Polarization in Contemporary Chile

To assess polarization in contemporary Chile, I focus on ideological and affective polarization and the rise of anti-identities in the electorate. Analyses of recent events in the country have identified two overarching trends: increasing polarization and persistent dealignment (i.e., the “electoral pendulum”), while focusing, in each case, on specific dimensions (e.g., left-right vs. affective polarization) or levels of analysis (e.g., political elites or the mass public).

Fabrega et al. (Reference Fábrega, González and Lindh2018) report evidence of growing ideological (left-right) polarization among the Chilean public during the 1990–2017 period and in the lead-up to the estallido. On this basis, the authors claim that Chilean society had left behind the politics of consensus that characterized its transition to democracy, especially from a demand-side point of view. Segovia (Reference Segovia2022), examining affective polarization and partisanship for the 1990–2021 period, reports two intriguing findings. First, affective polarization changes significantly over time: it is relatively high at the outset of Chile’s transition to democracy in the 1990s, declines until the mid-2000s, and then climbs again in recent years to levels similar to those observed in the early 1990s.

Second, Segovia also reports that in recent years affective polarization has been similarly high among both partisans and non-partisans. Even if the latter group does not identify with a party, it strongly dislikes parties on one side of the ideological spectrum (either on the left or the right) more than those on the other side. In other words, these findings point at the possibility of observing ideological polarization that is not effectively channeled into partisan representation. Similarly, Cubillos et al. (Reference Cubillos, Cox and Le Foulon2022) report significant affective polarization in the 2021 Gabriel Boric vs. José Antonio Kast presidential contest and among the Apruebo vs. Rechazo camps in the 2022 constitutional plebiscite.

These findings on affective polarization are consistent with Melendez’s (Reference Meléndez2022) claim concerning anti-identity structures in the contemporary Chilean party system. According to Melendez (Reference Meléndez2022), the Chilean system is structured around two negative (anti)partisanship identities such that the electorate mainly comprises two anti-movements (anti-Concertación on the right and anti-Chile Vamos on the left). The recent outflanking of the Concertación by the Frente Amplio (as well as the electoral growth of the Communist Party) on the left, and the outflanking of Chile Vamos by Republicanos on the right, might also be viewed as the consolidation of an anti-Concertación camp on the left, and an anti-Chile Vamos camp on the right. The growth of both the Frente Amplio and the Communist Party on the left, and Republicanos on the right, arguably have also strengthened both extremes at the expense of more centrist positions.Footnote ⁷ In this regard, the shrinking electoral space for centrist options and the fragmentation of “centrist” political parties are also noteworthy (Navia and Osorio Reference Navia and Osorio2015).

Melendez’s claim regarding anti-identities and findings on the nature of affective polarization in Chile are relevant to understanding a fundamental driver of Chile’s persistent interlude: strong anti-identities that stabilize electoral behavior. These anti-identities create a basis for the growth of “anti” candidates. However, candidates that can grow support by mobilizing one anti-identity usually collide with a low ceiling, which constrains further growth due to opposing anti-identities. For instance, it could be argued that the extreme Right’s potential growth is constrained by anti-authoritarian views associated with those that repudiate Pinochet’s regime and its legacy in Chilean society. The extreme Left’s potential growth is also constrained by deep-seated anti-communist identities. In the meantime, those who feel alienated from the “30 years” of center-left and center-right governments after the transition to democracy often repudiate candidates and party brands associated with that legacy. I will return to the implications of this scenario below.

Regarding the “pendulum” thesis, after Michelle Bachelet’s first term in office, alternation between ideological currents has been the norm (first to the right with Sebastián Piñera, then to the left with Bachelet’s second term, then to the right with Piñera’s second term, and more recently to a new left with the election of Gabriel Boric). The pendulum could also be observed in a more temporally compressed sequence if we consider the results of the elections related to the constitutional reform process. In the entry-plebiscite of 2020, nearly 80% voted for a completely elected constitutional assembly to draft a new constitution. The election of that constitutional assembly eventually ousted traditional party elites, especially those of the center and center-right, through the massive selection of independent candidates. Most of those independent candidates were elected as representatives of social movements and organizations active in the estallido.

