Party Discipline in Latin America: The Role of Party Ideology

Abstract

Party discipline is an area that has received considerable attention in the party cohesion literature. While existing comparative research has predominantly focused on the institutional environment that impacts party discipline, we know less about its relationship with party-related variables. In this article, we explore the effect of party ideology on legislators’ perceptions of party discipline in Latin America. The difficulties for parliamentary groups to act cohesively in Latin America make party discipline and its study a major task. We argue that party ideology shapes the perceptions of legislators vis-à-vis their relative autonomy and the resulting ability of the party to be cohesive. To perform the analysis, we have constructed a database containing the responses of 1272 parliamentarians from 71 political parties in 17 countries across the 2011–2022 legislatures. In addition to party-related variables, we have included individual- and system-level variables. The results suggest that Latin American parliamentarians belonging to left-wing parties tend to conform to party discipline more frequently than those of the right wing.

Keywords

party discipline cohesion ideology political parties Latin America

Introduction

Party discipline to achieve cohesion has been a major focus of the literature on party unity (Bowler et al., 1999; Carey, 2009; Hazan, 2013; Kam, 2009). In the context of internal party (dis)agreement, this parliamentary attribute makes legislator’s behaviour more predictable (Van Vonno et al., 2014) and assures the effectiveness of parliamentary groups by increasing the ability of governments to successfully implement their policies (Little and Farrell, 2017; Sieberer, 2006). It may also affect a party’s image and electoral performance as electorates might penalize what they see as disorder (Maravall, 2003).

Previous work has shown that the variation in party discipline is due to institutional arrangements, such as electoral rules (Kam, 2009; Owens, 2003) and form of government (Carey, 2007), as well as party-related variables, such as the government’s or opposition’s status (Carey, 2007), leadership powers and party candidate selection rules (Depauw and Martin, 2009). In addition, socialization processes, both within parties and parliaments, have effects on the extent to which discipline is interiorized (Kam, 2009). In this line, party ideology (Close and Nuñez, 2016) and party family (Close, 2018) affect parliamentarian’s approaches to party discipline. The majority of this work draws on empirical evidence from Europe and the United States.

This article analyzes the relationship between ideological party orientations and legislators’ values regarding party discipline in Latin America. We argue that the ideology of a party shapes the perceptions of the legislators vis-à-vis their relative autonomy and the resulting ability of the party to be cohesive. There are several studies on party cohesion in the region measured through roll call voting (Carey, 2009; Morgenstern, 2004), but we know less about how party discipline operates and what legislators think about it as a way to achieve party cohesion. Latin America offers various disincentives for discipline within parliamentary groups that might also affect legislators’ opinions on the matter. It is also known that parties in Latin America sometimes have difficulties to survive and become institutionalized (Davila and Wylie, 2021; Mainwaring, 2018). At the same time, the received wisdom is that there are less disciplined parties under presidential regimes than those operating in European parliamentary environments (Cheibub, 2012 [2006]). In addition, clientelism in the region lead parties to compete in non-programmatic ways (Kitschelt et al., 2010), with its consequent impact on parliamentarians’ goals and career opportunities (Amorim Neto, 2002). All this represents a contrast to what we usually find in parties and institutional settings in Europe. The aforementioned difficulties for parliamentary groups to act cohesively in Latin America that are not normally seen in Europe will allow us to deepen our knowledge of how party discipline operates in adverse environments. Specifically, our selection of cases from Latin America enables us to assess whether, despite institutional constraints and party characteristics that a priori play against legislators’ acceptance of discipline, a party’s ideology is capable of shaping legislators’ approach to discipline.

Our study of party discipline follows an attitudinal conception of party discipline that has already been applied to the study of European cases (Close, 2018; Itzkovitch-Malka and Hazan, 2017). The large database used for this study, created from the University of Salamanca’s Parliamentary Elites of Latin America (PELA, n.d.) surveys, covers 1272 parliamentarians from 71 Latin American parties in 17 countries. The multilevel logistical analysis includes, in addition to ideology, other associated variables, such as institutional factors and legislators’ party, as well as individual variables. The results show an impact of party ideology on legislators’ approach to discipline. Some parliamentarians tend towards defending party discipline and accepting their party’s ascendancy over legislators, whereas others may reject or rebel against party discipline, claiming freedom of action and choice.

The following section provides some theoretical background on the importance of ideology to party discipline. Next, we present the data and methodology, paying particular attention to the characteristics and measurement of the dependent and explanatory variables. The fourth section sets out the results of the explanatory model. The article concludes by summarizing the role of ideology found in attitudes towards party discipline, the implications of these findings and future research avenues.

Why Ideology Is Important in Explaining Party Discipline

Studies dealing with the relationship between ideology and party discipline highlight the need for empirical approaches to understand the impact of this variable. Gauja (2013) notes that left-wing legislators, when compared with those on the right, tend to prioritize their parties’ interests. In this same line, green parties, libertarians and alternative parties are more likely to defend discipline than traditional, authoritarian, and nationalist parties (on the GAL/TAN scale) (Volpi, 2019). The cases analyzed in these studies are from Europe, meaning this relationship has not yet been examined in studies relating to Latin American legislators and party dynamics within those parliaments.

Regarding ideology, there is a wide variation in the region with social democrats, Christian democrats, liberals and conservatives representing the most common party families. In addition, parties are different in the relevance they attribute to ideological substance, which is sometimes blurred due to personalism (Kitschelt et al., 2010). At the same time, many have evolved into catch-all parties but display different forms (Cyr and Liendo, 2020), such as the minor niche parties that have recently emerged (Kernecker and Wagner, 2019). As expected, given their varying nature, Latin American parties differ in their ‘vertical interest aggregation’ abilities (Luna et al., 2021). In this rich universe of diverse ideologies and varying roles played by systems of belief within parties, we examine whether a party’s ideology has any impact on legislators’ attitudes.

