Building democracy after war? Post-rebel electoral parties and the construction of stable party systems

Abstract

Since 1990, multilateral organizations have invested in the idea that multiparty political systems are an essential part of building durable peace, and that former insurgents must be converted to political actors. To what extent does this foster longer term democracy and stability? In this paper, we focus on the role of electoral inclusion for post-rebel parties in the creation of stable party systems as one feature that supports both democratization and durable peace. We find that political integration of former insurgents in electoral politics not only serves the short-term purpose of bringing such groups “inside the tent” and smoothing the transition from war to peace. It also lays a foundation for the creation of stable party systems. We test our propositions with an original dataset and offer explanatory analysis of small set of parties.

Keywords

Post-rebel electoral parties party system institutionalization post-war stability

Introduction

Since 1990, Western powers and multilateral organizations have invested in the idea that multiparty political systems are an essential part of building durable peace. A key part of this model of liberal peacebuilding is the opportunity for the political integration of armed combatants, chiefly through the transformation of armed opposition groups into electoral political parties. A considerable literature has emerged exploring the degree to which these parties have succeeded at the ballot box and how they have adapted or failed to adapt to political life (Ishiyama and Batta, 2011; Lyons, 2016; Manning and Smith, 2016, 2019; Sindre and Söderström, 2016). Much less has been written on the role of post-rebel electoral parties in party system institutionalization (hereafter PSI), considered by many scholars to be a key part of democratic consolidation (Diamond, 1997; Mainwaring and Scully, 1995).

Since the transformation of armed opposition groups to political parties is seen as the main mechanism of the liberal peacebuilding process, we want to explore whether these parties can become meaningful actors in post-war electoral politics. Do these parties contribute to the establishment of more stable political systems over the long term? We address this question by examining post-rebel parties’ contributions to the construction of institutionalized party systems. We argue that post-conflict electoral systems that include former rebel groups as electoral actors are particularly well-suited to the establishment of institutionalized party systems.

Party systems structure politics, rather than dictating its content. Why then is a stable party system important? Because stability and predictability can set the stage for the much more gradual and slow-moving processes of voter realignment and lasting change in the content of politics. First, consistent political actors set an enduring agenda for the content of politics. The presence of a consistent set of parties competing regularly in elections serves as both a marker and a driver of PSI. In an institutionalized party system, parties with clear links to a base compete regularly within a consistent set of rules that shapes the behavior of existing parties and of aspirants and newcomers.

Over time, this behavior becomes routinized. As a result, there is, “a widening range of political actors who come to assume democratic conduct … on the part of their adversaries.” (Diamond, 1997: xvii). This is part of the process of democratic consolidation, which, according to one broadly accepted definition, is “a discernible process by which the rules, institutions, and constraints of democracy come to constitute ‘the only game in town,’ the one legitimate framework for seeking and exercising political power.” (Diamond, 1997: xvi-xvii)

One key attribute of institutionalized party systems, consistency, enables these post-rebel parties to become key actors in the electoral field. Consistency occurs when a stable set of parties shows up regularly to compete in elections and to orient inter-party competition in predictable ways. A second attribute of an institutionalized party system is a stable political cleavage structure. This is linked to consistency since consistent parties shape the cleavages that structure politics in enduring ways. But cleavages are also rooted structural, historical processes. Lipset and Rokkan (1990) argued, for example, that cleavages shaped by macro-historical processes structure political competition long after the origins of that cleavage have been forgotten. Others point to the robust interaction between structural changes and the actions of rational political actors to set the master cleavage lines that structure politics (Caramani, 2004). We argue here that consistent participation by post-rebel parties has contributed to the setting of enduring master cleavages in politics, though these are not immutable.

Can strong party systems, even if they perpetuate the wartime cleavage, form a benign structural setting that dictates the types of party organizations that thrive, but not the content of their platforms? To answer these questions, we must first gain a better understanding of how post-rebel parties have shaped post-war party systems in which they compete. Thus, we focus on a subset of parties from the Post-Rebel Electoral Parties (PREP) dataset, which tracks the transformation of armed opposition groups into political parties (post-rebel parties) in all civil wars ending in 1990 or later. We examine the small but significant group of post-rebel parties (about 29% of all such parties) that have participated in all available national legislative elections over at least two decades. These long-haul post-rebel parties, we argue, have played an important part as agents of political system building, by anchoring party systems that have become institutionalized over two or three decades.

This paper contributes to the literature in several ways. First, while there is a burgeoning literature on the parties formed by insurgents after (or during) conflict, there has been relatively little research on the contributions of these parties to the construction of party systems or indeed to political processes after the first or second elections. An exploration of these contributions is essential to assessing the value of including insurgents in post-conflict politics as electoral parties. The paper also contributes to a broader debate on the viability of “democratization by elections,” in which elections are the first step, rather than the culmination, of a process of democratization (Lindberg, 2009). This sequencing for democratization efforts has become the rule since 1990, not only in post-conflict settings but also in many countries experiencing regime transition for other reasons. And it has been supported by donors and other international actors as a remedy for economic mismanagement and poor governance since the late 1980s, particularly in sub-Saharan Africa. Yet our understanding of this approach to regime transition has tended to overlook the role played by political parties, and particularly by opposition parties. Parties and electoral politics have been studied as means to other ends, as transitional agents and processes, rather than as laying down a foundation for durable politics. Finally, our research raises questions for future research regarding the conceptualization and operationalization of PSI.