However, in 2022, 62% of the electorate rejected the constitutional proposal produced by the assembly, which lost popular support during the process of drafting the proposal. Subsequently, in May 2023, a new assembly was elected, in which the extreme right (35.4% of the vote) and the center right (21.1% of the vote) obtained an absolute majority in the election of a new council in charge of drafting a new constitutional proposal. Finally, in December 2023, the option (“approve”) favored by Republicanos and center-right parties was defeated at the polls, with 55.8% of the people voting against their proposal.

This rapid oscillation might be interpreted in three ways. On the one hand, the shift might imply that in a system we often describe in terms of left vs. right positions, a significant chunk of the electorate rapidly switched preferences from the extra-systemic left to the radical right, and then again to the option favored by the governing center-left and left coalition.Footnote ⁸ This dynamic is consistent with “democratic protest voting” by an alienated electorate that partially benefits challengers (Hernández Reference Hernández2018). However, I suggest that the extremely compressed temporality in-between electoral shifts observed in Chile indicates an absence of “meaningful partisanship” (see Panicco and Anduiza Reference Pannico and Anduiza2023); i.e., the latter cannot explain the electoral behavior of those who continually switch allegiance from one camp to the other in between elections.

On the other hand, one could argue that Chile remains a country comprising three blocs of more or less equal size (Scully Reference Scully1992): a core bloc of voters, a core bloc of right voters, and a core bloc of centrists, who take turns joining the left or the right in shooting down any political project that empowers the rival bloc. If this conception is accurate, the basic discontinuity regarding previous periods would lie in the fact that each of these blocs is non-partisan (i.e., given the weakening of partisan organizations and the growing role of personalistic electoral mobilization, these blocs have now become more “sociological” than “political.”)

These two interpretations of recent electoral dynamics are not necessarily mutually exclusive; they may be complementary. Whereas the former interpretation more accurately describes the electoral behavior of non-politicized voters (who compose roughly half of the electorate now mandated to participate in elections), the second interpretation better represents the electoral behavior of social elites and highly politicized voters (roughly also half of the electorate.) In the following three sections I expand on this description by analyzing demand- and supply-side dynamics, and by illustrating the general dynamics of the system via a stylized narrative concerning the two recent (and failed) attempts at constitutional reform.

A third possible interpretation of recent electoral volatility is that the observed rapid oscillation responds to specific dynamics associated with recent turmoil and particular events (e.g., elections held during the pandemic and its immediate aftermath.) From this perspective, recent electoral change in Chile is, in a comparative perspective, rather mild. Although this third possibility is plausible and deserves closer scrutiny in future works, in the remainder of this essay, I will explore the first two interpretations and their possible implications.

Demand-side dynamics

To describe the evolution of different possible indicators of ideological polarization in the electorate, I draw on data from the CEP-Survey, which provides the longest public opinion series for analysis.Footnote ⁹ Overall, the different indicators presented in this section suggest that neither ideological polarization nor the crystallization of antagonist party identities has increased over time. Moreover, the data suggest that the events of the estallido correlate with several disruptions in long-term trends in public opinion, which have reverted “back to normal” in the most recent years.

Let us first look at ideological identification in the electorate. Figure 1 shows a great deal of continuity in where Chilean citizens locate themselves on the left-right spectrum since this question was first included in 2006, with centrist self-identification garnering more support than leftist or rightist identification.Footnote ¹⁰ Only in the aftermath of the estallido did the option “don’t know” become the majoritarian choice, while “rightist” self-identifications dropped significantly and “leftist” ones remained stable.

Figure 1. Ideological Self-placements in the Chilean Electorate. 1-10 Scale

The CEP-Survey also includes a question that asks respondents whether they identify with nominal ideological categories, such as “left,” “right,” and “center.” Data for this question date back to the beginning of Chile’s transition to democracy in 1989–1990. On this topic, the number of people who refuse to identify with an ideological category is systematically greater than the number of those who decline to locate themselves on a 10-point ideological scale, arguably because choosing a specific ideological category rather than a position on a number scale might imply a greater level of identification.

