelt's data further show that the Concertación regularly engages in patronage, rewarding voters through particularistic access to consumer goods, public benefits, employment opportunities, government contracts, and regulatory proceedings. According to the same survey, during the 2000s the overall recourse to patronage receded in Uruguay, while it increased in Chile (see also Álvarez-Rivadulla 2012;Luna 2014). In short, although Left parties in Uruguay and Chile are both rooted in a similar tradition and share a common history of political mobilization, repression, and reincorporation into democracy, they have evolved into very different kinds of organizations, adopting markedly divergent strategic orientations.
Institutional Constraints and Opportunities: A Rival Hypothesis? Significant differences in parties' character notwithstanding, an important question remains: To what extent can these differences be fully attributed to the institutional legacies of transitions from authoritarian rule? Put differently: How much leeway did political actors within the Left camp have to shape the course of democratization? Both Uruguay and Chile recovered their historically strong democratic regimes at roughly the same time through elite negotiations between opposition political parties, on the one hand, and the civilian and military arms of the dictatorial regime, on the other. But the pacts that enabled democratization were contingent upon very different institutional frameworks with very different legacies. Political institutions regulate conflict and allocate incentives and veto powers across different kinds of political actors (Mahoney 2010), providing (or limiting) the opportunities necessary for nonelite actors and peripheral polity members to organize collectively and intervene effectively in the democratic process. A plausible hypothesis is that center Left parties in each country simply adapted their organizational and strategic orientations to the rules of the game crystalized during the negotiated transitions.
Of those rules, the constitutional ones were all the more important (Romero 2009). In 1980, both regimes pursued the electoral ratification of new constitutions that institutionalized military participation in government, created countermajoritarian institutions, and granted extraordinary powers to the executive. However, while Pinochet succeeded after committing massive fraud, the Uruguayan generals refrained from manipulating the electoral process, only to find their constitutional bill unexpectedly defeated in the ballot. Hence the institutional blueprints that served the context for transition bargaining differed substantially. Whereas elite pacts in Chile ratified the constitutional transformation enacted by the military, the Uruguayan pacts restored the constitutional provisions in force before the military coup (Demasi 2009;González 1991;Valenzuela 1992).
Because in Chile the military constitutionalized the regime before formal negotiations for democratic opening began, the resulting pacts were very restrictive. First, democratization pacts severely reduced the range of inclusiveness of the bargaining cartel, splitting moderates from radicals, and creating incentives for defection from opposition solidarity. The rapid moderation of the Chilean Socialist Party and the total exclusion of the Chilean Communist Party from the Concertación coalition are a case in point. In Uruguay, in contrast, the fact that the FA was able to reenter the democratic game under relatively similar institutional conditions as those preceding the coup discouraged programmatic moderation and defection, keeping the Left united. Second, the Chilean pacts severely constrained the subsequent democratic game. Provisions such as super majorities in Congress, a Senate with nonelected seats, a majoritarian electoral system that virtually excludes third parties, and a politicized constitutional tribunal with preemptive powers over the legislature granted veto powers to regime incumbents. The Uruguayan Constitution, in contrast, allocated veto powers to opposition actors and nonpolity members such as unions or social movements through the provision of mechanisms of direct democracy. Proportional representation also favored the representation of the FA as an uncompromising third party. Finally, the pacts in Chile ratified highly restrictive procedures for constitutional change. Constitutional rigidity operated as an anchor of predictability that stabilized the polity. But it also slowed subsequent movements toward further democratization. Again, in Uruguay, the balance of stability and transformative potential was markedly different. 2 Transition pacts proved consequential. In Chile, the Right's overrepresentation in political institutions favored intraelite consensus building rather than conflict. The authoritarian enclaves also limited the effective representation of dissent and inhibited the formation of defiant political actors, seriously constraining the extraparliamentary actions of Left parties. Meanwhile, although important authoritarian legacies were also present in Uruguay's postdemocratization aftermath, the political game appeared more open to contestation. Eventually, the institutional environment provided the Uruguayan Left with unique opportunities to mobilize against neoliberalism. In particular, as I explain later, organized groups had the possibility to trigger binding mechanisms of direct democracy through popular initiative. These extraparliamentary mechanisms constituted a major incentive for the FA to challenge governing parties by reaching out to civil society.
Institutional constrains and opportunities certainly influenced Left parties' trajectories in the democratization aftermath. But opportunities need to be seized. Constraints, likewise, may be challenged and eventually lifted. Whether more or less open, democratization scenarios are always "moving targets" of political actors' struggles (Markoff 2011). An historical-institutionalist explanation that gives too much weight to parties' path-dependent trajectories risks obliterating their historical agency. Moreover, it overlooks that parties are not monolithic entities but contested fields formed in ceaseless struggles of coalition formation and mobilization. Intervening in these struggles are also competing factions of truth-claiming "experts" vying for the definition of political language and programs that shape parties' capacity to speak to (and for) different types of constituencies (Mudge 2018).
To be clear, I do not argue that the institutional environment is irrelevant for parties' organizational development and orientation. But I do hold that the constraints and opportunities parties face operate as enabling conditions for other political processes and mechanisms at work. Following the formal principles of conjunctural causation, I argue that, as a whole, Left-party articulation works as an INUS condition: an insufficient but necessary part of an unnecessary but sufficient causal sequence of democratic inclusion (Beach and Pedersen 2013;Schneider and Wagemann 2012). Parties' practices of political articulation are causally efficacious because, in their absence, we do not observe the outcome of interest. But these practices are just an insufficient but necessary part insofar as, by their sole action, they do not suffice to produce, for instance, a more inclusive democracy. Only when combined with specific institutional settings do these practices lead to an actual outcome, which would otherwise remain only possible. And yet, because in a hypothetical analysis of more than two cases other causal sequences might lead to the same outcome (which is known as the principle of equifinality), this theorized combination of institutional factors and political agency is just unnecessary but sufficient. In short, although my conceptual framework underscores agency, I do not claim that parties are endowed with unlimited capacity to shape cleavages and identities (Eidlin 2016), let alone that the particular combination of structure and agency observed (or not) in my paired comparison is the only pathway to more or less inclusive democracies.