Different analytical approaches to political parties and representation in parliaments help to provide a response to the question of the importance of ideology in explaining party discipline. The most common type of analysis of the impact of ideology is on legislators’ behaviour with regard to issues such as how votes are cast (Cox and McCubbins, 2005) or the gestation of some forms of inter-party support, such as parliamentary coalitions (Londregan, 2000). Alongside these issues, ideology may also affect legislators’ conceptions regarding representation and, specifically, their views of party discipline.

First, a party’s core ideology has an impact upon legislators’ values and attitudes, including on their views about their role and connections with the party. The Latin American Left tends to stress the value of collective action implemented through renewed alliances between parties and non-state actors (Anria and Chambers-Ju, 2019), at the same time as promoting solidarity as a source of identity (Arnson and Perales, 2007). In contrast, many right-wing parties emphasize the principle of individual freedom in areas such as economic policy (Aguilar Rivera, 2013; Bobbio, 1996). As a consequence of this sense of individual uniqueness, parties on the right of the spectrum have ascertained that internal diversity is electorally attractive, resulting in the ‘catch-all’ model (Middlebrook, 2000). This gives rise to a greater permissiveness of dissent in these parties, which reinforces the principle of independence among legislators. In this regard, it is plausible that the interiorization of this principle by members of right-wing parties is transferred to legislators’ conceptions regarding party discipline. Consequently, legislators from right-wing parties might find it difficult to conciliate a party’s core principles dominated by an individualistic perspective with the requirement of party coordination that leads to party unity.

Second, the relationship between parliamentarians and their parties, which frames their opinions regarding party discipline, is influenced by socialization processes. Apart from helping legislators to interiorize a party’s main values, socialization processes have an impact on how parliamentarians relate to the party and, consequently, on their self-perception of their role as representatives and their related stances on party discipline. Within parties of the left, and particularly among communist parties, socialization processes stress the primacy of the party and respect for hierarchies (González, 2014). Parliamentarians from such parties, many with a revolutionary background, would be more prepared to set aside their individual beliefs should these conflict with those of the party (Alcántara, 2008). This tendency is likely to predominate in polarized contexts and in those parties that have experienced persecution and the need to go underground, which is the experience of, among others, the region’s social organizations and communist parties (Carr and Ellner, 1993). This adverse situation, experienced by many left-wing parties in Latin America (Van Dyck, 2017), provides the basis for a sense of party unity. Regarding right-wing parties, Loxton (2021) argues that authoritarian successor parties on the right of the political spectrum that went through counter-revolutionary struggle and have an authoritarian inheritance are capable of generating unified parties. However, he shows that this is only the case for a few parties in the region. Besides, authoritarian experiences tend to have effects on opposing parties in other regions too (Levitsky and Way, 2012). Given all this, even though some parties on the right might show attachment to the idea of party discipline, it is an outcome that is more like to be found on the left.

Derived from these effects, our expectation is that different party ideologies shape attitudes towards party discipline among legislators. Parliamentarians of parties on the left would have more favourable positions towards party discipline having interiorized the value of party unity, as well as their own contribution to this. On the other hand, across parties on the right, one would expect to find a lower number of parliamentarians willing to defend party discipline because of the emphasis on individual freedom. This pattern might have been reinforced as a reaction of the right-wing elite to the shift to the left in the region (Cannon, 2016). At the same time, while it is true that more radical right-wing parties that emphasize order may, as in Europe (Close, 2018), put more stress on party loyalty than those of the moderate right, in our representative sample of the distribution of parties with parliamentary representation in Latin America, less than 10% of them are from the radical right.

Between these two extremes, it would be expected that parties occupying ideological positions close to the centre, many of them Christian democrat, would have mixed feelings towards party discipline. This would be a result of these parties’ emphasis on values such as solidarity. Nevertheless, despite the importance of this value to Christian democracy in Latin America, most of the parties belonging to this ideological family, now very blurred in terms of specific weight, promote platforms with policies that bring them closer to parties of the right (Mainwaring and Scully, 2003). In this same vein, increasing polarization (Singer, 2016) has also contributed to erode centrist positions in voters and parties. These factors may impact the conception of party discipline held by the centre, driving it closer to positions more likely to promote individual freedom.

In short, party ideology matters for party discipline: First, principles and values from a party’s ideology influence legislators’ attitudes on political representation and promote a culture that shapes their perceptions of their role and autonomy; second, socialization processes have an effect on legislators’ conceptions about how to relate to the party. The analysis below will not discuss which of these mechanisms is more relevant, but it provides a measurement of the effect of party ideology controlled for the effect of other individual-, party- and system-level factors.

Data and Methodology

The data on party discipline extracted from the PELA survey were obtained from face-to-face interviews with parliamentarians aimed at collecting information on their attitudes across a variety of issues. The number of interviewees for each party is proportional to each party’s size in the legislative chamber. Many studies have drawn on this database in their analyses (Kernecker, 2017; Lupu et al., 2017; Singer, 2016, among others). In addition to being questioned on political, social and economic issues, the parliamentarians provide responses on their conceptions around representation and party dynamics.¹ To perform the analysis, we have merged these data from 17 countries between 2011 and 2022. To this were added individual, party characteristic and country variables. The resulting database encompassed the responses from 1272 parliamentarians from 71 political parties across 17 countries (see details in Supplementary Appendix).² This sample is representative of Latin American legislatures because it includes parliamentarians with varying characteristics in terms of career history, political significance and ideological orientation.

Party Discipline as a Way to Achieve Party Cohesion

The concept of party discipline refers to mechanisms to secure party unity, generally in voting behaviour (cohesion), when this does not occur naturally: in other words, in the absence of party agreement (or party coherence). Party discipline ranges from financial penalties to threats to a parliamentarian’s current career, such as not appointing the legislator as chair or member of a particular committee within the legislature. It could also affect the legislator’s future prospects through mechanisms such as hindering possible re-election or impeding promotion to another body or institution (e.g. the Senate). Recent studies have considered party discipline to be part of legislators’ sequential decision-making process (Van Vonno et al., 2014): In case of party agreement, the legislator will naturally vote with the party. Otherwise, in a second stage, the legislator will decide out of party loyalty. The latter deals with the degree to which a legislator has interiorized the idea of following party directives, reflecting their predisposition to loyally follow the party line. When this attribute of party loyalty is lacking, legislators might consider positive and negative incentives in a third phase. It is in this phase that parties might employ coercion to ensure unity.