The rest of this paper proceeds as follows. First, we describe the theoretical framework for our analysis, defining what we mean by “institutionalized party system” and considering its role in democratic consolidation. To support our arguments, we then present evidence from our dataset and offer descriptive details about the participation and performance of the group of “long haul” post-rebel parties. Finally, we offer some insights on the implications of our findings for thinking about PSI more broadly. We conclude by considering the implications for both democracy and long-term peace.

Theoretical framework

Mainwaring and Scully’s (1995) groundbreaking work became a widely used comparative framework to understand party systems and process of party system institutionalization in the developing world. Conceptually similar work has also been carried out by others: PSI is examined in disparate political contexts (for instance, see Riedl’s (2014) work on Africa or Ufen’s (2008) work on Southeast Asia), and PSI is inferred as the key mechanism in the causal order of relevant political phenomena (for instance, see Mainwaring and Scully’s (1995) work on quality of democracy or Mainwaring and Scully (2008) on governance and quality of policy-making in Latin America). The PSI concept is very influential and paved the way for a stimulating discussion of party systems and electoral politics. But as far as we know, no previous research has investigated the role of PSI within the post-war electoral politics context.

What makes for an institutionalized party system? Mainwaring (2018) argues that party systems become institutionalized when they can generate a sense of future predictability. This occurs when a group of parties achieves stability through regular interaction. Scholars offer three characteristics that mark an institutionalized party system: (1) the main parties continue to be key contenders in elections over time, (2) the vote shares of these parties do not fluctuate gravely and remain stable from one election to another, and (3) these parties contend around a stable set of political cleavages (Mainwaring, 2018; Lipset and Rokkan, 1990).

In this paper, we focus on a single dimension that we suggest is a necessary component of PSI: consistency over time. For our measure of consistency, we combine consistent participation and stable vote shares over time. As we will detail below, post-rebel parties are remarkably regular participants in electoral politics, a sizeable share of them have participated in all post-conflict national legislative elections, and their vote shares are also fairly stable over time. Moreover, post-rebel parties often embody important political and social cleavages that enabled them to mobilize support in wartime. If a post-rebel party’s wartime “brand,” and the master cleavage it supports, carry over into post-war politics, it often translates into greater electoral success for the party. Although closely related, the establishment of stable cleavage structures is not examined in this paper, as a full examination of the political role of wartime cleavages deserves a separate paper and cannot be satisfactorily treated here. Instead, we focus on the first two attributes of an institutionalized party system, while acknowledging the importance of the party system’s cleavage structure and the role of post-rebel parties in sustaining it.

Consistency

Certainly, post-rebel parties have been strikingly consistent participants in electoral politics. The opportunity for armed opposition groups to form political parties at war’s end and to become legitimate political actors was a core part of the liberal peace paradigm, which held that durable peace depends in part on creating channels for peaceful competition and broad political inclusion. Most former rebel groups – 55% – seized on the chance to form parties and compete in the first post-war election (Manning and Smith, 2016). Manning and Smith (2019) find that post-rebel parties have staying power, with the majority (65%) competing in all available post-war legislative elections.

Post-rebel parties, as compared to new opposition parties in many nascent electoral systems, tend to have high name and “brand” recognition, an identifiable base of voters, and the organizational capacity to fight an insurgency. While organizational inheritances of these parties vary, many new opposition parties that were not formal rebel groups lack any organizational capacity at all, amounting to little more than a well-known personality and an attendant “clique of notables.” Thus, post-rebel parties enjoy some initial advantages that are not common for opposition parties in new democracies. This suggests that these parties have developed a stake in electoral politics that keeps them coming back again and again. Rustow (1970) argues for example that consistent participation in electoral politics can generate self-reinforcing dynamics that favor parties most willing to adapt to the new rules of the game. The more that parties engage in tactical or strategic adaptations to compete more effectively, the more invested they become in electoral politics and the more likely they are to enforce the system for which they have adjusted their organizational routines.

Post-rebel parties as consistent electoral actors

Evidence for this paper is drawn from the PREP dataset (1990-2021) (Manning et al., n.d).¹ In this section, we pay particular attention to a group of parties we call “long-haul post-rebel parties” for two reasons. First, PSI is a gradual process happening over several electoral cycles. Although inclusion of a post-rebel party is a significant step towards democratizing and long-lasting peace, transition from armed opposition groups to a full-functioning political organization takes time. The impact of these parties on their party systems depends on numerous factors: institutional and contextual factors can hamper or improve the disarmament, demobilization and reintegration process, or post-rebel parties may not be equipped with a political party toolbox that would allow them to compete in elections effectively and successfully.

Table 1 provides an overview of the performance of post-rebel parties in our dataset in national legislative elections. There were 321 national legislative elections in which the parties in our dataset could have participated. Looking at parties rather than elections as the unit of observation, we can observe distinct patterns of performance. We find that a post-rebel party failed to contest an available election in 24% of election-year observations. This includes skipped elections where parties participated in the contests before and after one that they skipped, and right-censored cases, where parties stopped contesting elections.