A shadow case borrowed from Fishman's (2019) analogous comparison between Spain and Portugal illustrates this theoretical proposition. For Fishman, the more inclusive democratic practices observed in Portugal vis-a-vis Spain relate to the social-revolutionary origins of the Portuguese transition to democracy, which engendered antihierarchical repertoires of popular engagement and consultation that with time became internalized by chief political actors across the board. Because these practices are rooted in the progressive cultural legacies of the Carnation Revolution, not in the articulation strategies of Left parties, the Portuguese road to democratic inclusion can be seen as an alternative causal sequence leading to the similar outcome observed for Uruguay. Figure 1 formalizes this proposition using Boolean logic. Chile's noninclusive democracy (∼Y) is the consequence of a transition through elite pacts (X) that bequeathed an authoritarian constitution (C), with no Left-party articulation (∼A) of popular sectors in the democratization aftermath. A transition from authoritarian rule via elite pacts (X) is a condition Uruguay shares with Chile, yet in the former a relatively open institutional framework enabled by the absence of an authoritarian constitution (∼C) and an articulating Left (A) led to an inclusive democracy (Y). Outcome Y is also observed in Portugal, but in this case democratic inclusion results from a transition via social revolution (Z), combined with the antihierarchical legacies that institutionalized over time (L). If Uruguay and Portugal arrive at Y through pathways X & ∼C & A and Z & L, respectively, it follows that these pathways are unnecessary but sufficient causal sequences leading to the equifinal outcome Y. It also follows that each condition in the Uruguayan pathway (X, ∼C, and A) is a necessary but insufficient part of the sequence, and therefore political articulation (A) is an INUS condition.

The Political Articulation Hypothesis
In 2007, two thousand nonunionized contract workers (contratistas) staged an illegal strike to demand CODELCO-Chile's state-owned copper corporation-provide them equal pay and workplace conditions to the company's permanent staff. The immediate trigger of the strike was CODELCO's reluctance to honor an agreement signed after a prior round of mobilization in 2006. At the time, the contratistas' incipient movement had received the endorsement of socialist Michelle Bachelet, then running for president as Concertación's candidate. Right after the election, Concertación had also passed the Subcontracting Act-sitting dormant in Congress since 2002-which required a holding company to employ all subsidiaries' workers under its direct supervision. Hence, in their list of demands, the strikers included the enforcement of this act. As detailed by Donoso (2017), the fact that the strike affected Chile's commodity-exports flagship sector and historical bastion of the labor movement attracted significant media coverage. Yet the unfolding of events evinced that the contratistas were very much alone. The strikers denounced the traditional central union's leadership for having acquiesced to Concertación's calls for demobilization. They also criticized Concertación's labor reforms, which had furthered labor market deregulation and harmed union organizing. The strikers even antagonized the union representing CODELCO's permanent workers, who feared the conflict could jeopardize their privileged status within the holding company.
The contratistas' strike lasted 37 days. Only after three weeks of violent protests and huge economic losses did the government intervene, requesting the mediation of the Catholic Church. In the end, the strikers gained important concessions, and President Bachelet created an Advisory Commission to consider a thorough reform of the labor code. Still, the contratistas declined to participate in the debate to not legitimize the commission's presumable business-friendly recommendations. Meanwhile, as the Commission deliberated, the National Labor Bureau (NLB) began enforcing the Subcontracting Act, ordering CODELCO to employ almost five thousand contractors, provoking a crisis within government ranks. The economic arm of the executive, in control of CODELCO's board, sought the legal protection of the courts, overtly challenging the NLB's authority. The legal dispute was finally won by the company after the Supreme Court revoked the essential provisions of the Subcontracting Act. Infuriated, the contratistas initiated a hunger strike, shortly ended after the president made sure that CODELCO thoroughly implemented the agreement signed with the workers. Victory notwithstanding, the contratistas' collective sacrifice helped improve only the individual conditions of the movement's participants.
The workplace safety campaign led by construction workers in Uruguay in 2013 stands in sharp contrast with the dynamic observed in Chile. In April of that year, SUNCA, the national federation representing all unionized construction workers, staged a day strike and mobilized more than 20,000 workers to Parliament in support of the so-called Employer Criminal Liability Act, which held employers liable for accidents resulting from violations of workplace safety standards. The bill was jointly drafted by the union and the Ministry of Labor and formally introduced by legislators (many of whom were former union leaders) of the governing FA. It was strongly supported by other union federations and the PIT-CNT, Uruguay's peaklevel central union, but fiercely resisted by all business associations. In the following months, SUNCA held hundreds of workplace assemblies and ran a public campaign to boost awareness of the safety hazards affecting workers in all industry sectors. It also staged day strikes in September and October. The union marched to Parliament again to present a petition with 300,000 signatures in November, when the bill reached the floor of the House. During the debate, the union's chief leader took his position as alternate representative of the FA to speak and vote in favor of the law.
In the Senate, several members of the moderate factions of the FA attempted to modify the bill to reach consensus with the opposition and the business community. The SUNCA pressured again. Eventually, a majority of FA senators passed the House bill in its original form, with workers demonstrating outside, and with the formerly rebellious senators celebrating the FA's historical alliance with the labor movement. Despite business groups' litigation, the Supreme Court ruled the law constitutional. By 2016, workplace accidents among construction workers had almost halved. But the SUNCA's struggle not only helped its principal constituency but also workers in all industries, whether unionized or not, benefited from a quarter reduction of workplace accidents.
These two examples illustrate the cross-country divergence in democratic practice, a concept coined by Fishman (2019: 6-18) to make sense of the more or less inclusive ways powerholders engage the collective actions and demands of relatively powerless citizens. Democratic practice constitutes a dimension of democratic variation that goes beyond the minimalist Schumpeterian standard of free and competitive elections. If all democracies need to meet this standard to be so regarded, some go further in scope and depth, promoting political inclusion beyond elections. While some democracies erect rigid barriers between elected officials and ordinary citizens, reducing the latter's participation to electoral instances, others promote their routine involvement as necessary for the exercise of political equality. Theories of democratic practice consider actor-centered, culturally embedded conceptions of democracy a critical dimension for the study of varieties of democracy; what political actors think the conditions for democracy's realization are plays a key role in how democracy is done (Fishman 2019: 16; see also Markoff 2011).