However, what matters most here is the evolution in the percentage of those who decline to identify with an ideological category, including the category of “independent.” Figure 2 shows that, while the percentage fluctuates, the long-term trend is an increase, with the percentage climbing from fewer than 20% in 1990 to more than 60% in recent years (with a sharp drop in the samples taken closer to the estallido of 2019). Although in the earlier years of this period non-identification was more characteristic of lower socioeconomic strata, it stands to reason that, as non-identification increases, the likelihood of respondents from the middle and upper sectors identifying with an ideological category decreases.

Figure 2. Ideological Self-placements in the Chilean Electorate. Nominal Scale: “Left, Center-Left, Center, Center-Right, Right.”

Because the left-right spectrum may fail to capture other salient ideological preferences well (or, more complicated still, because it may capture preferences for policies but not ideological identities), let us look at specific preferences for which longitudinal data are also available in the CEP-Survey, namely, attitudes regarding the need to generate greater income equality or to compensate individual efforts. Figures 3 and 4 display the evolution of these preferences. The estallido coincides once again with a moment of relatively high polarization, which increased during the decade leading up to 2019, with the proportion of people favoring moderate redistribution gaining ground against the most radical individualist view. Such polarization recedes in more recent measurements, where historically “normal” proportions (predominantly individualist) are observed (Figure 3).

Figure 3. Ideological Preferences Regarding the “Equality” (1-4) vs. “Individual Effort (7-10)” Tradeoff.

Figure 4. Ideological Preferences Regarding the “Individual liberties” (1-4) vs. “Security” (7-10) Tradeoff.

Another interesting trade-off is that between individual liberties and order and security. Although the CEP-Survey’s series on this question is shorter and is concentrated in the most recent period, the observed trend in Figure 4 is still telling. Once again, we observe temporarily heightened polarization of views around the time of the estallido. However, the preference for more order and security increases and gains hegemony over moderate and more radical views that prioritize individual liberties. The most recent results are similar to those observed in 2008, the only measurement available before the estallido. Recent results lean marginally more towards order and security, arguably due to the increased salience of violence and criminality in public opinion in recent years.

As we know, polarization might be non-ideological and, instead, may occur among different political identities that usually form two opposing camps (McCoy this volume). In turn, those identities might be partisan or non-partisan (Segovia Reference Segovia2022). To assess the extent of partisan affective polarization, Figure 5 shows the evolution of party identification in the Chilean citizenry since the mid-1990s, when close to a third of respondents identified with the Christian Democratic Party. However, since the early 2010s, those who do not identify with any political party have composed close to two-thirds of respondents, with this trend peaking at the time of the estallido and then receding somewhat in the most recent waves of the survey. The massive proliferation of new parties (see below) has not been able to compensate for the decline of traditional party identities.

Figure 5. Party Identification in the Chilean Electorate.

We know, however, that affective polarization might not be structured around strong party identities but, rather, around roughly defined camps (Segovia Reference Segovia2022) or even anti-identities (Melendez Reference Meléndez2022). Recent evidence on affective polarization in Chile, after the estallido, is consistent with the consolidation of a rightist and a leftist camp that do not necessarily map onto clear partisan cleavages (Segovia Reference Segovia2022). Moreover, qualitative and quantitative assessments of social elites’ attitudes and ideological preferences have argued that polarization has increased and runs deep in upper social strata (Gayo and Méndez-Layera Reference Gayo and Luisa Méndez2022; Undurraga et al. Reference Undurraga2023), which also coincides with the reasonable expectation that activist networks in civil society, given their greater intensity, are more polarized than the mass public. Opposing views on the “30 years” provide the structure for such elite-level polarization.Footnote ¹¹