Table 1.

Performance over time for all post-rebel parties.

Level of success	Count	Percentage
Did not contest election	77	24
Contested, did not win seats	7	2
Won at least one seat but less than 10%	105	33
Won 10–30%	55	17
Won 30–60%	53	17
Won over 60%	24	7
Total	321	100

In only 2% of elections did post-rebel parties contest without winning any seats. In these cases, either the party failed to offer a compelling campaign prior to the elections, or the electoral process was marred by the lack of meaningful competition where other parties dominate the electoral politics. For instance, in 1997 Liberian legislative election, National Democratic Party of Liberia (NDPL) does not win any seats and only receives 1.3% of the votes against Charles Taylor’s National Patriotic Party which won more than 75% of the votes.

In 33% of elections, these parties won at least one seat, but no more than 10% of total seats, while in 7% of elections, parties won more than 60% of seats. The rest of the time (in 17% of elections), post-rebel parties won between 30 and 60% of national legislative seats. Lastly, in 7% of all elections, the post-rebel party’s vote share was more than 60% which suggests that the party was able to eliminate any meaningful competition in the electoral politics.

While the majority of parties that compete in the first election continue to win consistent levels of representation in every available election thereafter, not all parties do so. A handful of parties win seats in the first election, but then fade to zero or very low levels of representation after a few electoral cycles. There are 24 parties in our dataset that only participated in one election and then faded away. Nine of these parties (37.5) won less than 10% of the votes in this first election. Such cases include parties formed by weaker rebel groups, breakaway parties from the main post-rebel party, and parties that must compete against a large number of other post-rebel parties. For instance, ALCOP (Liberia) was formed out of the weaker of the two main rebel factions in the 1990–1995 conflict episode. PDD (Mozambique) was an early secondary party formed out of the main opposition group, Renamo. Three Iraqi parties formed out of armed insurgent groups – PUK, DPK, and ISCI/SCIRI, got off to a strong start. Of these, ISCI/SCIRI had the advantage of a prior political existence and outside support from Iran, but like the others, it fell victim both to Iraq’s fragmented party system and to the resurgence of violence shortly after the first two electoral cycles.

Finally, of the 77 post-rebel parties as our unit of analysis, we find that around 30% of these parties have participated in fewer than two elections or never participated in a single election (see Figure 1). Yet the overwhelming majority (70%) of post-rebel parties have been active players in electoral politics, not the transitional, temporary fix that might have been expected given the circumstances of their entry into politics. And around 15% of post-rebel parties in the dataset have participated in more than seven election cycles and showing impressive commitment and continuance.

Figure 1.

Election participation of post-rebel parties (N = 77).

Does consistency improve party system institutionalization?

So far, we have argued that the crux of the post-war PSI is post-rebel parties’ consistency in elections. How do these parties, by becoming consistent electoral actors, contribute to the institutionalization of the post-war electoral politics? As explained in the previous section, we use consistency for a sort of first, keyhole view of PSI among our parties. Our data is composed of disparate political actors in disparate political contexts. Parties that have faded away or parties that have minimal success and presence in electoral politics are less likely to be influential party organizations. Particularly, we believe that post-rebel parties that have been present in electoral politics for a substantial amount of time have engaged in considerable and consistent participation and are thus able to influence institutionalization of the party system. So, we propose:

H₁: If post-rebel electoral parties are more consistent in their participation in legislative elections, institutionalization of the party system is more likely. We expect that the effect of consistency on institutionalization is likely to be seen if these parties have been present in electoral politics for more than 20 years.

Additionally, we think the existing structural factors or the quality of democracy where these parties operate matters. Without free and fair elections, a government accountable to its citizens, and guaranteed civil and political liberties, we cannot expect these parties or the party system itself to be institutionalized. Institutionalization of the party system is more likely to happen for post-rebel parties operating in countries where the quality of democracy is high.

H₂: Post-rebel parties operating in more democratic regimes are more likely to lead to institutionalized party systems.

In this paper we bring together two different ways of thinking about PSI. One has to do with the behavior of the parties in a system – how consistently do they participate in elections, how well do they perform, how do they choose to mobilize followers. Mainwaring’s (2018) conceptualization of PSI follows this approach. In the conceptualization of PSI that informs V-Dem, the focus is on the organizational attributes of parties that make up institutionalized party systems. Though V-Dem’s PI stands for party institutionalization (and not party system institutionalization), as we discuss below it has been advanced also as a measure of party system institutionalization. We then examined the relationship between consistency and PSI as measured in V-Dem. Thus, is there a strong correlation between parties that show high consistency and those that rank highly on V-Dem’s PI variable? If so, our long-haul parties show both organizational and behavioral traits consistent with parties that anchor institutionalized party systems.