As Itzkovitch-Malka and Hazan (2017) note, different mechanisms intervene in party unity, such as institutional arrangements, as well as attitudes and norms of behaviour. It is these latter elements that are captured when studying how legislators view party discipline. Unlike in Europe, party discipline in Latin America has been the subject of less attention, as studies of legislative dynamics have prioritized the study of voting cohesion from a comparative perspective (Carey, 2009; Morgenstern, 2004), together with case studies on Chile (Campos and Navia, 2022), Mexico (Su and Fonseca, 2021) and Brazil (Desposato, 2005), to give some recent examples. Party agreement has also been examined (Hawkins and Morgenstern, 2010; Otero-Felipe and Ruiz Rodríguez, 2022), as well as party loyalty (Kernecker, 2017). Existing works on party discipline in Latin America give us a sense of the practices employed to maintain party unity in parliament (Alemán and Tsebelis, 2016; Carey, 2003), with attitudinal aspects of party discipline (such as how it is envisioned, defended or fought against by legislators across parties) remaining unexamined.

Our dependent variable is ‘attitudes towards party discipline’. These attitudes are related to parliamentarians’ modes of representation, which encompass not only who the legislator represents and who to consider when making decisions but also how autonomous a legislator is in fulfilling the representative role (Andeweg and Thomassen, 2005). The PELA survey includes a question on legislators’ opinions on their parliamentary group’s discipline, which will be the basis to construct our dependent variable:

‘The issue of Members’ voting discipline has traditionally been the subject of widely differing opinions. Which of the following opinions do you agree with most?: (1) Voting discipline will always be necessary in the parliamentary group. (2) Each Member should be able to vote as he or she sees fit. (3) Some issues should be subject to voting discipline and others should not’.

These response options on party discipline are based on the categorization of legislators’ roles by Eulau et al. (1959). Advocacy of party discipline is linked to a legislator’s role as a party delegate (Option 1 from PELA question). In contrast, those who advocate freedom may act as a constituency delegate or to follow their own judgement as trustees (Option 2). The response, depending on the situation, fits the characterization of legislators as politicians (Option 3).

The question has been codified on a binary basis between those in favour of party discipline and those in favour of voting based on their own criteria and on an issue-by-issue basis. This dual operationalization of party discipline allows an assessment of whether the legislator defends the need for party discipline (being a partisan defender in this case) or declines it (placing themselves as distant from the party’s leadership and arguing for their freedom to decide). This dichotomous view obviously suffers from limitations. First, we do not know whether a particular parliamentarian tends to agree with their party’s dictates, in which case, discipline would not be required. In this regard, a legislator’s degree of agreement with the party could undermine our measurement of legislator attitudes if the response options referred to the legislator’s own personal voting experience. On the contrary, the response options are posed in a normative way to see whether discipline is desirable or avoidable for each legislator (each member should [sic]; some issues should [sic]). That is how our chosen question captures those parliamentarians that, no matter what, would argue that discipline is a legitimate resource to achieve party unity. However, as an additional measure to ensure that attitudes towards discipline are not conditioned by prior consensus in the party, a variable controlling for this issue has been included in the statistical model, as discussed below. Second, the binary codification of the answers provides no information on the mechanisms entailed in a legislator’s decision. We remain unaware about whether the legislator’s decision entails consensus through debate or is imposed by the leader of the parliamentary group and/or political party.

Despite these limitations, we were able to test the main idea of our approach, which is whether parliamentarians believe in the need to adhere to party discipline for party cohesion. Table 1 shows the frequency distribution for our dependent variable ‘party discipline’. It shows that a significant proportion of legislators in the region do not see the need for party discipline.

Table 1.

Distribution of Parliamentarians Who Accept/Reject Party Discipline.

	N	%
Accept party discipline	405	31.84
Reject party discipline	867	68.16
Total	1272	100.00

Source: the authors, based on the PELA survey.

PELA: Parliamentary Elites of Latin America.

Figure 1 shows the distribution in percentages of our dependent variable. As a general trend, there are fewer parliamentarians that defend party discipline than those against it. Nevertheless, there are great differences between and within party systems, pointing to the existence of contextual factors that may lie behind these differences. In some countries, parliamentarians who accept discipline exist in all political parties. This is the case of Brazil, Nicaragua, Uruguay and Ecuador. On the other extreme, there are countries, like Guatemala and Chile, that scarcely have any prodiscipline parliamentarians or where they are concentrated in just one or two parties, as with Argentina and El Salvador. Another noteworthy feature is the higher number of pro-discipline parliamentarians in leftist parties, particularly in the cases of the FSLN (Nicaragua), MPAIS (Ecuador), FA (Uruguay) and MAS (Bolivia), of which more than half of the legislators state the belief that parties should demand discipline. Nevertheless, there are different sensitivities among the left-wing parties with respect to party discipline, which are to be expected given the coexistence of different types of left-wing formations (Webber, 2022). Thus, in contrast to the parties that we saw as more enthusiastic about party discipline, the parliamentarians of the PS (Chile) and the PRD (Panama) are opposed to the imposition of party discipline. On the other hand, among the right-wing parties, there is a tendency to reject party discipline. Examples include Chile’s RN and UDI, El Salvador’s PCN and Arena and Mexico’s PAN. However, it is true that some parliamentarians of right-wing parties also support party discipline in significant numbers, such as the PN of Uruguay, the PNH of Honduras and the ANR of Paraguay. Regarding centre parties, there is a trend towards rejecting party discipline in the same way as many rightist parties. Very few of the parliamentarians belonging to centre parties hold pro-party discipline positions.

Ideology

Ideology constitutes a supradimension, or summary measure, of political parties that tends to be measured with the categories ‘left’ and ‘right’. Regarding Latin America, some doubt has been cast on the usefulness of measuring party ideology due to personality politics and clientelism, which might mean that substantive issues are, when compared with European parties, of less importance in ideological competition (Mainwaring, 2018; Ruth-Lovell, 2016). Nevertheless, a number of authors have highlighted the importance of this dimension in inter-party competition and also in voting behaviour in Latin America (Kitschelt et al., 2010).

