Loyalty Dilemmas and Market Reform: Party-union Alliances Under Stress in Mexico, Spain, and Venezuela
Katrina Burgess *
During the twentieth century workers in Latin America and Western Europe were often incorporated into unions with institutionalized links to political parties. In the wake of economic crisis, globalization, and market reform, these links are coming under stress. Unions find themselves struggling to maintain their membership and to respond effectively to the double movement of globalization and decentralization, and they see their mediating capacity and political legitimacy being severely tested in many cases by the adoption of market reform by their party allies in government. The result is a crisis of representation that threatens to alter profoundly the mechanisms by which workers are (or are not) linked to the state.
This article analyzes the impact of these trends on historic alliances between governing parties and labor unions in Mexico, Spain, and Venezuela. All three countries were governed in the 1980s and early 1990s by parties with historic ties to the labor movement. In Mexico the Institutional Revolutionary Party (PRI) or one of its antecedents had been organically linked to the Confederation of Mexican Workers (CTM) since the 1930s. In Spain the Socialist Party (PSOE) had been closely associated with the General Workers' Union (UGT) since the 1880s. And in Venezuela, Democratic Action (AD) had dominated the Confederation of Venezuelan Workers (CTV) since the 1940s. These parties owed much of their success to the support of affiliated unions, which, in turn, enjoyed a degree of influence that outweighed their organizational presence among workers.
In the 1980s and early 1990s all three parties adopted economic reforms that violated their long-standing principles by imposing painful sacrifices on organized labor. These reforms--which included fiscal and wage austerity, trade liberalization, price deregulation, flexibilization of the labor market, privatization of state-owned enterprises, industrial restructuring, and welfare reform--subjected workers to declining incomes, unemployment, decreased job security, and cuts in social services. They also reduced the mediating capacity of unions by shrinking the size of the public sector and expanding the influence of neoliberal technocrats in policy-making circles. Yet party-affiliated labor leaders responded differently in each case. In Mexico they complained bitterly about the reforms but neither organized serious protest nor took steps to extricate themselves from the alliance. In Venezuela they came closer to breaking their ties with the party but ultimately retreated to their traditional strategy of negotiating behind closed doors. Only in Spain did they abandon their historic ties with the party in response to its market reforms.
I argue in this article that the strategic choices of disaffected labor leaders can be understood with reference to Albert Hirschman's schema of exit, voice, and loyalty. 1 I propose that voice and exit constitute points along a path of alliance change, and that loyalty interacts with authority structures to affect the pace and direction of this change....
Loyalties under Stress
Political parties and labor unions began to forge alliances in Latin America and Western Europe with the advent of industrialization and mass politics in the late nineteenth and early twentieth centuries. 2 Although the earliest unions tended to be apolitical or anarchist, unions and parties discovered shared interests as workers developed a class consciousness, the state increased its involvement in the economy, and working-class support became a valuable commodity in the struggle for state power. These conditions laid the foundation for an exchange whereby the union would help the party win or maintain control of the state in return for the party's commitment to use state intervention to channel benefits to labor. Although many labor-backed parties initially advocated radical state intervention in the form of socialism or communism, participation in the political system usually led them to moderate their programs in favor of using state power to promote the interests of workers in the context of a capitalist economy. 3
These exchanges facilitated the construction of stable, institutionalized alliances between parties and unions that were informed by strong bonds of loyalty. Rather than being ad hoc and dictated by short-term interests, these alliances resembled "regimes" as understood by international relations scholars. Stephen Krasner defines regimes as "sets of principles, norms, rules, and decision-making procedures around which actors' expectations converge." 4...A defining feature in both cases is the nonunitary character of the central actors. Like nation-states, parties and unions have complex internal structures, identities, and constituencies. They can enter into regularized patterns of interaction with one another, but they do so as collective entities rather than as individuals. Thus, they must consider more than their individual preferences; they must also take into account the potential responses of their followers. In addition, these central actors are relatively few in number, which means any one of them can have an impact on institutional outcomes....
Most party-union alliances in Latin America and Western Europe in the twentieth century developed around the principle of state intervention in the economy. This principle generated similar norms regarding the rights and obligations of each partner. In the socioeconomic arena the party was expected to use the power of the state, when available, to deliver wage and nonwage benefits to labor, as well as to adopt procedures that gave labor leaders a role in formulating and implementing relevant policies. In return affiliated unions were expected to moderate their strike activity and to refrain from maximizing their wage demands when the party was in government, particularly in times of economic crisis. In the political arena the party was expected to give labor leaders access to public office and party posts in return for the union's delivery of the working-class vote in elections. Finally, in the organizational arena the party was expected to seek (or maintain) the hegemony of its labor allies in the labor movement in return for legitimation as a representative of the popular classes. Although the party could deliver many of these benefits only when in government, it exchanged the promise of their delivery for the union's support when in the opposition....