To examine the relationship between consistency and PSI, we employ pooled ordinary least squares (OLS) with robust-clustered standard errors to regress our independent variables on the change in PSI.² This analysis includes all legislative elections from 1990 to 2021, and our unit of analysis is election-year. These hypotheses are tested in separate models using scores from Varieties of Democracy (V-Dem) dataset with PREP dataset (Coppedge et al., 2022b; Manning et al., n d.; Pemstein, et al., 2022). We also lag our dependent variable, the institutionalization measure, at 5-year interval since we consider institutionalization as a slow-moving variable. Adopting lagged dependent variables allows us to examine the delayed effect of consistency on institutionalization – the effects of our main covariates may be delayed or otherwise manifest itself after a long period of time. To measure the consistency in the long run for H₁, we use a subset of our dataset to include only those parties that have participated in legislative elections for at least 20 years.

We have chosen the 20-year time period for several reasons. Given that most countries hold national elections every four or five years, this allows us to capture four to five elections. Mainwaring and other scholars stipulate at least three electoral cycles before assessing PSI. Parties adapt to electoral politics over time, as each election requires them to make choices about whether and how much to invest in electoral politics. Only over time can we identify parties that consistently choose to participate. We aimed for the longest period of participation that we could while maintaining a reasonable number of cases. Using a 30-year time frame for continuous electoral participation reduces the sample size prohibitively. Thus, we opt for a 20-year threshold.

Measuring PSI and consistency

We use V-Dem’s party institutionalization (v2xps_party) variable to measure our dependent variable. While party institutionalization (PI) is a complex concept, V-Dem’s index includes various attributes of the political parties in a country.³ Many analysts use electoral volatility as an indicator of institutionalized party systems.⁴ In this research, we opt for an additive index created by V-Dem, for several reasons. First, the Pedersen’s index of volatility of vote shares is not available for post-rebel parties. We can manually code all elections and calculate electoral volatility for these parties, but this is a time-consuming exercise that might have limited payoff given the second limitation. Another limitation is that the Pedersen index might be an inadequate proxy for PSI since most of the countries in our dataset comes from regimes with limited electoral politics experience.

Although V-Dem’s PI variable aims to capture the extent to which political parties, rather than party systems, are institutionalized, the variables that go into this index of PI are widely considered to be theoretically linked to broadly accepted conceptualizations of PSI. The index captures features of party organization and cohesion (level and depth of party organization, cadres, coherence of platforms and ideologies, party-line voting), as well as social legitimacy and social rootedness of parties (party supporters in the electorate, links to civil society).

The V-Dem index is based on a two-dimensional concept of PI: “the degree to which party organizations are ‘routinized’” and “the extent to which voters and party elites value the party label and program.” (Bizzarro et al., 2020: 2). They define party routinization as “the extent to which parties constitute stable, permanent, autonomous institutions” and the second dimension aims to capture “value infusion” or the intrinsic value that party elites and voters assign to the party itself (Bizzarro et al., 2020: 4). V-Dem suggests that PI implies party system institutionalization. The authors of the index note that a high score on the attributes of PI by the major parties in the system indicate more institutionalized party systems. In the V-Dem measure, parties are the building blocks of the system, so PSI requires that the attributes of the larger parties, “those that may be said to dominate and define the system” (according to V-Dem variable definition), be considered in the index. Given this requirement, the authors feel confident asserting that high scores in PI are likely indicative of an institutionalized party system.

Ultimately, the level of party institutionalization contributes to the PSI level: scholars consider PI as the internal mechanism or the causal underpinning of PSI (Hicken and Kuhonta, 2015; Mainwaring, 2018). The PI index is defined as, “the extent to which parties build stable organizations (routinization), party followers develop lasting connections with the parties, and come to prioritize party interests in addition to their individual short-term interests (value infusion)” (Bizzarro et al. 2020: 4) The authors note that the measure is an indication of how closely a party approximates an ideal type, rather than a measure of institutionalization as a process. While we are measuring observed behavior, the V-Dem PSI variable measures organizational structure. If our measure of consistency correlates with this measure of PI, it suggests that post-rebel parties that participate consistently in legislative elections also approximate this ideal type of an institutionalized party. The authors of V-Dem’s index argue that this measure can also be seen as an indication of party system institutionalization, asserting that, “a high score on these attributes generally indicates a more institutionalized party system” (Coppedge et al., 2022a: 315).

Our main independent variable, consistency, ranges from 0 to 1. Consistency is the ratio of the number of elections in which a party has participated to the total number of elections that this party could have participated in up to and including that election year.⁵ Thus, this measure allows us to see how consistent these parties are over time. We measure the level of democracy using V-Dem’s Electoral Democracy Index (v2x_polyarchy) which measures the extent to which the ideal of electoral democracy is achieved in a country. Given that the choice of democracy measurement significantly impacts the results of research findings (Coppedge, 2012), we use different measurements as a robustness check. Lastly, we add regional controls to account for potential neighborhood effects. We also control for pre-war parties, the cases in which the post-rebel party existed as a political organization before the first episode of conflict, since presence of these organizations might suggest developed electoral system and institutionalization.

Analysis

Our analysis shows that consistent participation by post-rebel parties in legislative elections can improve post-war PSI, but only if they participate regularly over a long period of time. The effect of consistency is present if these parties have been participating in electoral politics for at least 20 years. Importantly, this effect continues to be present if we lag the dependent variable to account for its possible delayed effect. These results suggest that consistency in post-war politics matters, and commitment to electoral politics has a slow but important impact on the institutionalization of the system.⁶

Our second hypothesis posited a positive relationship between the level of democracy and PSI. We find clear evidence in support of this hypothesis. As the level of democracy increases it positively affects the party institutionalization within a country. We find consistent evidence for this hypothesis throughout our models.