In our work for this article, ‘party ideology’ has been taken from a PELA question in response to which each parliamentarian situates the political party they belong to.³ From the legislators’ responses, we calculated the ‘ideological average’ of each political party. As in other cited studies using the PELA survey, instead of using expert surveys or analyzing electoral programmes, here we leverage the direct perceptions of parliamentarians as to what their party is. Our database includes Latin American parties distributed across the entire ideological spectrum from 1 (extreme left) to 10 (extreme right). Figure 2 shows the ideological distribution of the parties included in our sample. The first thing to note here is, once again, the regional heterogeneity. There are countries where the ideological distribution of parties is spread across the spectrum, as in Chile and Costa Rica, and others where there is a clustering towards the left, as in Nicaragua, Uruguay and Ecuador, where there is a left-wing force of great electoral importance. At the other extreme are those cases in which parties are clearly found on the right of the scale, as in the Dominican Republic, Paraguay and Guatemala. Figure 2 also shows that approximately one third of the parties included would be located in the ideological centre. Argentina’s UCR, Chile’s PDC, Ecuador’s CREO and Costa Rica’s PUSC occupy this ideological space in their respective party systems. It should also be noted that in some cases, the centre space is shared by more than one political party, as in Argentina, Guatemala, Panama and Peru. In other countries, on the contrary, no party is found in this space, as is clearly the case in Nicaragua and in Bolivia to a lesser extent.

Control Variables

Individual-level control variables include both those associated with a parliamentarian’s political career and experience, as well as socio-demographic factors. A legislator’s previous experience is associated with the history of socialization within the parliamentary group or in other instances, which should (in theory) foster the internalization of party discipline (Close, 2018; Ghergina and Chiru, 2014). Here, it has been operationalized with three variables. First, we have included a variable indicating if it was the first time s/he was elected as a parliamentarian; second, a variable that controls whether the legislator has previous experience as a representative (at any electoral level); third, we have controlled for the political ambition of the parliamentarian, differentiating between those who wish to continue in politics at the end of their term of office and those who do not. For all these dichotomist variables, the expectation is to find a positive relationship with acceptance of party discipline. We have also included as a control variable the difference (in absolute value) between the parliamentarian’s ideological position and that of his or her political party, expecting those closest to him or her to show a greater tendency to vote in a disciplined manner following an external authority (Kam, 2009; Sudulich et al., 2020). The two socio-demographic variables included are age and sex. Young parliamentarians are expected to be more loyal to their party leaders given their more probable belief in a long future career in which they can secure more important positions within the party or the parliamentary group (Kam, 2009). Sex has been shown to have an impact upon voting behaviour (Barnes, 2012; Norton, 1999), but its relationship with party unity has received less scrutiny. Some studies have found that women tend to exhibit higher levels of party discipline than men (Clayton and Zetterberg, 2021; Cowley and Childs, 2003).

At party level, the model controls for previous consensus inside the political party operationalized by means of the variable called ‘padisa’ from the V-Party database (Lührmann et al. 2020) for the period immediately before the election under consideration. This variable measures experts’ evaluation of the extent to which the elites of each party show disagreement on the party’s strategies (understood in broad terms, including campaign strategies, public policy positions, distribution of resources, etc.). Its theoretical range is between 0, where there is complete disagreement among elites, and 4 when party elites show virtually no visible disagreement about party strategies. We know that internal agreement influences legislators’ decision-making process (Van Vonno et al., 2014). Given this, it could be the case that consensus leads legislators to reject party discipline as an unneeded tool as unity in vote is a natural result. However, we could argue the opposite: High internal party agreement leads legislators to accept party discipline as a natural expression of having the party at the centre of the decision-making processes.

The model also controls for party size, measured as the percentage of votes secured in elections to the legislature, and party age, measured as the years between its foundation and the election under review. Both variables reflect a party’s institutionalization and durability (Cyr, 2017), which reinforce the aforementioned socializing experience for party discipline. In addition, we are controlling for party’s control over candidate nomination. The literature has highlighted the positive effect of centralized candidate selection methods on internal cohesion. Therefore, our expectation is to find a positive relationship between attitudes towards party discipline and party-dominated candidate-selection processes (Sieberer, 2006). This variable has been operationalized on the basis of the V-Party data set, in which experts have classified the degree of centralization of legislature-level candidate selection. This variable theoretically oscillates between 0, where national leaders are exclusively responsible for picking candidates, and 4, where there are primaries for selecting candidates for legislative elections. In our regional sample, the minimum is 0, and the maximum for legislative election candidate selection decentralization is 2.83. The last control included at party level is it being in government or opposition. The literature has noted the additional pressure that may be felt by parliamentarians by being in government (Carey, 2007; Stecker, 2015), and so it is foreseeable that such a relationship will be found in our data. The variable is taken from V-party, where parties are ranked according to their degree of support for the government, ranging from 0 (when the party leads the government) to 3 (if the party is in opposition).

Finally, we control for four system-level factors. First, we have incorporated two dichotomous variables related to the electoral system: one relating to the existence of preferential voting and the other relating to average district magnitude. It would be expected that where voters can mark a preference for a candidate or where district magnitude is high, the importance of discipline would diminish (Owens, 2003).⁴ We have also incorporated a control for the degree of programmatic or clientelistic ties that predominate in the system based on the Varieties of Democracy (V-DEM) classification. Although higher levels of discipline might be expected in more programmatic environments, the evidence is inconclusive (Chaves Jucca, 2022). This variable theoretically ranges from 0, for cases where voters are rewarded with material goods or jobs, to 4, where linkages or benefits are mainly programmatic. In addition, we have included a variable that refers to the multilevel structure of the state. We use the Regional Authority Index drawn up by Shair-Rosenfield et al. (2021) to measure decentralization, on the basis of an analysis of self-rule and shared rule. In Latin America, this index ranges from 0.08 (Costa Rica) to 23.5 (Argentina). As Close (2018) found for Europe, the expectation is that, by multiplying the levels of power, decentralized settings tend to increase intra-party divisions (Carey, 2007), and this could reduce the number of parliamentarians favourable to party discipline.⁵

Party Discipline and Ideology in Latin America: Results

In order to explain the impact of ideology on Latin American parliamentarians’ attitudes towards party discipline, a multilevel analysis has been performed. Hierarchical regression models are the most appropriate when dealing with predictors measured at different levels of aggregation – in this case, at individual (parliamentarian), political party and political system levels. The most important feature to take into account in these models’ specification is that the observations analyzed are not statistically independent. The cases belonging to a single category of contextual-level variables (party and political system) are subject to a common influence that creates a dependency in their behaviour. In this article, given that the dependent variable has a binary structure – whether the parliamentarian in question is or is not in favour of party discipline – we have estimated multilevel logistic regression models (Hox, 2010; Sommet and Morselli, 2017).