Fine-grained quantitative electoral analyses of the recent election of the Constitutional Council (May 2023) and the 2023 plebiscite (December 2023) also suggest the crystallization of two camps among the wealthier districts in Metropolitan Santiago: a rightist camp, voting for Republicanos, but also in greater proportions for the traditional right in Las Condes, Vitacura, and Lo Barnechea (“las tres comunas”); and a leftist camp voting in Ñuñoa for candidates of Apruebo Dignidad formed by Frente Amplio parties and the Communist Party (Bro Reference Bro2022, Reference Bro2023). Bro (Reference Bro2023) also shows that the great majority of popular initiatives proposed during the constitutional process originated from organizations and individuals registered in these same wealthy municipalities in Santiago, with conservative proposals originating in Las Condes, Vitacura, and Lo Barnechea, and progressive ones originating disproportionately in Providencia and Ñuñoa. There is also a generational component that interacts with this elite-level cleavage. While the “Ñuñoa” camp is younger and predominantly composed of new generations of university-educated aspiring elites, the “tres comunas” camp is older and consists of traditional elites. This growing polarization among elites is also replicated (Sarmiento et al. Reference Sarmiento, Bravo-Marquez, Graells-Garrido and Poblete2022) and amplified (Scherman et al. Reference Scherman, Etchegaray, Browne, Mazorra and Rojas2022) in social media. Political shows in mainstream media and morning television and radio shows that, after the estallido, became strongly politicized, increasingly reproduce this cacophony of warring camps.

Despite this roar, the mass public does not (yet?) seem aligned with elite-level polarization. This is relevant because affective polarization often assumes the mobilization of large groups in society in terms an overarching “us” vs. “them” conflict structure (McCoy Reference McCoy, Rahman and Somer2018). Examining the mass-public is even more important today since the reinstalment of mandatory voting in 2022 implies the electoral incorporation of the roughly forty percent of (predominantly low-income) voters who had opted out of the system during the 2012–2022 period (PNUD-Chile 2017, 2020). Figure 6 shows the evolution of interest in politics in Chilean society. Nearly two-thirds of Chilean citizens consistently declare having low or no interest in politics. Again, the estallido coincides with a temporal disruption (inducing a bit more interest), but that trend recedes closer to traditional levels in the follow-up wave. If polarization entails massive politicization—i.e., that “major groups in society mobilize politically to achieve fundamental changes in structures, institutions, and power relations” (McCoy et al Reference McCoy, Rahman and Somer2018, pp. 16)—the latter condition is also absent among the mass public in Chile.

Figure 6. Interest in Politics in the Chilean Electorate

Low interest in politics has coincided with the consolidation over time of the so-called electoral pendulum. Since 2010, presidential incumbents and their governing coalitions have been systematically chastened in elections. Presidential approval ratings, shown in Figure 7, clearly depict the consolidation of an inverted-U trend (see Navia and Perello Reference Navia and Perello2019; Cabezas and Navia Reference Cabezas and Navia2019). Presidents start with high levels of popularity, which they lose rapidly, and then recover some of the original support by the end of their term in office. Losses have occurred increasingly rapidly in recent periods, and late recoveries have proven less solid, too. Recent presidential elections have become marriages without honeymoons. Again, if polarization entails a modicum of preference crystallization between opposing camps, the fluidity of electoral preferences and presidential approval does not show much crystallization in the Chilean mass public.

Figure 7. Sample WhatsApp messages sent to “my representative” in the Convention.

Source: PlataformaTelar (2021)

To sum up, in contemporary Chile, we observe not merely polarization (in some of its various possible interpretations) but, rather, two general trends on the demand side. First, especially after the estallido, we observe the radicalization of two camps among social elites: a younger, university-educated, aspiring elite vs. a relatively older, traditional economic and social elite (Bro Reference Bro2023). That polarization might reflect a generational cleavage, at least at the elite level, between those who came of age politically at the outset of the transition, and those who were politically socialized after 1990. This emerging structure among social elites provided the electoral bases for outflanking traditional parties with new challenger parties that mobilize discontent with the system on the left (Frente Amplio) and right (Republicanos) of traditional parties.