In this context party-affiliated unions had the opportunity to mediate between the party and workers, particularly when the party was in government. In effect, they became managers of a virtuous circle of loyalty. Party leaders reinforced the loyalty of labor leaders by offering them policies favorable to workers and unions, union subsidies, access to public office, and participation in policy formulation. Labor leaders, in turn, used these resources to reinforce the loyalty of workers by offering them higher wages, increased employment, job security, welfare benefits, and other socioeconomic improvements. The loyalty of these workers was then translated by labor leaders into electoral and organizational support for the party, thereby reinforcing the party's loyalty to labor. Although these loyalties were never entirely unconditional, relatively stable identities and predictable economic trends allowed for the accumulation of reserves of loyalty that could be tapped in times of stress and replenished in times of prosperity.
These conditions favoring the building of reserves of loyalty between parties and unions began to deteriorate in the late twentieth century, however, as both identities and economic conditions became more fluid....
Hirschman posits exit and voice as alternative responses by customers of a firm or members of an organization to a decline in the quality of the goods and services they are receiving. Exit occurs when they abandon the relationship altogether. Voice occurs when they make "any attempt to change, rather than to escape from, an objectionable state of affairs." 11 Hirschman finds that exit is the most likely response because it is relatively costless and straightforward. Once an individual has gathered information about alternative products and organizations, he or she can exit without investing in political resources or organizing collectively. Voice, by contrast, "is costly and conditioned on the influence and bargaining power customers and members bring to bear within the firm from which they buy or the organization to which they belong." 12
Party-affiliated labor leaders face an analogous set of choices when the party adopts market reform and thereby devalues their terms of exchange. Assuming they are unwilling to remain silent, they can either pressure the party to reverse the decline (voice) or defect from the alliance altogether (exit). Unlike the atomized individuals in Hirschman's model, however, they are likely to find voice to be less costly than exit. Exit is discouraged by the uncertainty and risk associated with abandoning regularized patterns of interaction. 13 Meanwhile, voice is facilitated by the channels of communication already built into the alliance....
In addition, the small number of actors in a party-union alliance is likely to raise both the resonance of voice and the costs of exit. In part, this is true because exit by a single actor (for example, the union) will lead to a collapse of the alliance, whereas organizations or firms can usually withstand the defection of numerous members or customers. Assuming the party still wants to preserve the alliance, this possibility lends greater weight to the threat of exit, thereby enhancing the effectiveness of voice. But by the same token, exit itself cannot be undertaken quietly or inconspicuously. It is inevitably a loud, political act. Combined with the other obstacles to exit, these factors are likely to make voice the default option for disaffected labor leaders in a party-union alliance....
Venezuela's labor leaders came closer to breaking their alliance with the governing party. After several years of limiting their voice to rhetorical attacks on the economic cabinet and threats to mobilize workers against the government's austerity policies, the CTV joined with Venezuela's other labor confederations in May 1989 to call a general strike against the reforms of the new AD government of Carlos Andrés Pérez. The strike was an unprecedented act of norm-breaking voice that brought the economy to a halt and severely damaged the government's credibility. Venezuela had not experienced a general strike since 1958, when AD and its labor allies had joined with other prodemocracy forces to protest an attempted military coup. Nor had the CTV ever taken such a combative stance while AD controlled the presidency. But rather than representing a step down the path toward exit, the general strike turned out to be an exception to the rule of party-union collaboration. Despite the failure of the general strike to reverse the reforms, labor leaders quickly returned to a strategy of norm-based voice. 16...
I argue that the authority structures in which labor and party leaders find themselves shape the choice set among these different solutions. For labor leaders, the most relevant structures are those that determine the relative power of the party and workers to sanction them for disloyal behavior. The party's sanctioning power is closely linked to its past and present relationship with the state. A hegemonic party that has been in power for many years will control greater resources than a party that has spent decades in exile or must share power with other parties. Likewise, a party with the authority to withdraw benefits already granted will enjoy significantly more sanctioning power than a party restricted to using inducements as a bargaining chip only prior to their delivery. Both of these factors are affected, moreover, by the nature of the state itself. Controlling a weak state in an economy dominated by private interests will bring the party far fewer resources than controlling a strong state capable of imposing its will on domestic and foreign actors.