Table 2 shows the results of our testing for these hypotheses.⁷ We tested four different specifications of the model using alternative dependent variables (current year, lagged 5). The first two columns test the relationship between party participation and party institutionalization for all parties in our dataset. The last two columns include only those parties that have participated in legislative elections for 20 years or more. Taken together, these results show support for hypothesis 1 and 2.

Table 2.

Pooled OLS regressions of party system institutionalization and consistency.

	Dependent variable: Party system institutionalization
	PSI	PSI lagged 5	PSI	PSI lagged 5
Consistency	0.027 (0.048)	0.000 (0.046)	0.231** (0.055)	0.172** (0.066)
Democracy	0.353*** (0.135)	0.375*** (0.121)	0.305*** (0.103)	0.349*** (0.103)
Pre-war party	0.143***	0.171***	0.354***	0.358***
Pre-war party	(0.051)	(0.049)	(0.043)	(0.038)
Region controls	Yes	Yes	Yes	Yes
Constant	0.342*** (0.086)	0.300*** (0.082)	0.546*** (0.042)	0.482*** (0.042)
R-squared	0.486	0.504	0.826	0.775
N	293	288	125	125
p	0.000	0.000	0.000	0.000

Note. PSI: party system institutionalization.

*p < 0.1; **p < .05; ***p < .01, two-tailed test. Clustered robust standard errors in parentheses.

Additionally, as a robustness check, we measured democracy differently because how we measure democracy significantly impacts the results of research findings. When we use other democracy measurements, our results do not change. We find that the level of democracy measures by Polity IV and Freedom House’s (FH) Civil Liberties measurements support our findings.⁸

A closer look: Consistent, long-haul post-rebel parties

The correlation between consistency of participation (regardless of electoral performance), and PI for post-rebel parties invites further investigation. In this section, we explore these parties further. We show that these parties contribute to stabilization of their party systems by simply participating consistently over multiple electoral cycles. There are 23 post-rebel parties that have an electoral presence of at least 20 years. To have a more nuanced understanding of these long-haul parties we adopt two scope conditions: (1) an electoral presence of at least 20 years, and (2) participation in at least five national legislative elections. PSI by definition plays out over multiple electoral periods. Mainwaring and Torcal (2006) call for an observation period of three elections. We opt for five elections, which allows us to observe party system development over at least two decades.

Fifteen parties – two-thirds of the 23 parties in our dataset – meet our criteria⁹: 13 of these have also participated in every available national legislative election over that period. Table 3 lists all 15 parties. In this list, we see that a surprisingly high number of parties with very different backgrounds and levels of inherited organizational capacity have participated in all available elections across quite different political contexts. All told, post-rebel parties are a very consistent group.

Table 3.

Parties with participation over 20 years or more and at least five or more elections.

	Country	Party name	Number of elections participated
1	Azerbaijan	ARF	5^a
2	Bosnia and Herzegovina	HDZ BiH	8
3	Bosnia and Herzegovina	SDS	8
4	Cambodia	FUNCINPEC	6
5	Croatia	HDZ Croatia	9
6	Djibouti	FRUD	5
7	El Salvador	FMLN	10
8	Ethiopia	EPRDF	6
9	Guatemala	URNG	5^a
10	Kosovo	PDK	8
11	Liberia	NPP	5
12	Mozambique	Renamo	6
13	South Africa	ANC	6
14	Sri Lanka	JVP	7
15	United Kingdom	Sinn Fein	6

^aFor both parties, there were six elections available to participate, but each missed one election.

Interestingly, these parties have little in common with one another. First, they are not all strong electoral performers. Within this group, we can identify three broad levels of performance. First are parties whose average legislative seat share is greater than zero but less than 10%. These we call persistent parties given that they are not discouraged by poor electoral performance, but remain active electoral players. These parties win representation but lack the power to be decisive except as minority members of a larger legislative bloc. The second group consists of parties we label competitive. These parties have an average performance level of between 10% and 50% of national legislative seat share. The third group – dominants – tends to crowd out their competitors to the point of eliminating challenges. Dominant parties win, on average across all elections, more than 50% of legislative seats.

Table 4 shows post-rebel parties that fall into each of these performance categories. It is interesting to note that all of the dominant parties in this long-haul set are African parties (though not all African parties in this group are dominant, for example NPP in Liberia). But it is also important to point out that as dominant parties EPRDF, FRUD, and ANC are quite different from one another in terms of their origins, the character of their links to an electoral base and their internal organization. They also differ in terms of the competition they face. ANC competes against other parties with considerable electoral experience. EPRDF was able to craft an electoral environment to suit itself, while FRUD confronts no competing parties with any significant electoral experience or other party experience.

Table 4.

Persistent, competitive, and dominant parties.