The two instances of labor conflict succinctly presented here reveal a fundamental discrepancy in the ways center-Left parties holding governmental power in Chile and Uruguay address and articulate the collective demands and aspirations of working-class actors. Like Fishman, I argue that democratic practices are highly consequential for the realization of democratic theory's normative principle of political inclusion. Yet, unlike Fishman, I do not attribute variation in democracy's inclusiveness to differing cultural practices rooted in either revolutionary or top-down transitions from authoritarian rule. Instead, I contend that what parties do with and beyond the legacies of transitional political junctures is also critical for democratic inclusion. How parties understand and "do" democracy when relating to economically disempowered actors matters. In particular, center-Left parties' articulation of class interests proves consequential for forging more or less inclusive democratic practices.
The political articulation school posits that parties' autonomous practices play a key role in the "political constitution of the social," that is, the "suturing" of politics, culture, and society through intensive integrative work across myriad social fields (de Leon et al. 2009(de Leon et al. , 2015Eidlin 2016). As a concept, articulation denotes the work of connecting and integrating diverse social constituencies to organize their inclusion and representation in battles over political authority. This work of "bringing together the constituents of the social" is done through "means of articulation" (de Leon et al. 2015: 2-3) such as party programs, platforms, and policies, but also through other less tangible practices such as grassroots organizing and mobilization, or even more informal work within cultural or academic fields.
An articulation-based explanation attributes causal efficacy to parties' practices, which researchers can only establish in counterfactual terms. The inclusion of comparative case studies that follow the logic of the most similar systems design (Moore 1993;Skocpol and Somers 1980) aids in identifying the relevant counterfactual processes and mechanisms. If anything, the articulation hypothesis posits that parties' political activities are accorded outcomes that would have not been observed had these activities not been undertaken (Desai 2002: 623-26;Eidlin 2016: 490-91). That parties' practices are endowed with sociologically consequential historical agency is, ultimately, the central claim of political articulation theory: "events generated by parties at particular historical moments can redefine the rules of the game and thus transform rather than merely translate the class struggle generated by underlying social structural variables" (Desai 2002: 625). I contend that the Uruguayan Left's sustained mobilization of a diverse array of means of articulation worked as a driving force of democratic inclusion. In particular, I highlight an important factor the comparative literature on the Latin American pink tide has largely overlooked: Left parties' contribution to strengthening the institutional power and autonomy of organized labor as representative class actor (but see Carneiro 2019).

Methodological Strategy
My analytical strategy is inspired by Riley and Fernández's (2014: 443-44) adaptation of Lakatosian epistemology, which seeks to move theories forward by identifying empirically salient historical anomalies that are at odds with existing explanations. Thus, I use the Chile-Uruguay comparison as a probe of an expanded political articulation framework, "both in the sense that it incorporates an anomaly"-one established typologies of Latin American Left turns failed to account for-"and [in the sense that it] predicts new facts" (ibid.: 444)-variations of Left parties' democratic practices.
The assessment of the political articulation hypothesis is informed by the historical-comparative method, which grounds the empirical evaluation of theories in in-depth interpretation of historical cases. The aim of this inquiry, however, is not to establish past events as an end, but rather to organize said events in temporal causal sequences, thereby identifying theoretically relevant similarities and differences across cases. For this specific research, I follow Mill's method of difference (Moore 1993;Skocpol and Somers 1980). I examine the processes and mechanisms of Left-party articulation in Uruguay-again, an INUS condition-as a part of a causal sequence leading to an outcome, against another plausibly similar sequence in which neither these mechanisms/processes nor the outcome are present (Chile).
First, I analyze data on labor statistics, protest events, and party platforms. Borrowing from Mudge's (2018) "inside-out" approach to political parties, I also look at the party experts in charge of postneoliberal social policy reforms. These analyses favor the political articulation hypothesis-a positive "hoop test" in Mahoney's (2012) sense. Yet, to prove the actual workings of Left-party articulation, I revisit key events organized in temporal sequences: within-case "causal-process observations" that evince party practices as "intervening steps between an initial cause and a final outcome" (ibid.: 579). I thus use secondary literature to develop a process tracing account of paramount historical instances of political articulation -the sort of "smoking gun" test (ibid.) that confirms the existence of the hypothesized causal pathway.

Labor and the Articulation of Class Politics
Capitalist democracies vary widely across many relevant dimensions, but those most egalitarian unmistakably grant special institutional status to relatively centralized industrial relations that (partially) decommodify labor (Esping-Andersen 1990). How specifically the state protects workers' collective rights to organize and awards unions access to collective bargaining institutions conditions the organizational and structural power of labor and, more generally, the dynamics of class conflict (Korpi 2018). Historically, moderate Left parties rooted in the social-democratic tradition have served as key vehicles for the institutionalization of class conflict.
On this dimension, the divergence between Chile and Uruguay could not be sharper. Despite enjoying congressional majorities during most of their terms, the Concertación governments in Chile-even when headed by socialist presidents-did not introduce substantial changes to the business-friendly labor code, politically marginalizing organized labor and further contributing to its fragmentation. In contrast, by coalescing with the labor movement to fight neoliberalism and then building neocorporatist institutions of centralized collective bargaining, the Uruguayan Left more deeply embedded the "class idea" (Eidlin 2018: 10) in the routine workings of democratic politics (see also Carneiro 2019: 9-10). Hence, the Uruguayan democracy more widely recognized the power imbalance inherent in class divisions. This recognition, in turn, furthered the universal provision of social rights.