Second, such elite-level polarization thus far has not produced new and encompassing partisan organizations, capable of realigning the system. Although more stable sociological blocs exist on the left and the right, their roots in society are shallow. Moreover, the capacity of each bloc to craft an electoral majority is contingent on the (un)structured mobilization of centrist (and less intense voters), who vote in either direction based on contingent issues and negative identities. According to public opinion trends, the estallido created a window of opportunity to structure social conflict around different ideological camps (Noccetto et al. Reference Nocetto and Rosenblatt2023), which could eventually produce a meaningful realignment in the party system. However, the window closed quickly, arguably due, in part, to the social malaise associated with the economic and security effects of the COVID-19 pandemic, which bolstered a rapid conservative backlash.

Today, elite-level polarization is highly visible in Chilean society and is constantly spoken about in the press and social media. However, it still has shallow roots in society and seems to have alienated the mass public instead of aligning it. We thus observe serial discontent among a public that feels unrepresented by a political system (which now comprises not only traditional but also new electoral vehicles) that has become utterly unable to address social issues that have become increasingly salient since the estallido and the social and security crisis that followed it.

An illustration of Chile’s polarization syndrome: The constitutional reform process

The constitutional reform process envisioned by the political system to address the demands of the estallido is similarly illuminating. While it remains uncertain whether this process effectively appeased the movement—protests continued until the COVID-19 pandemic in March 2020—the surprising outcome of the constituent elections granted the process unprecedented legitimacy. Although the October 2020 plebiscite delivered a resounding mandate to replace the current constitution by involving new actors, the election of constituents in May 2021 provided even greater legitimacy. The traditional political parties were severely punished at the polls, as evidenced by comparing party-affiliated candidates’ participation in congressional and constituent elections. While party members consistently accounted for over 90% of those elected to Congress during the democratic transition, party lists composed only about 60% of the Convention members. Additionally, even those elected from party lists, such as socialist candidates grouped in the Socialist Collective, maintained a distant and tense relationship with their original parties.

The anti-system nature of the Convention is further exemplified by the socioeconomic background of its members. The Convention was the least elitist elected assembly in Chile’s history, as estimated by Bro (Reference Bro2022), and candidates who received the most financial support and spent the most on their campaigns received fewer votes than popular candidates (Plataforma Telar 2021). The unprecedented levels of descriptive representation in terms of gender parity and indigenous representation also contributed to the Convention’s popularity among the popular sectors.

The degree of descriptive representation achieved with the election of convention members generated adhesion at the level of the popular sectors. In this sense, a series of opinion panels among different socioeconomic groups and focus groups conducted by Plataforma Telar at the beginning of the constituent process reflected high citizen enthusiasm (Plataforma Telar 2021). On the one hand, positive feelings about the process predominated, except among the panel of business elites. On the other hand, in qualitative terms, the focus groups conducted among the popular sectors suggested that descriptive representation was especially valuable to this segment of voters.

In this sense, for example, the election of a Mapuche woman as president of the constituent body was seen as a positive sign as it distinguished the Convention from traditional politics. One participant in these focus groups stated, for example: “If they elected a Mapuche woman as president, it gives me hope that it will be something different from the usual politicians.” Accordingly, when participants were asked in the same study to send a message to “their” representative, the messages collected underscored the need for differentiation from traditional politics. Figure 7 lists a series of such messages, dominated by the expression “No…”, which directly states the demand for a different relationship than the one assigned to the traditional political system. Other responses convey a similar sentiment with different phrasing.

However, the Constitutional Convention did not fulfill those expectations and failed to represent the popular sectors. The inability of the “new” political system to structure the representation of popular sectors echoed that of the “old” party system; such inability has been recursive (Nocetto et al. Reference Nocetto and Rosenblatt2023), and more significant than other factors highlighted elsewhere (see Piscopo and Siavelis Reference Piscopo and Siavelis2023). Indeed, the collapse of the Convention was rooted in its inability to articulate the legitimate representation of social discontent while structuring a connection to the popular sectors that initially embraced the process (Alemán and Navia Reference Alemán and Navia2023).

Insights from a series of focus groups conducted by Plataforma Telar at the end of the constituent process, but before the exit plebiscite, shed light on the prevailing frustration among both the popular and middle sectors regarding the actions of the Convention. While there were initial expectations that the Convention members would redefine political representation, given their unprecedented levels of legitimacy compared to the traditional political system, by the end of the process, convention members and the Convention itself were perceived as “more of the same.”