Table 2 shows the results of both the null and full model with the principal explanatory and control variable. The null model informs us of the mean probability of being a parliamentarian who supports party discipline for the sample as a whole, which stands at 21.21%.⁶ Also calculated is the intra-class correlation coefficient, which tests the percentage variation represented by the aggregate-level predictors (Austin and Merlo, 2017; Sommet and Morselli, 2017). In this case, this stands at 0.17 at country level and 0.16 at political party level. This informs us of the correlation between the responses of two parliamentarians belonging to the same country or political party, respectively. In other words, approximately 33% of the variation of being in favour of party discipline would be explained by contextual factors. All variables have been centred on the grand mean.

Table 2.

Results.

Fixed effects	Variance components model		Three-level model		Odds ratio
Fixed effects	Coefficient	Std. error	Coefficient	Std. error	Odds ratio
Constant	–1.312***	0.289	–0.977***	0.146	0.377
Individual level
First legislature			–0.035	0.203	0.965
Prior representative experience			0.448*	0.190	1.566
Political ambition			0.667*	0.312	1.949
Ideological distance with party			–0.008	0.089	0.992
Age			0.021*	0.009	1.021
Sex (female)			0.041	0.194	1.042
Party level
Ideology			–0.152*	0.066	0.859
Previous consensus			0.093	0.186	1.098
Size			0.038**	0.011	1.034
Party age			0.093**	0.004	1.009
Candidate selection			–0.451*	0.218	0.637
Government support			–0.188^#	0.099	0.829
System level
Preferential vote			0.325	0.316	1.384
Average district magnitude			0.044	0.073	1.045
Linkages			–0.188	0.154	0.828
RAI			–0.028	0.022	0.973
Random effects
Party level	0.785	0.303	0.113	0.164
Party system level	0.848	0.470	0.140	0.169
N (parties) (country-years)	1009 (59) (16)		905	(59) (16)
AIC/BIC	1050.82/1065.57		916.3456/1007.696

Source: own work.

RAI: Regional Authority Index; AIC: Akaike’s Information Criteria; BIC: Bayesian Information Criteria.

p < 0.1, *p < 0.05, **p < 0.01, ***p < 0.001.

The following column displays the results obtained in the full model, which support the article’s principal hypothesis on the impact of party ideology and confirm some of the literature’s findings regarding the control variables. Given that this is a multilevel logistic model, the coefficients are not directly interpretable, so, to evaluate the magnitude of the effect, the odds ratios have been calculated.

The results of the analysis confirm the proposed hypothesis. As shown by the negative coefficient, leftist Latin American parliamentarians are more likely to support party discipline than those on the right. The odds ratio for this effect is 0.859, which represents a decrease of 14.1% from the baseline probability for each extra point towards more rightist positions. Since the model’s baseline probability of accepting party discipline is 26.93% (e^−1.312), each one-point increment towards more rightist positions decreases this probability by 3.8% points. This result goes in the same direction as comparable studies among legislators of other geographic areas (Gauja, 2013).⁷

The impact of the variable can be seen in the figure below, which presents the estimated probabilities of supporting party discipline, according to the ideological positions assigned to the party (Figure 3). We can see that the lower the probability of backing party discipline, the further the party is situated towards the right of the scale. At the same time, it can be seen that the positive effect of ideology on the probability of defending discipline is more prominent in parties located further to the extreme left and decreases as they move to the ideological centre. Thus, there are some parties of the more radical left whose parliamentarians are stronger advocates of discipline than other parties of the moderate left. According to our argument, these differences are due to the emphasis generated by the parties on the value of unity, as well as the varied experiences of socialization.

Regarding the control variables, experience in representative posts reveals its importance in attitudes towards party discipline, but experience as a legislator and distance from party ideology do not. In this sense, the results suggest that upholding party discipline is not related to the degree of the parliamentarian’s ideological alignment with the party. Our expectation of a relationship between these variables was based on the impact that the degree of distance to the party has on legislators’ effective behaviour (party cohesion). While this expectation may prove true when explaining legislators’ behaviour, our findings of the impact of distance from the party on legislators’ attitudes highlight the fact that despite a legislator’s ideology being very close to their party’s ideology, if the latter lacks group identity and also fails to transmit to their parliamentarians the importance of unity, this legislator is more likely to consider that following his or her own principles is appropriate. Socialization while in representative office makes experienced legislators more pro-discipline, increasing this likelihood by 57%. Political ambition is also related to attitudes towards discipline. Those legislators who aim to continue in politics register an increase in the probability of being pro-discipline of 94.9%. The age of the parliamentarian is another individual-level control that has been found to be statistically significant: Older legislators are more likely to support party discipline. This effect goes in a different direction from that which is reported in the comparative literature. Furthermore, the sex control variable for parliamentarians does not show any impact upon the dependent variable.