As figure 2 shows, at the peak of state-led development-the so-called Import Substitution Industrialization period-Chilean workers were able to unionize at a much higher rate than their Uruguayan counterparts. The military regimes of the 1970s and 1980s virtually destroyed the organizational power of labor in both countries. After redemocratization, union density remained low through the early 2000s. However, unionization began growing in Uruguay in the mid-2000s, after the FA took power. The series ends in 2013, but unofficial documents of the central union assert that more than 35 percent of workers are currently unionized. Moreover, 95 percent of Uruguayan workers are covered by collective bargaining agreements negotiated at the industry level and overseen by a top-level tripartite wage council. The situation of Chilean labor is quite different. Take, for instance, union density in construction, a sector that generally employs large numbers of lowskill, blue-collar workers. In Chile, less than 7 percent of construction workers belong to a union; against 32 percent in Uruguay. More generally, only 15 percent of Chilean workers belong to a union, and the last available data shows that collective bargaining agreements, only extensive at the firm level, covered merely 18 percent of the total labor force.
Labor incorporation gaps reflect markedly different labor policy regimes. While Chilean institutions of industrial relations fit the individualistic, pluralist model of "interest group" mediation, the Uruguayan ones have more broadly recognized the collective rights of labor as a "class representative" (Eidlin 2018: 14-18). This different articulation of what Eidlin terms "the class idea" finds concrete expression in union certification. In Uruguay, firm-level unionization only requires the informal recognition of an industry-level federation, which lowers the cost of organizing and discourages fragmentation. In contrast, Chile follows a firm-level certification model that boosts the incentives for employers to resist unionization. In turn, collective bargaining laws restrict contract negotiations to the firm level without state mediation, while in Uruguay a centralized, hierarchical, tripartite structure dictates contract terms at the industry level, arbitrated by the state in case of disagreement. Likewise, right-to-work provisions in the Chilean labor code hold down union membership and undermine union security. Strikes are also tightly regulated in Chile, as well as severely restricted by employers' legally recognized right to hire replacement workers. Uruguayan lawmakers, in contrast, have been traditionally reluctant to regulate the right to strike, which is generally seen as an extension of freedom of speech and assembly. More generally, the internal characteristics and orientation of the labor movement in each country follow these contrasting patterns of working-class incorporation. A weak, fragmented, economistic unionism that mostly seeks to provide individual benefits to their members has prevailed in Chile (as illustrated by the contratistas movement). In contrast, the Uruguayan labor movement follows the social unionism model. More collectively oriented to broader social changes, labor has developed stronger ties with a broader social movement Left, advancing the interests of workers as a class beyond the "special interest" of individual union members, and endorsing universalist claims for redistribution and social rights provision (e.g., the SUNCA campaign for workplace safety). 3 Although comparable longitudinal data on protest events are currently unavailable, some preliminary data for 2009 suggests that different labor policy regimes and unionism also translate into divergent patterns of collective action. 4 That year, 71 percent of all protests events in Uruguay were led by unions, compared to only 27 percent in Chile. Movements' demands in Chile seem more heterogeneous: only 25 percent refer to labor issues (workplace conditions, wages, and layoffs), compared to 58 percent in Uruguay. Meanwhile, demands related to education represent 15 percent of total claims making in Chile, but only 4 percent in Uruguay. The use of peaceful tactics (marches, demonstrations, and rallies) is roughly similar in Uruguay (27 percent of protests) and Chile (25 percent of protests), while nonviolent but disruptive tactics such as strikes, occupations, and road blockages are used in 49 percent of protest events in Uruguay, but only 30 percent of protest events in Chile. Protesters' use of violent methods, such as destruction of private property, fights with the police, lootings, or bombings, also differ; violent tactics occur during 21 percent of the protests in Chile, but protesters only use violent tactics in 1 percent of protests in Uruguay.
The centrality of industrial conflict in Uruguay speaks to the presence of institutions that enable the exercise of workers' collective rights through union activity rather than to the level of grievances. Periods of high and low strike activity, indeed, typically follow a cyclical dynamic patterned by collective bargaining rounds and congressional budget appropiations (Etchemendy 2021). Likewise, the marginal presence of labor and the prevalance of violent tactics in Chile result from less institutionalized and correspondingly more autonomous forms of collective action (Somma et al. 2020).

Party Programs and Experts
The Uruguayan and Chilean democracies diverge remarkably in their efforts at recognizing and institutionalizing class politics as central to the political process, an outcome that is directly linked to Left parties' programmatic commitments to economic redistribution and welfare. I analyze data from the Manifesto Project (Lehmann et al. 2016)  time. These data quantify policy positions by systematically coding and counting text from official manifestos and programs presented by political parties during national elections across the world. I calculate the neoliberalism index for Left parties in Chile and Uruguay based on Mudge's (2018) analysis of programmatic convergence across Western parties since World War II. The index reclassifies and scores 16 variables of the Manifesto dataset in three dimensions: what the state should do, through what means, and for whom. For Mudge, neoliberal programs assume that the state's responsibility is to promote human capital investment, control social spending, and pursue law enforcement and social order, to the detriment of social protection and decommodification. Neoliberal party programs also prefer decentralization, administrative efficiency, privatization, labor productivity, and orthodox fiscal policy over economic planning, protectionism, universal welfare provision, and market regulation. Finally, neoliberal appeals generally forsake trade unions and blue-collar labor for business, finance, and professionals. 5 Figure 3 shows the scores of Chilean and Uruguayan Left party programs on the political neoliberalism index. Two caveats are necessary. First, because the Manifesto Dataset only surveys parties with seats in congress, the series for the Chilean Communist Party begins in 2005. Second, data on predictatorship Chile are not available, so I imputed an index score to the Popular Unity-the alliance of Communist, Socialist, and other leftist organizations that brought Salvador Allende to power in 1970-that is roughly similar to the Uruguayan FA in 1971. This imputation makes sense because the FA explicitly took inspiration from the Popular Unity government's program. 6 Overall, the trend in both countries is consistent with the general neoliberal shift in Western democracies with traditionally strong Left parties. Similarly, the index scores for both Chile and Uruguay indicate that even when the Center-Left was in power, government platforms included significant neoliberal elements. Political analysts have observed both these phenomena. What figure 3 evinces is some important differences across the two cases. The Left's neoliberalization was more incremental in Uruguay. In addition, the Chilean Concertación began flirting with neoliberal ideas much earlier. Although the scores at the end of the time series suggest that the FA may have shifted further to the right, what matters most is not the actual score but the timing and pace of neoliberalization. The Uruguayan Left adapted slowly to an increasingly neoliberalized political field. The FA's index scores were negative when it was in the opposition during the heyday of the Washington Consensus (the 1990s). Market ideas became more entrenched after the FA won national elections. Meanwhile, the Concertacion enthusiastically embraced the Third Way consensus already in the early 1990s. Politically marginalized, the Chilean Communist Party initially resisted programmatic moderation, until it was finally incorporated into the governing coalition.