Performance group	Parties
Persistent (won less than 10% of votes on average, across all elections)	(1) JVP, (2) URNG, (3) HDZ BiH
Competitive (won between 10% and 50% of votes on average, across all elections)	(4) ARF, (5) FUNCINPEC, (6) NPP, (7) SDS, (8) Sinn Fein, (9) PDK, (10) Renamo, (11) FMLN, (12) HDZ Croatia
Dominant (won more than 50% of votes on average, across all elections)	(13) EPRDF, (14) FRUD, (15) ANC

Next, they don’t share a history of having been electoral parties or even political parties of any kind prior to the post-war period. For example, four of the parties in this group have disparate values on many of the variables one would think might explain electoral performance. Manning and Smith (2019) find that parties with pre-war experience are likely to perform better than those without. Neither Renamo nor PDK were pre-war parties. Sinn Fein was a prewar party, though not a prewar electoral party. Some of the politico-military groups that made up the FMLN in wartime had previously been political parties, but like Sinn Fein, they had not participated in elections.

There is also considerable variety in the electoral contexts in which they participate. Previous research suggests that proportionality of the electoral system has a slightly positive impact on vote share for parties at the low end of the electoral performance spectrum, but is not significant otherwise (Manning and Smith, 2019). PDK, the Bosnian and Croatian parties, the ANC and Sinn Fein operate in parliamentary systems, while FMLN, Renamo, EPRDF, FRUD, URNG and NPP compete in presidential systems.

Thus, among these parties that have competed consistently in elections for 20 years or more, there is considerable variation in the factors that appear to affect electoral performance in the first or second electoral cycle. Additionally, neither organizational nor environmental characteristics line up systematically with these parties’ electoral performance over the long term as seen in the following Table 5. In this table, we look at different characteristics of these parties, the polity that they operate, and the rebel group that precedes the party. These categories (country’s mode FH status after PREP actor became a political party and electoral system, whether the party is a pre-war electoral party, ideology and the goal of the rebel group which precedes the post-rebel party, and whether there was any peacekeeping operation that ended the civil war) further support our conclusion that these parties come from disparate political contexts and background.

Table 5.

Some organizational and environmental characteristics of post-rebel parties.

	Country	Party name	FH category	Electoral system	Pre-war electoral party	Rebel group ideology	Rebel group goal	Peace keeping operation
1	Azerbaijan	ARF	Not free	Majority	Yes	Nationalist	Independence for Karabakh	No
2	Bosnia and Herzegovina	HDZ BiH	Partly free	PR	Yes	Right leaning	Self-defense	No
3	Bosnia and Herzegovina	SDS	Partly free	PR	Yes	Nationalist	Independence for Bosnian Serbs	No
4	Cambodia	FUNCINPEC	Not free	PR	No	Nationalist	Regime change and anti-occupation	Yes
5	Croatia	HDZ Croatia	Free	PR	Yes	Right leaning	Independence for Croatia	No
6	Djibouti	FRUD	Not free	Mixed	Yes	Left leaning	Overthrow of the government	No
7	El Salvador	FMLN	Free	PR	No	Left leaning	Ouster of incumbent regime	No
8	Ethiopia	EPRDF	Partly free	Majority	No	Right leaning	Establish democracy	No
9	Guatemala	URNG	Partly free	PR	No	Marxist, other communist	Ouster of incumbent regime	No
10	Kosovo	PDK	Partly free	PR	No	Nationalist	Independence for Kosovo	Yes
11	Liberia	NPP	Partly free	Majority	No	Unclear/vague	Overthrow Doe regime	Yes
12	Mozambique	Renamo	Partly free	PR	No	Nationalist	Form a government of unity through democratic elections	Yes
13	South Africa	ANC	Free	PR	No	Left leaning	End Apartheid; establish racial equality	No
14	Sri Lanka	JVP	Partly free	PR	No	Marxist, other communist	Socialist revolution, pro-Sinhalese rights	No
15	United Kingdom	Sinn Fein	Free	PR	No	Nationalist	Independence for Northern Ireland	No

Note. Information about the goal of the rebel group comes from FORGE dataset (Braithwaite and Cunningham, 2020).

Indeed, the diversity of this group suggests that it is something about electoral participation, rather than anything else that has led these parties to survive. In other words, institutionalized parties (which are necessary anchors for institutionalized party systems) are not those that have a longer history as a pre-war party, or parties that have external funding, or even parties that perform well in elections.

A second question pertains to the relationship between party institutionalization and party system institutionalization. Hicken and Kuhonta (2015), Mainwaring (2018) and others argue that PI is integral to PSI. Of course, much work remains to be done on the theoretical relationship between party institutionalization and party system institutionalization.

Our analysis suggests that post-rebel parties that have participated in electoral politics over the long haul are indeed contributing to PSI. Moreover, we find that it is consistent participation in elections, regardless of electoral performance, that is important. Post-rebel parties in this group of long-term players are not always high performers in terms of electoral results – they certainly do not always dominate the party systems of which they are a part. But they arguably do define them.

Our results admittedly rest on a small group of parties, so any conclusions must remain modest. But since our data suggest that a majority of post-rebel parties that take part in the first election go on to participate in every election thereafter, we may safely infer that this group of long-haul parties will grow in size.