Participants in these conversation groups noted that convention members “forgot about the people,” “never spoke to them again,” and fostered a “polarized” and “radical” debate perceived as detached from the interests and demands of their constituents. Instead, convention members emphasized specific “identity” issues such as the environment, indigenous rights, or feminism. However, the main emerging complaint in these conversation groups points to a relational deficit. Convention members were seen as having “abandoned their voters” while engaging in a “shouting match” that nobody understood. This perception closely resembles how traditional (elite) politics was viewed before and after the constituent process of 2021 and 2022.

Fragmentation and the incapacity of the larger blocks to coalesce around consensual proposals were also prominent in the legislative behavior of the Constitutional Convention between 2021 and 2022 (Plataforma Telar 2021). In this regard, independents and party representatives could not induce minimum levels of consensus and internal discipline to facilitate the approval of different proposals. Moreover, internal disputes occurred, such as within the Socialist Party, whose representatives in the convention and the Senate engaged in an open dispute, challenging each other’s legitimacy.Footnote ¹³

The available research on the Lista del Pueblo, the list of independent candidates that was most successful in contesting seats in the Convention, is also telling. According to Nocetto et al. (Reference Nocetto and Rosenblatt2023), independent lists missed a crucial opportunity to cleave a new division in Chilean politics (one that represented the anti-establishment sentiment of the estallido) while fostering new party organizations. Low barriers to entry, brought about by a permissive electoral system and the crisis of established parties (old and new), created an opportunity for massive electoral growth without organization. Such a formula nonetheless proved evanescent (Nocetto et al. Reference Nocetto and Rosenblatt2023).

Indeed, the trade-off the Lista del Pueblo exemplified between rapid electoral growth and organizational development could be extended to the new parties that emerged in the system recently. These new parties grew rapidly, filling the void left by discontent with the traditional political establishment and taking advantage of more permissive electoral rules enacted in 2015. However, these parties lack strong territorial and functional organizations besides their visible leadership. In that regard, otherwise “illegitimate” established parties have a competitive edge: they can run candidates and win elections at the local and regional levels where the new parties have remained inchoate beyond their specific electoral strongholds. These dynamics resonate well with extant arguments on the paucity of successful party-building instances and the conditions that enable it in contemporary Latin America (Levitsky et al. Reference Levitsky, Loxton, Van Dyck and Domínguez2016; Rosenblatt Reference Rosenblatt2018; Luna et al. Reference Luna, Rodríguez, Rosenblatt and Vommaro2021; Sánchez-Sibony Reference Sánchez-Sibony2024).

In sum, the Convention ended up reproducing systemic flaws and exacerbating the crisis of representation that has plagued the Chilean political system for years. After having represented citizens’ anger against traditional elites, the Convention replicated the historical disconnect between political representatives and the popular sectors. The latter’s silence delivered a resounding electoral result in the exit-plebiscite held on September 4, 2022. That plebiscite, the first election with compulsory voting in the country’s recent history, crystallized into a massive rejection of the constitutional proposal.

After the first Convention’s failure, the political system agreed upon a second constitution-making process. Mainstream political actors designed such a process under the premise that independents’ failure to represent popular demands had re-legitimized traditional politicians. Thus, the new process included highly restrictive electoral rules, a central role for “experts” proportionally chosen by political parties with representation in Congress, and tight constitutional boundaries to be imposed on the new text.Footnote ¹⁴

The election of the Constitutional Council in May 2023 produced two telling results. On the one hand, a resounding win (35.4%) for Republicanos, the right-wing party led by José Antonio Kast, the resilience of Unidos por Chile (28.6%), representing the governing Frente Amplio, the Communist Party, and the Socialist Party, and the shrinking role of traditional political parties on the center-left (8.9%) and the center-right (21%), which played a central role in negotiating the new process and in designing electoral rules (replicating the election system for the Senate) that in theory worked strongly in their favor. The Partido de la Gente, led by outsider figure Franco Parisi, failed to get a seat, garnering 5.5% of the vote.