Party programs provide important means for articulating demands and interests in civil society, but, alone, they cannot deliver. How party experts translate these 5 A full description of the index methodology can be found in Mudge (2018). 6 The coincidence between the two programs is striking: both the Popular Unity and the FA advocated state planning, sought to nationalize foreign trade and banks, and promoted agrarian reform by confiscating and partitioning large estates. programs into policy is also a key aspect of parties' articulation work. I draw from Mudge's (2018) "inside-out" approach that sees party politics as a specialized field within which party experts vie over the formulation of political problems and solutions. Speaking for both party leaders and those parties claim to represent, truthclaiming experts work from both within and in-between parties, bureaucracies, and knowledge-producing institutions to shape parties' capacity to intermediate.
A brief description of the major social policy reforms implemented during both the Concertación and FA governments reveals important differences in the affiliations and biographies of the party experts in charge. In Chile, the healthcare reform implemented during the administration of socialist Ricardo Lagos was designed by high-ranking public health experts and originally included provisions that partially redistributed private insurance funds into the public sector (Pribble 2013). But these provisions were eliminated after harsh congressional negotiations with two influential economists and Concertación senators: Alejandro Foxley (PhD University of Wisconsin) and Eduardo Boeninger (PhD UCLA). Besides limiting cross-sector redistribution, health insurance coverage was further reduced after the Minister of Finance, Nicolás Eyzaguirre (PhD Economics, Harvard University), resisted the incorporation of informal workers through a mandatory contribution. In a similar vein, the reform of social security of 2008 that secured government-funded minimum pensions while generalizing the privately run individual capitalization system was led by Alberto Arenas (PhD Economics, University of Pittsburgh) and Mario Marcel (PhD Economics, University of Cambridge).
Importantly, all these top technocrats followed similar career pathways: they were all trained in economics, pursued PhDs in Anglo-Saxon universities, and built professional reputations within finance-oriented institutions of economic governance. Marcel, for instance, had affiliations with the Central Bank (CB), the Ministry of Finance (MF), the World Bank (WB), the Organization for Economic Cooperation and Development (OECD), and the Interamerican Development Bank (IDB). Arenas served in the MF in both Bachelet governments. Eyzaguirre, too, served in the CB and as executive director of the International Monetary Fund (IMF). Foxley occupied high-ranking positions in the WB, IDB, and IMF. Both Foxley and Boeninger served as Minister of Finance and Chief of Staff, respectively, during the first Concertacion administration. And both, along with Marcel, played a leading role in CIEPLAN, the influential think tank behind the Concertación's intellectual conversion from state dirigisme to neoliberalized third-wayism (Silva 2008). Collectively, these party experts shared the common mission of ensuring that top-down social policy reforms preserved fiscal orthodoxy (Pribble 2013). Marcel, in fact, was behind the design of the severe fiscal rule that imposed quantitative caps on government budgets.
The party experts involved in the making of social policy reforms in Uruguay look rather different. The implementation and subsequent universalization of family allowances was led by party officials (some of them trained as sociologists) appointed to the Ministry of Social Development. Importantly, these bureaucrats worked in close collaboration (Amarante and Vigorito 2012) with a team of heterodox academic economists specialized in labor markets, poverty, inequality, and social security, that is, Andrea Vigorito (MA London School of Economics), Verónica Amarante (PhD University of Sussex), and Rodrigo Arim (MA University Torcuato di Tella). All these economists had a past of informal participation in the FA's programmatic committees and fora, and all held full-time faculty positions at Universidad de la República (UdelaR), the country's public university and largest research institution-Arim, in fact, currently serves as University Rector. Another party official with a say in social policy reforms in Uruguay was Ernesto Murro. Educated as a schoolteacher, Murro's political career started as a leader of the timber union, and then occupied the central union's seat at the board of the Social Security Bank (SSB). Upon the FA's victory, Murro became president of the SSB and then served as Minister of Labor. Murro's chief advisor during his two terms as head of the social security administration was Gabriel Lagomarcino, a career bureaucrat of the SSB who also had an activist past in the labor movement and was trained as an economist at UdelaR. Both Murro and Lagomarcino oversaw the reform of family allowances and presided over the expansion of the publicly managed social security system. Finally, the architect of healthcare reform was Marxist economist Daniel Olesker (MA Leuven University), also a UdelaR professor and senator of the Socialist Party. Before heading the Minsters of Health and Social Development during different FA governments, Olesker collaborated with the Uruguayan Federation of Housing for Mutual-Support Cooperatives. He also directed the labor-affiliated Instituto Cuesta Duarte, a research organization that provides the central union with crucial technical advice in economic and labor policies.
The biographical trajectories of Uruguayan and Chilean reformers fit nicely the two ideal-typical categories of party expert Mudge (2018) constructs to illustrate Left parties' transition from Keynesianism to neoliberalism. Most party experts with a say in social policy in Uruguay epitomize late incarnations of the economic theoreticians that prevailed in the heyday of the Keynesian era. At once academic-and party-affiliated (if not labor-affiliated), these experts of Keynesian persuasion embodied a peculiar ethics: a professional imperative of translating popular demands into sound socioeconomic policies. Thus, the economic theoreticians undertook the crucial intermediation work of providing the technopolitical language and knowledge necessary for decommodification and social protection in the context of Polanyian-like double movements. In contrast, the social policy reformers who weild the upper hand in Chile resemble the transnational, finance-oriented economist that, according to Mudge, ended up dictating the programmatic language and policies of neoliberalized Left parties. Relieved of their intermediation duties, these party experts no longer need to speak for labor or any other voiceless constituency whatsoever, except for abstract and-allegedly -apolitical markets.