Regarding party-level variables, the analysis shows that prior agreement within the party is not relevant to parliamentarians being more or less favourable to discipline. The remaining party-level variables have a statistically significant impact albeit at different intensities. The coefficients of size and age show the positive effect of these variables on the probability of being loyal to the party. In other words, members of larger and older parties are more likely to support party discipline. The odds ratio for both variables shows a small increase in the probability of being favourable to party discipline, of 3.4% and 0.9% for every additional percentage point of votes secured in elections and for every year of life of the political party, respectively. For their part, legislature candidate selection processes, whose coefficient is negative, appear to have an impact upon Latin American parliamentarians’ attitudes towards party discipline, in line with that indicated in the literature (Sieberer, 2006). In other words, the centralization of candidate selection is related with the probability of defending party discipline. Keeping all other variables constant, a shift in the scale towards less centralization in candidate recruitment processes would reduce the probability of support party discipline by 36.3%, against the mean probability of the entire sample. Another variable shown to be statistically significant is whether a parliamentarian’s party is in power or not. In this case, being a parliamentarian, whose party is either in government or who supports the government, decreases the probability of defending discipline compared with parliamentarians whose parties are in opposition, with all other variables constant.⁸

In contrast to the findings relating the control variables to the legislator and their party, none of the systemic variables come close to significance. The empirical evidence suggests that system-level variables, including those relating to institutions and linkages, do not seem to be relevant when understanding attitudes towards party discipline in Latin America. In contrast, in our model, attitudes towards discipline are more related to the traits of the legislator and their party; specifically, we have seen the decisive impact of party ideology on legislators’ acceptance of party discipline.

Conclusion

This article has studied the role that party ideology plays in shaping attitudes towards party discipline in Latin America. In our analysis, party ideology has proven decisive in understanding parliamentarians’ views towards party discipline. In other words, parliamentarians appear differentially to frame the need for party discipline to achieve party cohesion, depending upon the ideological orientation of the party to which they belong.

There are two reasons to consider our approach to party discipline as less traditional than other studies. First, our dependent variable is the degree to which legislators have interiorized the possibility of party discipline. Attitudes play a key role in party unity, and in this case, we have focused on how willing parliamentarians are to accept party dictates. Second, our dependent variable measures the impact of party ideology, whose effects are sometimes underestimated due to the effects of clientelism and personalism on parties’ ability to structure themselves around ideas. A party’s ideology not only provides a summary measure of its identity but also represents a set of values that legislators internalize and that affect their conceptions of representation. In this regard, the analysis has shown that, in Latin America, parliamentarians belonging to leftist parties tend to support party discipline more frequently than those of right-wing parties. We argue that this outcome can be explained through values and principles promoted via a party’s ideology, as well as conceptions about their role and autonomy towards their party, interiorized through socialization processes.

This article contributes, with new empirical evidence, to studies on the dynamics of legislative bodies, as well as to the literature on party organization. It shows that a party’s ideology does not only relate to how legislative votes are cast and the implementation of public policies but also has an impact upon its legislators. Given that parliamentarians belonging to right-wing parties are less prone to accept discipline, these parties and their leaders might work on promoting party consensus over policies and loyalty as a way to ensure party cohesion, especially regarding issues and laws of special importance for the party. Furthermore, our work offers new insights into our understanding of the determinants of party cohesion that affect legislative performance and the governability of political systems, as well as the image held of political parties in the Latin American region.

The importance of these topics calls for continued analysis of party unity and also of the impact of party ideology in legislative bodies. Apart from the need for a further study of the effects of some of our control variables related to parliamentarians that seem to have a role in party discipline, such as political ambition and political experience, at least two other aspects deserve attention. First, the study of related topics, such as party loyalty, may bring a deeper understanding of party unity in Latin American legislative bodies, regarding which ideology might account for differences among parties. Second, it remains unexplored how party ideology affects other parliamentary group dynamics, which are also relevant for party unity. In this sense, styles of leadership and decision-making processes might differ between right- and left-wing parties.

Supplemental Material

sj-docx-1-psw-10.1177_14789299231176699 – Supplemental material for Party Discipline in Latin America: The Role of Party Ideology

Supplemental material, sj-docx-1-psw-10.1177_14789299231176699 for Party Discipline in Latin America: The Role of Party Ideology by Patricia Otero-Felipe and Leticia M Ruiz Rodríguez in Political Studies Review

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the project “Más allá del organigrama: centros de poder en los partidos latinoamericanos” (Beyond the organization chart: Centres of power in Latin American parties) by the Ministerio de Economía y Competitividad (Spain) under Grant PID 2019-104787RB-I00.

ORCID iD

Patricia Otero Felipe

Supplemental Material

Additional supplementary information may be found with the online version of this article.

Name of political parties included in the analysis by country

Table 1A. Descriptive Statistics.

Figure A1. Effect of Party Ideology on Parliamentarians’ Support for Party Discipline Depending on Prior Political Experience (Left) and Political Ambition (Right).

Figure A2. Effect of Party Ideology on Parliamentarians’ Support for Party Discipline Depending on Government Support (Left) and Candidate Selection Methods (Right).

Appendix B. Robustness Check

Table B1. Results (Ideology Squared).

Figure B1. Effect of Party Ideology (Squared) on Parliamentarians’ Support for Party Discipline.

Table B2. Results (Ideology With Experts).

Table B3. Results (Redistribution).

Figure B2. Effect of Party Ideology (According to Experts) (Left) and Party Positions About Redistribution (According to Experts) (Right) on Parliamentarians’ Support for Party Discipline.

Notes

Author Biographies

Patricia Otero-Felipe is an Associate Professor at the University of Burgos (Spain). She holds a PhD in Political Science and a master’s degree in Latin American Politics from the University of Salamanca. She has been a visiting scholar in the Departments of Political Science at the University of North Carolina at Chapel Hill and the University of Pittsburgh. Her research interests are political parties, political behaviour and political representation from a comparative perspective. She has published in Latin American Politics and Society, Electoral Studies, Revista de Ciencia Política (Chile), Política y Gobierno, Colombia Internacional and Revista de Investigaciones Sociológicas.