The sanctioning power of workers is strongly influenced by the dynamics of union competition and, relatedly, by the legal framework governing union formation, collective bargaining, strike activity, union elections, and membership in union organizations. 27 This framework structures the universe of workers ostensibly represented by affiliated labor leaders. Within that universe it affects the extent to which these workers can mobilize independently, remove leaders within the union, switch allegiances to a competing union, and control inducements that labor leaders may want to deliver to the party, such as wage restraint, industrial peace, or working-class votes. To some degree, the sanctioning power of workers varies inversely with that of the party, since a party empowered to intervene extensively in union affairs is likely to constrain the independent actions of workers....
Venezuela: Party Autonomy and Pressure from Both Sides
In contrast to both Mexico and Spain, party-executive authority structures in Venezuela gave AD leaders enough autonomy to challenge their own government. Supreme authority within the party resided with the National Executive Committee (CEN) rather than with the chief executive. Once in office, an AD president had no formal position in the party executive and exercised no direct control over the party apparatus. The CEN, by contrast, controlled the placement of party candidates on closed lists for the Senate and the Chamber of Deputies, preselected the head of the party's parliamentary fraction, dictated the party line in congress, and even had final say over when an AD president could form a multiparty coalition to pass legislation. Moreover, the president had only limited influence over the nomination of his successor. 50 In contrast to Mexico, the party's nominating convention enjoyed both formal and de facto power to select its presidential candidate, and its delegates often voted down the president's choice.
These spaces for autonomous action enabled AD leaders to join the CTV in resisting the government's reforms at a critical moment in party-union relations (solution 3). Not long after CTV leaders organized the 1989 general strike, AD eased their loyalty dilemma by defying Pérez and behaving much like an opposition party. 51 AD's bitter differences with the executive broke into the open after the party's disappointing showing in state and local elections in December 1989. The AD leadership publicly blamed the results on the economic reforms and the lack of AD representation in the cabinet. The struggle became increasingly vitriolic in 1990 and 1991, culminating in a landslide victory by an anti-Pérez faction, led by Luis Alfaro Ucero, at AD's national convention in October 1991.
This feud had two salutary effects on CTV leaders. First, it effectively freed them to treat the Pérez administration like a non-AD government, thereby transforming actions that would normally constitute norm-breaking voice into norm-based voice. As in the years when AD's main rival, COPEI, was in power, the CTV got the green light to support antigovernment initiatives in congress, carry out joint actions with other labor confederations, and mobilize workers. Second, AD's congressional delegation took the CTV's side against the executive on several important issues. In 1990 AD supported a labor law enhancing worker rights and union privileges despite strong opposition by Pérez and his cabinet. In 1991 AD voted down an executive decree establishing a salary increase below the rate of inflation and suspended deliberations regarding a controversial bill offered by the executive to rationalize severance benefits.
These developments brought CTV leaders important relief from the loyalty dilemma created by Pérez's economic reforms. Besides being able to resist the reforms without engaging in disloyal behavior toward the party, they received crucial help from the party in postponing reforms that would have put even greater strain on their relations with workers. This relief arrived, moreover, at a critical moment in party-union relations. By the time outright confrontation between AD and the Pérez administration subsided in 1992, the momentum had shifted back to norm-based voice. A pro-Pérez leader had become president of the CTV, and two attempted military coups had diverted attention from the government's reforms. Furthermore, the episode reaffirmed the possibility that CTV leaders could find allies within the party in their struggle against the reforms.
Had CTV leaders been highly vulnerable to pressures from below, temporary access to solution 3 might not have been sufficient to salvage the alliance. But instead they faced an intermediate distribution of sanctioning power. On the party side AD had a significant degree of leverage but not as much as the PRI. After a decade in exile AD controlled the presidency for twenty-five of the thirty-five years following the democratic transition in 1958 and always held the largest share of seats in the congress. In addition the 1936 Labor Law, which was modeled after Mexico's LFT, gave the state extensive authority to intervene in union formation, collective bargaining, strikes, and union elections. Furthermore, AD's organic ties with individual unionists translated into significant influence over leadership selection within the labor movement. The party decided which AD labor leaders would be placed on candidate lists for union elections; these positions belonged to party in the sense that it could replace a suspended or expelled leader. 52 Thus, AD unionists depended on the party not only for party posts or elected office but also for their upward mobility within the union.