Process-Tracing Analysis
For more than five decades, both Uruguay and Chile have had electorally competitive Left parties embracing programs and policies that broadly seek to represent the voices and demands of powerless and socially marginal actors. Yet a closer look at what these parties have done and how they have developed reveals they are in fact organizations of very different sort. To complete the articulation-based explanation, this section presents a succinct process-tracing account of how these parties have evolved strategically and organizationally to exploit and transform the opportunities and constraints of the postdemocratization aftermath.

Chile: A Demobilizing/Disarticulating Left
In the wake of democratization, the Socialist Party of Chile broke its historical alliance with the Communist Party, underwent deep ideological revisionism (Ostiguy 2005), and united the Christian Democratic Party and other smaller progressive groups into the Concertación coalition, which then ruled uninterrupted for two decades. Initially, both the Socialists and the Christian Democrats exploited their traditionally strong links with popular sectors to promote demobilization (Oxhorn 1994;Posner 2008). Overtly committed to securing consensus with the civilian arm of the dictatorial regime, the government coalition subordinated effective representation to the achievement of political and economic stability (Motta 2008)-the so-called democracy of agreements (Joignant 1999;Luna and Mardones 2010). This strategy meant that Concertación parties rapidly evolved into professional-electoral organizations, relinquishing their ties to their social bases. The Communists, however, were completely marginalized from the political arena, and lost their historically high number of seats in Congress, to which they would not return for more than a decade.
Certainly, left-of-center Chilean governments promised to reverse the most regressive aspects of neoliberal reforms bequeathed by Pinochet. Promoting social justice and popular participation figured at the top of the political agenda, particularly after the Socialists took over the presidency in 2000. Nevertheless, the prevalence of a technocratic, consensual style of policy making in the context of widespread demobilization permitted only timid, incremental reforms. Regulating marketlike schemes of social service delivery, rather than decommodifying and universalizing social rights, became the Concertación's raison d'être, thereby limiting popular sectors' demands for reincorporation and redistribution.
In this context, the Chilean state remained reluctant to institutionalize and promote class compromise. Unions' demands for state protection and collective bargaining rights produced only tepid responses from political elites. With no exception, all Concertación administrations sought to amend the authoritarian labor code originally instated by Pinochet, incrementally expanding workers' individual rights. Yet, workers' collective rights to organize, strike, and bargain beyond the firm level remained seriously constrained (Carneiro 2019), even as social protests began to erupt and the Communists joined the Concertación coalition for the second Bachelet presidency (2014-18). Ultimately, reformers prioritized consensus with the strongly organized business community and the powerful right-wing parties in Congress. Reduced to an interest group in advisory committees mainly staffed with consensus-seeking state technocrats, an increasingly weak and fragmented labor movement abandoned any hope of elite concessions and shifted to more confrontational tactics (Traverso et al. 2012).
In sum, Chile rapidly evolved into a low-intensity, highly institutionalized, elitebased market-democracy-a competitive oligarchy, according to Luna and Altman (2011). The traditional programmatic structuration of party competition continued only as a façade. Increasingly relying on personalist/clientelistic strategies of vote seeking to the detriment of programmatic appeals, political parties in general, and Left parties in particular, relinquished their linkages with their grassroots. Chilean civil society, in turn, grew increasingly fragmented and autonomous from political parties. Confronting the political class as whole, mass protest movements broke into the political arena as ordinary people's last resort for voicing the multiple grievances created by widespread market insecurities (Somma 2017;Somma and Medel 2017). The recent social uprising only confirmed the sharp divide between the political class and ordinary people. To end pervasive street violence and uncontrolled police repression, political parties across the political spectrum acceded, as a last resort, to a binding consultation on the election of a Constitutional Convention. Whether citizens' disaffection will change with the enactment of a new constitution remains to be seen.

Uruguay: A Mobilizing/Articulating Left
Unlike in Chile, Uruguay's dominant party of the Left camp-the FA-still serves as the political expression of its founding constituency: organized labor, student and housing cooperative movements, the cultural intelligentsia, and, more generally, salaried middle sectors. The FA was founded in the aftermath of the cycle of protests of 1968 (Markarian 2016) to contest the hegemony of elite-based, catchall, patronageoriented parties that virtually monopolized access to political office. The FA originated as a hybrid organization combining a movement base of social activists with a political coalition of Socialists, Communists, Christian Democrats, National Liberation groups, and progressive fractions that defected from traditional parties.
After redemocratization, the FA maintained the original features of an externally mobilized, mass party with strong linkages with the labor, student, feminist, human rights, and cooperative movements. With time, the party developed into a complex and highly institutionalized organization that preserved its dual movement/partycoalition structure. On the one hand, a dense network of local sections run by a massive and territorially spread cadre of activists became disproportionally represented in the party's governing bodies, imbuing these activists with a sense of collective efficacy that fosters their engagement, recruitment, and mobilization between elections (Bentancur et al. 2019). On the other hand, the founding organizations (e.g., the Socialist Party, the Communist Party, the Christian Democratic Party) and other groups such as the former guerrilla MLN-Tupamaros became integrated as party fractions in the party's directorate and permanent committees. Importantly, not only did the statutory rules adopted in the wake of redemocratization foster consensus and balance among fractions; by affording voice and veto power to base activists at all levels of the party's structure, the party's bylaws prompted all fractions to compete for grassroots support to influence the party's orientation, thereby furthering the organization's openness and responsiveness to bottom-up participation (Bentancur et al. 2019;Yaffé 2005).