We must also be cautious regarding the broader significance of PSI. Institutionalized party systems are not necessarily more democratic. While our analysis confirms that democracy scores have a positive relationship with consistency, party institutionalization, PSI, can certainly be high in authoritarian regimes as well (Bizzarro et al., 2020). Democracy scores naturally would be expected to increase as more elections are held – consistent elections boost democracy scores. But perhaps what see reflected in our findings is that elections are more likely to be held consistently over time when post-rebel parties participate.

This brings us to another caveat that is linked to the role of political cleavages in PSI. Institutionalized party systems are defined by parties with strong roots in society and clear links to an electoral base. In other words, they operate in systems with clear and stable political cleavage lines, which they also help to define and sustain. Stable political cleavage structures stabilize expectations for political actors, establish the terms on which new entrants are likely to succeed, and entrenches the advantages to parties that are already established, regular participants. These cleavages, in turn, continue to shape politics long after the initial divisions are no longer relevant. Party systems give rise to families of parties that use mobilizing appeals that resonate with voters on one side or the other of this master cleavage. This cleavage helps new and existing parties figure out how to pitch their appeals to draw voters toward themselves and away from other parties. While Lipset and Rokkan’s (1990) strongest claims about “frozen cleavages” have sparked much debate, the essential ideas set out in their seminal piece remain relevant. Party systems are structured by macro-historical processes, often conflictual ones, and the cleavages formed in the process have enduring effects on the kinds of parties that emerge and succeed.¹⁰

Where electoral politics has emerged from a protracted struggle between organized, contending forces, identifiable cleavages are more likely to occur. Levitsky and Way (2013) argue that parties forged out of military organizations form hard boundaries that prevent defection. Lebas (2011: 16) finds in her study of parties in southern Africa that conflict is an important ingredient in the formation of strong parties: “Just as intractable conflict may be the only route to democracy, mobilization that intensifies conflict and erects sharp social boundaries may be the only means of effective party-building in hybrid regimes.”

Countries that have experienced civil wars are likely to have established cleavage lines born of the conflict, which may carry over into post-conflict politics. The wartime cleavage is the divide that defined the two (or more) sides during the conflict, the cleavage that enabled belligerents (rebels and incumbents alike) to mobilize support during the war, helped shape belligerents’ organizational competencies and ideational capital, and influences the degree to which the party must rely on international financial support or moral authority. War creates cleavages that have staying power, and they can carry over into politics. And the armed actors that become peacetime parties capitalize on them, because they already have identities that fit with these wartime cleavages. These parties often rely on the carry-over of their wartime identity of champions of a particular cause as much as possible to generate electoral support.

If wartime cleavages continue to structure politics and become part of the PSI, they could sow the seeds of future conflict. Our data suggest that so far this risk has been minimal among long-haul parties. Only two parties returned to conflict after competing in two or more electoral cycles: Yemen’s Socialist Party and Renamo. In Yemen, the party returned to conflict after the third election. However, this occurred within the context of the outbreak of a wider set of conflicts over control of the country after 2011, which prevented additional elections. In Mozambique, Renamo and the ruling Frelimo party returned to conflict after four elections, but both Renamo and Frelimo took pains to arrange ceasefires to allow national elections to go forward with Renamo’s participation in 2014 and 2019 (Manning et al., n.d.).

Conclusion

This article has argued that post-rebel parties that have participated in electoral politics for two decades or more have contributed to building stable post-war party systems. We argued that party system stability in the post-war context is achieved through two mechanisms: consistency and stable cleavage structures. The consistency with which such disparate parties in very different contexts have engaged with electoral politics over decades is striking. Post-rebel parties are certainly contributing to the creation of systems of regular competition among political actors that have endured over time.

A closer examination of the other cases in our dataset reveals interesting insights. For instance, having a stable party system does not necessarily mean stable or effective governance. In Northern Ireland, the major parties have consistently failed to form an executive, more than once occasioning Westminster’s re-imposition of direct rule. Consistent accusations of corruption and fraud plague party politics in Kosovo and El Salvador. Nor is having a stable, even institutionalized party system always positively related to democratic quality. In Mozambique, a strong and stable two-party system persists even as the quality of elections has steadily declined. Even a two-year return to armed conflict after nearly two decades of peace did not significantly disrupt the rhythm of electoral politics there. Even a two-year return to armed conflict after nearly two decades of peace did not significantly disrupt the rhythm of electoral politics there.

This article pioneers in understanding the link between the post-rebel parties’ participation and party system stability. Nevertheless, more could be done to engage fully on both the empirical evidence so briefly sketched out here, as well as with the theoretical literature. How do consistency and cleavages play a role in other disparate political contexts? What kind of mechanisms enable these parties to establish stable cleavages? Why do some parties heavily rely on wartime legacies to build a broad base of voters while others let go of the past? Since these questions remain unaddressed, studies of post-war politics and political parties should anticipate and address the unique contribution of these parties in future studies.

Supplemental Material

Supplemental Material - Building democracy after war? Post-rebel electoral parties and the construction of stable party systems

Supplementary Material for Building democracy after war? Post-rebel electoral parties and the construction of stable party systems by Ozlem Tuncel and Carrie Manning in Party Politics.

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) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Ozlem Tuncel

Carrie Manning

Supplemental Material

Supplemental material for this article is available online.