Leticia M Ruiz Rodríguez is an Associate Professor at Universidad Complutense de Madrid (Spain). She is currently Academic Visitor at the Latin American Centre from University of Oxford, Doctor in Political Science by Universidad de Salamanca (Spain), Master in Comparative Politics by University of North Carolina (USA) and Diploma in Data Analysis and Collection by University of Essex (United Kingdom). Her work is focused on party dynamics and parliamentary elites from a comparative perspective. She leads a 3-year project, funded by the Spanish government, on spheres of power within Latin American political parties. Author of different books, she has also published in the Czech Journal of Political Science, Perfiles Latinoamericanos, Revista Mexicana de Sociología, Revista de Ciencia Política (Chile) and Colombia Internacional (in press), among others. She was recently awarded with the Prize for Best Congress Paper (2021) by the Asociación Española de Ciencia Política.

References

Aguilar Rivera

(2013) Latin American Political Ideologies. In: Freeden

Stears

(eds) The Oxford Handbook of Political Ideologies. Oxford: Oxford University Press, pp.583–606.

Alcántara

(ed.) (2008) Politicians and Politics in Latin America. Boulder, CO: Lynne Rienner Publishers.

Alemán

Tsebelis

(eds) (2016) Legislative Institutions and Lawmaking in Latin America. New York: Oxford University Press.

Amorim Neto

(2002) The Puzzle of Party Discipline in Brazil. Latin American Politics and Society 44 (1): 127–144.

Andeweg

Thomassen

JJA

(2005) Modes of Political Representation: Toward a New Typology. Legislative Studies Quarterly 30 (4): 507–528.

Anria

Chambers-Ju

(2019) Parties and Non-State Actors in Latin America. Oxford: Oxford Research Encyclopedia of Politics.

Arnson

Perales

(eds) (2007) The ‘New Left’ and Democratic Governance in Latin America. Washington, DC: Woodrow Wilson International Center for Scholars.

Austin

Merlo

(2017) Intermediate and Advanced Topics in Multilevel Logistic Regression Analysis. Statistics in Medicine 36 (20): 3257–3277.

Barnes

(2012) Gender and Legislative Preferences: Evidence from the Argentine Provinces. Politics & Gender 8 (4): 483–507.

10.

Bobbio

(1996) Left and Right: The Significance of a Political Distinction. Chicago, IL: University of Chicago Press.

11.

Bowler

Farrell

Katz

(eds) (1999) Party Discipline and Parliamentary Government. Columbus, OH: Ohio State University.

12.

Campos-Parra

Navia

(2022) La cohesión legislativa en la cámara más valorada: el caso del senado de Chile, 1990–2018. América Latina Hoy 90: 1–20.

13.

Cannon

(2016) The Right in Latin America. Elite Power, Hegemony and the Struggle for the State. New York and London: Routledge.

14.

Carey

J M

(2003) Discipline, Accountability, and Legislative Voting in Latin America. Comparative Politics 35 (2): 191–211.

15.

Carey

(2007) Competing Principals, Political Institutions, and Party Unity in Legislative Voting. American Journal of Political Science 51 (1): 92–107.

16.

Carey

(2009) Legislative Voting and Accountability. Cambridge: Cambridge University Press.

17.

Carr

Ellner

(eds) (1993) The Latin American Left: From the Fall of Allende to Perestroika. Boulder, CO: Westview Press.

18.

Chaves Jucca

(2022) Explaining the Puzzle of Weak Parties With Strong Legislative Leaders in Brazil: Towards a Theory of Clientelistic Party Discipline. PhD Thesis, University of Minnesota, Minneapolis, MN.

19.

Cheibub

(2012 [2006]) Presidentialism, Parliamentarism, and Democracy. Cambridge: Cambridge University Press.

20.

Clayton

Zetterberg

(2021) Gender and Party Discipline: Evidence from Africa’s Emerging Party Systems. American Political Science Review 115 (3): 869–884.

21.

(2018) Parliamentary Party Loyalty and Party Family. The Missing Link? Party Politics 24 (2): 209–219.

22.

Nuñez

(2016) At the Root of Parliamentary Party Cohesion: The Role of Intraparty Heterogeneity and Party Ideology. Cahiers du CEVIPOL/Brussels Working Papers 1: 1–29.

23.

Cowley

Childs

(2003) Too Spineless to Rebel? New Labour’s Women MPs. British Journal of Political Science 33 (3): 345–365.

24.

Cox

McCubbins

(2005) Setting the Agenda: Responsible Party Government in the U.S. House of Representatives. New York: Cambridge University Press.

25.

Cyr

(2017) The Fates of Political Parties: Institutional Crisis, Continuity, and Change in Latin America. Cambridge: Cambridge University Press.

26.

Cyr

Liendo

(2020) Party Change and Adaptation in Latin America. In: Thompson

(ed.) Oxford Research Encyclopedia of Politics. Oxford: Oxford University Press, pp.1–29.

27.

Davila

Wylie

(2021) Party Leadership and Institutionalization in Latin America. In: Thompson

(ed.) Oxford Research Encyclopedia of Politics. Oxford: Oxford University Press Available at: https://doi.org/10.1093/acrefore/9780190228637.013.1667.

28.

Depauw

Martin

(2009) Legislative Party Discipline and Cohesion in Comparative Perspective. In: Giannetti

Benoit

(eds) Intra-Party Politics and Coalition Governments. New York: Taylor & Francis, pp.103–120.

29.

Desposato

(2005) Correcting for Small Group Inflation of Roll-Call Cohesion Scores. British Journal of Political Science 35 (4): 731–744.

30.

Eulau

Wahlke

Buchanan

, et al. (1959) The Role of the Representative: Some Empirical Observations on the Theory of Edmund Burke. American Political Science Review 53 (3): 742–756.

31.

Gauja

(2013) The Politics of Party Policy: From Members to Legislators. Houndmills: Palgrave Macmillan.

32.

Gherghina

Chiru

(2014) Determinants of Legislative Voting Loyalty under Different Electoral Systems. International Political Science Review 35 (5): 523–541.

33.

González

(2014) Communism in Latin America. In: Smith

S A

(ed.) The Oxford Handbook of the History of Communism. New York: Oxford University Press, pp. 252–267.

34.

Hawkins

Morgenstern

(2010) Ideological Cohesion of Political Parties in Latin America. In Kitschelt

Hawkins

Luna

, et al. (eds) Latin American Party Systems. Cambridge: Cambridge University Press, pp.145–168.