Unlike the Chilean Left, the FA remained basically united when democracy was restored. Indeed, the Socialists', Communists', and far-left activists' dominance over the party congress-entirely integrated by base committee delegates and tasked with approving the electoral program and nominating the presidential candidates-provoked defections from the Right, not from the Left. Thus, the Uruguayan FA reincorporated into democracy (1980-85) as a challenging, independent third party, refusing electoral compromise with centrist and progressive fractions of traditional parties, now forced to cooperate to keep control of both the legislature and the presidency through the mid-2000s. Left alone in the opposition, the FA maximized what Luna (2014) calls a "strategic harmonization of party-linkage segmentation" (pp. 92-94, see also ch. 5): the electoral mobilization of a broad and diversified social base through a combination of programmatic appeals, strong political leadership, and party identity. Indeed, the articulation of heterogeneous political demands against the dismantling of contributory welfare schemes and in favor of social assistance for the poor and the unemployed reinforced programmatic alignment of the party system along the state/market cleavage. This in turn allowed the FA to expand its popular base, reaching out to those who, in the context of pervasive fiscal crisis and state retrenchment, were increasingly marginalized from the clientelistic networks that tied traditional parties to low-income citizens (Handlin and Collier 2011;Luna 2014;Luna and Filgueira 2009;Yaffé 2005).
A crucial component of the FA's mobilization/political articulation strategy was the systematic use of binding mechanisms of direct democracy activated by popular initiative (Altman 2010;Bidegain and Tricot 2017;Moreira 2004). Citizens' use of mechanisms of direct democracy was the order of the day especially in the 1990s and the early 2000s, when the Left was in the opposition and traditional parties advanced their neoliberal agendas. During that period, opposition forces resisted neoliberalism through campaigns to block reforms at the ballot box. Trade unions, social movements, and leftist activists all converged, time and again, on civic campaigns aimed at collecting petition signatures to trigger referenda and constitutional plebiscites. The sustained mobilization required for these national campaigns dominated the repertoires of collective action throughout the period. Between 1989 and 2015, electoral authorities convoked more than 14 legally binding, mandatory popular consultations, and many other failed attempts nevertheless entailed significant campaigning and stimulated debate. These actions contributed to a culture of unity and common identity within the Left camp. A new grammar to designate the party/ movement alliance emerged among activists: the so-called social bloc for change represented intangible political capital that all leftist groups and social organizations felt a duty to preserve (Bidegain and Tricot 2017).
Through mechanisms of direct democracy, therefore, an active civil society, allied to an opposition Left party, translated collective protests against neoliberalism into electoral mobilization. An important milestone in this struggle was the popular repeal, in a referendum held in 1994, of a law allowing the government to privatize state-owned companies. In the years that followed, several referenda against demonopolization, privatization, and deregulation of state-controlled industries took place. Although the government coalition managed to demonopolize insurance markets, privatize the national airline, outsource the administration of port services, and introduce private retirement accounts into social security, the opposition bloc led by the FA successfully blocked many other reforms. In 2003, amid a severe economic recession, the government coalition was clearly defeated in a national referendum against a law allowing the state-owned oil company to become publicly traded. That year, to avoid more embarrassing defeats, the government retracted a similar effort to privatize equity in the state-owned telecommunications company. The threat of direct democracy eventually convinced traditional parties to abandon their privatization plans.
This decade of protest victories through direct democracy closed with the turn to the left in the 2004 national elections, when the electorate further approved a constitutional amendment initiated by the water utility union that ensured state monopoly on drinkable water production and distribution. By that time, the FA's efforts at electoral mobilization peaked. About 400,000 supporters, more than 40 percent of the franchised population of the capital, attended the rally that closed the electoral campaign. Police reported a similarly high turnout at the rally leading up to the national election of 2009, where the FA renewed its presidential mandate, again, with congressional majorities. The FA's deep roots in civil society were reconfirmed during nonmandatory elections of party authorities: 20 percent of FA's voters showed up at the party's local sections to elect the party authorities in 2006, and 15 percent in 2012.
Already in power, and putting its congressional majorities to good use, the FA expanded both contributive and noncontributive family allowances, created a mandatory single-payer health insurance, reformed pension law to ease access to public pensions, created a progressive income tax, and significantly increased spending in public schools and hospitals. More crucially, the Left honored its historical alliance with the labor movement, passing more than 50 labor laws that, among other things, significantly reduced informality, protected union rights, and extended centralized, corporatist forms of collective bargaining at the industry level in all sectors of the economy.

Discussion: Mass Parties, the Left, and Democracy
This article advanced a Left-party articulation account of divergence in democratic inclusion. Chile and Uruguay looked very much alike in the aftermath of military dictatorship. In both cases, democracy could only be achieved through complicated (and contested) pacts between a strong military regime and a mobilized opposition led by political parties. In exchange for the military's gradual withdrawal from the center of the political arena, party elites promoted demobilization. Neoliberal democracy grew fearful and conservative, soon generating widespread disenchantment. Institutionalized, electorally competitive center-Left coalitions therefore had to navigate an historical period in which market-oriented forms of socioeconomic regulation alienated popular sectors and strained the legitimacy of democratic regimes. These similarities notwithstanding, as I have shown, the Left developed as a tightly coupled, mass party organization, strategically oriented toward mobilizing civil society and articulating-in social actions, political programs, and public policy-a politically inclusive democratic practice in Uruguay but not in Chile.
How Left parties engaged popular sectors and articulated on a class basis proved consequential for political inclusion (or exclusion). While the Uruguayan Left managed to push the traditional catchall parties to the right, realign the party system, and increasingly represent low-income voters (Luna 2014), the Chilean Left could never recover the capacity to speak for working class voters it had so skillfully built in the predictatorship period (Scully 1992). As figure 4 shows, in Chile, average marginal effects of citizen's socioeconomic status (SES) on electoral support for Concertacion parties remained stable and statistically nonsignificant during the period 1995-2015. 7 In Uruguay, however, the FA's electoral growth and eventual ascent to power made citizens' vote progressively align with class cleavages. Indeed, at the beginning of the series, higher SES was positively associated with Left vote. But the direction of this association reversed as the FA made inroads into informal workers and salaried middle sectors outside major urban centers. By the end of the period, as the FA government implemented much of its program, highincome status became negatively associated with Left vote. 8 Thus, the Left's capacity to foster the political inclusion of the powerless increased in the postneoliberal aftermath.