Notes

Author biographies

Ozlem Tuncel is a doctoral candidate in the Political Science Department at Georgia State University in Atlanta, Georgia. Her research addresses competitive autocratic regimes, political parties, and post-rebel parties.

Carrie Manning is Professor of Political Science at Georgia State University in Atlanta, Georgia. She is the author of three books and numerous articles on post-conflict politics and democratization. Her works include Costly Democracy: Peacebuilding and Democratization after Civil War (Stanford 2013), jointly edited with Christoph Zuercher et al.; The Making of Democrats: Elections and Party Development in Postwar Bosnia, El Salvador, and Mozambique (MacMillan 2008); and The Politics of Peace in Mozambique: Post-Conflict Democratization, 1992-2000 (Praeger 2002).

References

Bizzarro

Hicken

Self

(2020) The V-Dem Party Institutionalization Index: A New Global Indicator (1900-2018). Available at: https://www.darinself.com/assets/files/party_institutionalization_index_2020.pdf (accessed 4 April 2022).

Braithwaite

Cunningham

(2020) When organizations rebel: introducing the foundations of rebel group emergence (FORGE) dataset. International Studies Quarterly 64(1): 183–193.

Caramani

(2004) The Nationalization of Politics: The Formation of National Electorates and Party Systems in Western Europe. Cambridge, UK: Cambridge University Press.

Coppedge

(2012) Democratization and Research Methods. Cambridge, UK: Cambridge University Press.

Coppedge

Gerring

Knutsen

, et al. (2022a) V-Dem Codebook v12. Varieties of Democracy (V-Dem) Project.

Coppedge

Gerring

Lindberg

, et al. (2022b) V-Dem [Country-Year/Country-Date] Dataset v12. Varieties of Democracy (V-Dem) Project. DOI: 10.23696/vdemds22

Diamond

(1997) Introduction: in search of consolidation. In: Diamond

Plattner

Chu

et al. (eds) Consolidating the Third Wave Democracies: Regional Challenges. Baltimore, MD: Johns Hopkins University Press, xiii–xlvii.

Hicken

Kuhonta

(eds) (2015). Party System Institutionalization in Asia: Democracies, Autocracies, and the Shadows of the Past. Cambridge, UK: Cambridge University Press.

Ishiyama

Batta

(2011) Swords into ploughshares: the organizational transformation of armed opposition groups into political parties. Communist and Post-Communist Studies 44: 369–379.

10.

Lebas

(2011) From Protest to Parties: Party-Building and Democratization in Africa. Oxford, UK: Oxford University Press.

11.

Levitsky

Way

(2013) The durability of revolutionary regimes. Journal of Democracy 24(3): 5–17.

12.

Lindberg

(2009) Democratization by elections? A mixed record. Journal of Democracy 20(3): 86–92.

13.

Lipset

Rokkan

(1990) Cleavage structures, party systems, and voter alignments. In: Mair

(eds) The West European Party System. Oxford, UK: Oxford University Press, 91–138.

14.

Luna

(2014) Party system institutionalization: do we need a new concept? Studies in Comparative International Development. 49(4): 403–425.

15.

Lyons

(2016) From victorious rebels to strong authoritarian parties: prospects for post-war democratization. Democratization 23(6): 1026–1041.

16.

Mainwaring

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

17.

Mainwaring

Scully

(eds) (1995) Building Democratic Institutions: Party Systems in Latin America. Stanford, CA: Stanford University Press.

18.

Mainwaring

Scully

(2008) Latin America: eight lessons for governance. Journal of Democracy 19(3): 113–127.

19.

Mainwaring

Torcal

(2006) Party system institutionalization and party system theory after the third wave of democratization. In: Katz

Crotty

(eds) Handbook of Party Politics. Thousand Oaks, CA: Sage, 204–227.

20.

Mair

(1997) Party System Change Approaches and Interpretations. Oxford, UK: Oxford University Press.

21.

Manning

Smith

(2016) Political party formation by former armed opposition groups after civil war. Democratization 23(6): 972–989.

22.

Manning

Smith

(2019) Electoral performance by post-rebel parties. Government and Opposition 54(3): 415–453.

23.

Manning

Smith

Tuncel

(n.d.). Rebels with a cause: introducing the post-rebel electoral parties dataset. Journal of Peace ResearchIn preparation

24.

Pemstein

Marquardt

Tzelgov

, et al. (2022) The V-Dem Measurement Model: Latent Variable Analysis for Cross-National and Cross-Temporal Expert-Coded Data. V-Dem Working Paper No. 21. 7th edition. Gothenburg, Sweden: University of Gothenburg: Varieties of Democracy Institute.

25.

Riedl

(2014) Authoritarian Origins of Democratic Party Systems in Africa. Cambridge, UK: Cambridge University Press.

26.

Rustow

(1970) Transitions to democracy: toward a dynamic model. Comparative Politics. 2(3): 337–363.

27.

Sindre

Söderström

(2016) Understanding armed groups and party politics. Civil Wars 18(2): 109–117.

28.

Ufen

(2008) Political party and party system institutionalization in Southeast Asia: lessons for democratic consolidation in Indonesia, the Philippines and Thailand. Pacific Review 21(3): 327–350.

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

0.20 MB