On the other hand, in an election in which 84.8% of those mandated to vote turned out, 20.9% of voters cast an invalid (16.5%) or blank ballot (4.4%). Indeed, casting an invalid or blank ballot, taken together, was the third-most-popular option. The concentration of such unprecedented levels of invalid and blank voting among less politicized segments of the citizenry was eventually explained by people’s lack of information (and interest) in the new constitution-making process (the government, which actively promoted voting in the exit plebiscite of September 2022, implemented a much more hands-off approach the second time around.)Footnote ¹⁵ However, low interest reflects, in and of itself, people’s disengagement from the political process. Indeed, according to Medel and Kouba’s (Reference Medel and Kouba2023) careful reconstruction of the geographic distribution and correlates of invalid voting, it is plausible to interpret such voting as an outright rejection of the process and of the political pact that produced it.

Popular rejection of the process was later ratified in the December 2023 exit plebiscite, in which the “against” option obtained 55.4% of the vote. During the constitution-making process, the Republican Party missed an opportunity to strike an efficient pact with the center-right, which could have been instrumental in moderating some of the more controversial norms that made it into the final text. In any case, the popular rejection of the proposal confirms voters’ persistent alienation from a political system from which they feel disconnected, a system they perceive as a distant shouting match.

Chile’s Polarization Syndrome: Disjointed Polarization

The Chilean party system has increased its ideological polarization on the supply-side of politics (political party elites), while also displaying affective polarization in the electoral arena (i.e., around identities and anti-identities that get mobilized in specific electoral races). In line with Hernandez (Reference Hernández2018), one could argue that Chile is a case of “democratic protest voting” whereby old parties get replaced by new challenger parties. However, the stability of allegiance to emerging challenger parties and its possible role in “cueing” voters’ subsequent political behavior (Pannico and Anduiza Reference Pannico and Anduiza2023) is not yet observed (i.e., in a significant segment of the electorate we observe dealignment without realignment; in other segments we observe the emergence of ideological blocs that lack partisan encapsulation and organization on the demand side). Indeed, the supply side has significantly increased its fragmentation, while encompassing partisan organizations (and even broad electoral coalitions) that could cue and mobilize sizable electoral groups have continued to weaken.

The trends mentioned above are complemented by other relevant (and partially inconsistent) demand-side dynamics. Ideological and partisan dealignment, as well as discontent with political elites have continued to grow, and have crystallized in the serial penalization of incumbent elites at the polls, as those that successfully challenge incumbents in one election, subsequently prove unable to efficiently structure conflict and societal discontent while in office. Consider, for example, national-level elections from Michelle Bachelet’s first term in office (2006–2010) until the recent election of a Constitutional Council in 2023, thus encompassing the two plebiscites and two constitutional convention elections that took place in that period. In those eight national-level elections, the only common denominator is the resounding defeat of incumbents.

The broader implication of this type of cycling is the weakening of legitimate order in society caused by national incumbents’ serial incapacity to restructure representation and sustain a modicum of legitimacy that enables them to govern effectively. In such a context, political challengers can mobilize “negative” popular majorities in opposition to whoever is in office, but serially fail to consistently mobilize “positive” majorities structured around specific political projects. Such serial failure induces the repeated cycling of electoral preferences that characterizes the emerging configuration in Chile. This repeated cycling, which helps make disjointed polarization durable in Chile, is driven by alternating but serial discontent with incumbent parties and figures.

Since remedying such discontent has proven beyond the grasp of those who successfully challenge incumbents, we observe radicalization on the supply side without well-crystallized and encompassing electoral poles on the demand side. For instance, if the 2021 runoff revived the old authoritarian-democratic cleavage, the positioning of the old political elites around the Approve and Reject options in the 2022 exit plebiscite seems to have definitively buried that cleavage (since many on the center-left supported the Reject option, along with the Right). Chile’s polarization syndrome is thus characterized by increasing polarization and fragmentation on the supply side, and a combination of recurrent discontent and affective polarization on the demand-side. While elites are increasingly radicalized, popular sectors have sequentially rejected the system as a whole.