Some conclusions follow. First, the formal conceptualization of Left-party, laborbased articulation as an INUS cause of democratic inclusion speaks to the need of better specifying party-centered accounts in comparative-historical sociology. Beyond the general claim that Left parties' articulation practices matter, this article empirically identified an array of strategies that articulate on a specifically class basis, and that are distinct from articulation strategies based on, say, religion, national identity, or other identity-based social movements. A more strongly organized and mass-oriented Left in Uruguay enabled a deeper recognition of class politics in democratic arenas. Fostering the power and national integration of organized labor also provided a stronger organizational infrastructure for the political expression of other civil society organizations and, more generally, the promotion of social citizenship. This was done through a series of articulating strategies. Specifically, the Uruguayan Left, unlike the Chilean one, granted considerable sway to its grassroots, which acted as a buffer against full-fledged neoliberalization. Before winning national elections, the Left actively encouraged mass mobilization, building alliances with labor and social movements to resist neoliberal reforms through binding mechanisms of direct democracy. Thus, the FA programmatically committed to restructuring the institutional bases of labor-capital relations, boosting the associational power of unions, and elevating their status as a unified class representative, that is, above special interests, in the political arena. The Uruguayan Left also staffed the government with social policy experts recruited from within the labor movement and progressive academic circles, who helped devise a more universal welfare state.
The autonomous practices of the Uruguayan Left strengthened the political articulation of class interests but, seen in historical terms, this articulation strategy was far from one-directional. The labor movement and its allied organizations had first contributed to the formation of the FA in the 1970s after a bottom-up process of unification, coalition formation, and mobilization. The "social bloc" remerged after dictatorship, articulating popular interests in civil society, and again resorting to the FA as the electoral vehicle that was better poised to channel such interests into the 7 Marginal effects are estimated through a moderation effects model, i.e., by interacting SES and year in a logistic regression that predicts the odds of voting for FA or any Concertacion party, controlling for sociotropic assessment of the state of the economy, trust in political parties, and self-declared ideology. Models are run separately for each country. The data comes from nationally representative public opinion surveys conducted by Latinobarometro Corporation. Replication syntax available upon request. 8 Other studies based on different data have also shown this trend. See e.g., http://www.razonesypersonas.com/2020/07/la-emergencia-del-voto-de-clase-en.html, accessed March 5, 2021. political arena. Class-based articulation was therefore instilled organically from below, too. It worked as a two-way road. This reasserts the core thesis of power resource theory (Esping-Andersen 2017; Korpi 2018), namely, that the structural power of labor within capitalist democracy is not an historical given, but a byproduct of mutually reinforcing processes of organization and mobilization of trade unions and Left parties.
Secondly, instead of simply restating the causal import of party-driven articulation in general, this research evinces that, under certain conditions, Left parties' class-based articulation strategies produce a specific outcome: inclusive democracy. Taking heed of this finding is relevant to contemporary debates on democratic decay. The role of Left parties in forging inclusive democratic practices speaks to the centrality of what Peter Mair (2013: 29) called the "popular component" of democracy. Theorizing the rise of the cartel party, Mair argued that besides constitutional checks and balances and the protection of civil liberties, mass democracy requires mass involvement in political affairs. Crucially, this "popular component" hinges on the flourishing of mass parties as multipurpose organizations that mobilize the citizenry, recruit political leaders from civil society, aggregate demands into programs, deliberate in parliament, formulate policy, and staff state institutions, enabling government by (and not only for) the people. For Mair, when parties relinquish their ties to civil society and retreat into state bureaucracies, when they become specialized, state-regulated agencies for teams of professional politicians of an undifferentiated ruling class to seek office and influence governance, we are left "with something that might still be called democracy, but which has been redefined so as to downgrade or even exclude the popular component" (ibid.: 32). Highly consequential for democracy's popular component, my analysis shows, is whether Left parties develop a capacity to act as mass-based organizations, externally mobilizing civil society and articulating the voices of subaltern sectors.
This finding, in turn, supplements Ziblatt's (2017) thesis that the organization of parties on the Right is decisive for democracy's survival. Ziblatt's well-grounded analysis shows that conservative parties' organizational features such as centralized leadership, the professionalization of party cadre, their networking with civic associations, their autonomy in fundraising, and the shielding from capture by antipragmatic ideologues ultimately determine whether democracy consolidates and thrives. Only strongly organized and autonomous conservative parties can overcome collective action problems, successfully articulating the interests of recalcitrant upper classes within, and not against, the democratic process.
My argument is symmetric with respect to Left parties and democratic inclusion. For Ziblatt, economic elites' support for democracy results from conservative party formation, strengthening, and institutionalization. Implicit is the premise that (capitalist) democracies are born inherently conservative, or they do not survive at all. That the fate of democracy as a political regime is inevitably attached to conservative parties' success at defending the interests of a wealthy and politically powerful minority clashes with the imperative of political equality and popular sovereignty on which the myth of democracy rests (Markoff 1996). In other words, democracies exist in an ever-fragile equilibrium between the factual impossibility of their full realization and the always latent threat of complete illegitimacy.
Chile's and Uruguay's Left parties faced a similar dilemma during the democratic transition. Reincorporating into the democratic game required complying with the conservative path of economic modernization initiated by the military and rightwing civilian elites. Some mobilization containment and programmatic moderation was in order. Otherwise, regime incumbents would relapse into democratic breakdown. Too much compliance, however, would also blur the Left's raison d'être and completely alienate its founding constituencies. In light of this dilemma, two possible strategic organizational responses were at hand. One was to prompt full-fledged ideological revisionism, do without the traditionally strong programmatic linkages with popular sectors, and incorporate into the ruling coalition to enact incremental fixes to the neoliberal regime. The other strategy was to compromise less, remain in the opposition, and resist market reforms, still preserving some of the mobilizational, organizational, and programmatic characteristics of a mass-based Left party. Each strategy came about with quite different consequences. The Chile-Uruguay divergence confirmed, again, that a strongly organized mass party of the Left is necessary to overcome collective action problems and articulate popular interests within capitalist democracies, thus serving as a driving force of democratic inclusion. Left parties may not be a necessary condition for a given democracy's evolving in a continuous, incremental way. But they may be a sufficient condition for a particular democracy's approximation to the normative ideal of political equality.