35.

Hazan

(2013) Cohesion and Discipline in Legislatures. London: Routledge.

36.

Hox

(2010) Multilevel Analysis, Techniques and Applications. London: Routledge.

37.

Itzkovitch-Malka

Hazan

(2017) Unpacking Party Unity: The Combined Effects of Electoral Systems and Candidate Selection Methods on Legislative Attitudes and Behavioural Norms. Political Studies 65 (2): 452–474.

38.

Kam

(2009) Party Discipline and Parliamentary Politics. Cambridge: Cambridge University Press.

39.

Kernecker

(2017) Ambition as a Micro-foundation of Party Loyalty. Party Politics 23 (5): 538–548.

40.

Kernecker

Wagner

(2019) Niche Parties in Latin America. Journal of Elections, Public Opinion and Parties 29 (1): 102–124.

41.

Kitschelt

Hawkins

Luna

, et al. (eds) (2010) Latin American Party Systems. Cambridge: Cambridge University Press.

42.

Levitsky

Way

(2012) Beyond Patronage: Violent Struggle, Ruling Party Cohesion, and Authoritarian Durability. Perspectives on Politics 10 (4): 869–889.

43.

Little

Farrell

(2017) Party Organisation and Party Unity. In Scarrow

Webb

Poguntke

(eds) Organizing Political Parties: Representation, Participation, and Power. Oxford: Oxford University Press, pp.285–306.

44.

Londregan

(2000) Legislative Institutions and Ideology in Chile. New York: Cambridge University Press.

45.

Loxton

(2021) Conservative Party-building in Latin America: Authoritarian Inheritance and Counterrevolutionary Struggle. Oxford: Oxford University Press.

46.

Lührmann

Düpont

Higashijima

, et al. (2020) Varieties of Party Identity and Organisation (V-party) Dataset V1. Varieties of Democracy (V-dem) Project. Available at: https://doi.org/10.23696/vpartydsv1(accessed on 28 September 2022).

47.

Luna

Rodríguez

Rosenblatt

, et al. (2021) Political Parties, Diminished Subtypes, and Democracy. Party Politics 27 (2): 294–307.

48.

Lupu

Selios

Warner

(2017) A New Measure of Congruence: The Earth Mover’s Distance. Political Analysis 25 (1): 95–113.

49.

Mainwaring

Scully

(2003) Christian Democracy in Latin America: Electoral Competition and Regime Conflicts. Redwood City, CA: Stanford University Press.

50.

Mainwaring

(ed.) (2018) Party Systems in Latin America: Institutionalization, Decay, and Collapse. Cambridge: Cambridge University Press.

51.

Maravall

(2003) El control de los políticos. Madrid: Taurus.

52.

Middlebrook

(ed.) (2000) Conservative Parties, the Right, and Democracy in Latin America. Baltimore, MD: Johns Hopkins University Press.

53.

Morgenstern

(2004) Patterns of Legislative Politics: Roll Call Voting in the Latin America and the United States. Cambridge: Cambridge University Press.

54.

Norton

(1999) Uncovering the Dimensionality of Gender Voting in Congress. Legislative Studies Quarterly 24: 65–86.

55.

Otero-Felipe

Ruiz Rodríguez

(2022) Internal Unity and Ideological Orientation in Latin American Parties (1993–2020). Czech Journal of Political Science 1: 50–70.

56.

Owens

(2003) Part 1: Cohesion. The Journal of Legislative Studies 9 (4): 12–40.

57.

Proyecto de Élites Parlamentarias de América Latina (PELA) (n.d.) Universidad De Salamanca. Available at: http://americo.usal.es/oir/elites/bases_de_datos.htm

58.

Ruth-Lovell

(2016) Clientelism and the Utility of the Left-Right Dimension in Latin America. Latin American Politics and Society 58 (1): 72–97.

59.

Shair-Rosenfield

Schakel

Niedzwiecki

, et al. (2021) Language Difference and Regional Authority. Regional & Federal Studies 31 (1): 73–97.

60.

Sieberer

(2006) Party Unity in Parliamentary Democracies: A Comparative Analysis. The Journal of Legislative Studies 12 (2): 150–178.

61.

Singer

(2016) Elite Polarisation and the Electoral Impact of Left-Right Placements. Evidence from Latin America, 1995–2009. Latin American Research Review 51 (2): 174–194.

62.

Sommet

Morselli

(2017) Keep Calm and Learn Multilevel Logistic Modeling: A Simplified Three-Step Procedure Using Stata, R, Mplus, and SPSS. International Review of Social Psychology 30 (1): 203–218.

63.

Stecker

(2015) How Effects on Party Unity Vary Across Votes. Party Politics 21 (5): 791–802.

64.

Fonseca

(2021) Explaining the Party Unity of Governing Parties in Mexico. Revista de Ciencia Política 41 (3): 515–537.

65.

Sudulich

Trumm

Bridgewater

(2020) Parliamentary Representation: A Cross-National Study of Candidates’ Views. Parliamentary Affairs 73 (3): 522–542.

66.

Van Dyck

(2017) The Paradox of Adversity: The Contrasting Fates of Latin America’s New Left Parties. Comparative Politics 49 (2): 169–189.

67.

Van Vonno

Malka

Depauw

, et al. (2014) Agreement, Loyalty, and Discipline: A Sequential Approach to Party Unity. In: Deschouwer

Depauw

(eds) Representing the People: A Survey Among Members of Statewide and Substate Parliaments. Oxford: Oxford University Press, pp.110–136.

68.

Volpi

(2019) Ideology and Party Switching: A Comparison of 12 West European Countries. Parliamentary Affairs 72: 1–20.

69.

Webber

(2022) Rhythms of the Left in Latin America. NACLA Report on the Americas 54 (2): 155–166.

70.

Wiesehomeier

Singer

Ruth-Lovell

(2021) Political Representation, Executives, and Political Parties Survey: Data from Expert Surveys in 18 Latin American Countries, 2018–2019. PREPPS Latam V2. Available at: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/JLOYIJ.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.32 MB