Despite these powerful levers of control, AD's sanctioning power was constrained by two factors. First, unlike the PRI (or the PSOE in the 1980s), AD had to negotiate the terms of its hegemony with other parties. Both the political system and the labor movement operated on the basis of power sharing and proportional representation among the major parties, which meant that AD could not act with impunity in either arena. Second, labor leaders enjoyed a degree of influence within the party that their counterparts in both the CTM and the UGT lacked. Like the CTM (and in contrast to the UGT), AD's Labor Bureau constituted the most unified and organizationally autonomous grouping within the party. 53 But unlike the CTM, the Labor Bureau could translate this unity into pivotal influence over leadership selection within the party. 54 As a result, party leaders had strong incentives to win the loyalty of the Labor Bureau, particularly during periods leading up to internal elections.
On the worker side, CTV leaders enjoyed significant protections from worker dissent but were more vulnerable than their counterparts in Mexico. A combination of indirect elections, negotiated slates, disproportionate voting power for small unions, and "delegates for life" protected high-level leaders from the kind of voters' unionism that prevailed in Spain. At the level of the firm, however, Venezuelan workers had real if limited access to both voice and exit. The Venezuelan congress removed exclusion clauses from an early draft of the 1936 Labor Law, and parallel unions were widespread, particularly in the public sector. In addition, nonunionized workers had the right to bargain collectively if they represented at least 75 percent of the workers in the firm. These features gave workers the option of exiting from a union that did not serve their interests.
More importantly, party pluralism made if possible for workers to punish incumbent leaders by supporting challengers from another party. Although the parties continued to negotiate united slates at the federal and confederal levels into the 1980s, this practice became less and less common at the level of the firm. As a result AD faced serious challenges to its dominance in strategic sectors of the economy. In 1987 it lost control of Venezuela's largest union (SUTISS, located in a state-owned steel complex in the industrial state of Bolívar) to the Radical Cause Party. 55 Although a one union-one vote rule prevented SUTISS and other radical unions from gaining seats on the governing boards of higher-level organizations, AD faced the possibility of losing its majority on the CTV executive if enough firms came under the control of rival parties. This threat put additional pressure on AD labor leaders to outbid these parties for worker support.
This intermediate distribution of sanctioning power meant that CTV leaders were likely to suffer serious consequences regardless of which claimants they betrayed in a loyalty dilemma. As a result, they tended to vacillate between different strategies and to consider both leadership privileges and generalized benefits for workers as part of their core interests. But as long as they could manage occasional demonstrations of loyalty toward workers, their preference was to remain in the good graces of the party. From 1984 to 1988 AD accommodated this preference by combining high-level positions for CTV leaders with benefit packages for workers and expansionary spending just before the 1988 elections. These concessions kept the CTV from moving beyond norm-based voice despite the government's brief experiment with austerity and a secular decline in the living standards of workers.
The CTV's balancing act demanded more radical action, however, after Pérez's reforms prompted violent demonstrations in the streets of Venezuela's major cities in February 1989. This unanticipated burst of opposition by the popular sectors dealt a serious blow to the CTV's legitimacy and emboldened those leaders within the CTV who favored a more combative stance toward the government. The CTV president, Juan José Delpino, concluded that "if we do not maintain conduct that is autonomous, independent, critical, rebellious toward the parties and the government, then this labor movement will escape our control." 56 Against the explicit wishes of the party and the government, the CTV engaged in an unprecedented act of norm-breaking voice by joining with rival labor confederations to organize the May 1989 general strike.
Having restored some of their lost credibility, CTV leaders quickly retreated to norm-based voice despite Pérez's refusal to abandon his reforms. 57 In part, they were resorting to their established practice of trying to appease competing claimants by vacillating between alternative strategies. But their retreat is unlikely to have been so rapid or complete had the party not offered them relief from the loyalty dilemma in 1990 and 1991. Rather than having to defy the party to demonstrate their loyalty to workers, as in May 1989, they could now do so with the party's blessing and support. Although this window of opportunity eventually closed, it proved sufficient to swing the pendulum back toward norm-based voice for the duration of the Pérez administration....
Katrina Burgess is Assistant Professor of Political Science in the Maxwell School of Citizenship and Public Affairs at Syracuse University. She has published several articles on union politics in Spain and is coeditor (with Abraham F. Lowenthal) of The California-Mexico Connection (1993). Currently she is completing a book manuscript on party-union relations in Mexico, Spain, and Venezuela and is beginning a new comparative project on the adaptive capacities of mass-based parties faced with crises of representation.
<http://muse.jhu.edu/journals/world_politics/v052/52.1burgess.html> -- Yoshie
* Calendar of Events in Columbus: <http://www.osu.edu/students/sif/calendar.html> * Anti-War Activist Resources: <http://www.osu.edu/students/sif/activist.html> * Student International Forum: <http://www.osu.edu/students/sif/> * Committee for Justice in Palestine: <http://www.osu.edu/students/CJP/>