Chile’s polarization syndrome might be inductively characterized as disjointed polarization. In disjointed polarization, polarizing trends at some levels of analysis (e.g., political elites) can coexist with significant dealignment at other levels of analysis (e.g., a large group of voters). Moreover, in disjointed polarization, some actors can polarize along the left-right dimension (e.g., established political parties), whereas other actors can polarize along other dimensions, such as the competition between system insiders (the establishment) and outsiders (e.g., anti-establishment politicians, as well voters alienated from traditional elites).

Although the Chilean syndrome might be interpreted as an indicator of McCoy et al.’s (Reference McCoy, Rahman and Somer2018) “gridlock and careening” scenario, which results from growing affective polarization between two relatively even camps, Chile’s current configuration in this regard is more nuanced. On the one hand, political elites often align around two camps (government vs. opposition forces), but are also increasingly fragmented, which has made for growing levels of internal conflict and volatility within each camp. In turn, wealthy and well-educated voters are more affectively-polarized, especially after the social protests of 2019.

However, a significant fraction of the electorate, where poor and lower middle-class voters are overrepresented, and who since 2022 are required to vote, have grown alienated from politics and do not consistently align with either of the two camps. Those voters either support outsider candidates (particularly under permissive electoral rules such as those guiding presidential and municipal elections, as well as the ones instituted for the Constitutional Convention elected in 2021) or switch allegiance from camp to camp because of contingent events and candidate specific traits. As a result, close to 50% of Chilean voters display durable and politically unstructured discontent.

Disjointed polarization is not, to be sure, a new phenomenon. We can find transient processes of disjointed polarization in the classically theorized interregnum between dealignment and realignment (Key Reference Key1955; Flanagan and Dalton Reference Flanagan and Dalton1984). What is new about disjointed polarization, as empirically observed in contemporary Chile, is its durability.

The Chilean political system displays durable disjointed polarization because new party organizations, which have flourished and gained traction in recent years, have rapidly been inoculated with systemic traits. One such trait is rootlessness (Luna and Altman Reference Luna and Altman2011). In this regard, for instance, in comparison to movement parties like the Bolivian MAS or the Brazilian PT in the 1980s (Anria Reference Anria2018), the Chilean Frente Amplio is virtually unable to structure the representation of the popular sectors. The counterpart of rootlessness is structural illegitimacy and the fluidity of electoral adherence to a party’s leadership. At the same time, anti-identities remain strong, and affective polarization around specific electoral instances has arguably grown. However, these anti-identities can only be mobilized circumstantially by political parties and candidacies, which rapidly lose their grip on public opinion once elected to office. Party rootlessness in society is thus essential in perpetuating polarization disjointed.

Might disjointed polarization be self-reinforcing?

Part of what we observe in Chile’s disjointed polarization might be framed as an instance of “populist” demand. A possible interpretation of the Chilean situation is, thus, that the country is headed toward a populist realignment. As elsewhere, drawing on widespread but hitherto unstructured sources of discontent with the system and its representatives, an apt leader might eventually be able to constitute a new cleavage from above by mobilizing “the people” against “the elites.” Increasing ideological and institutional polarization levels should follow, as in other instances of polarization via successful populist mobilization (Roberts Reference Roberts2022). Although plausible, this outcome has consistently failed to crystallize in Chile.

Under specific contexts, moreover, we might alternatively observe “generative polarization,” which instead of leading towards top-down populist polarization, leads to the institutionalization of a new cleavage structured around a couple of encompassing partisan organizations (Le Bas 2018). However, in contemporary Chile, party building attempts by emerging leaders (either from the left, from the right, from the center, or from outside the system) have systematically failed to capture “people’s imagination.” Again, although plausible, successful party building that could provide the basis for generative polarization has not yet been observed in the case since functional substitutes for traditional parties have consistently failed to consolidate new enduring representation structures, leaving the system in a long-lasting interlude.

To address this puzzle, in the remainder of this section I entertain a different, perhaps less plausible possibility: that disjointed polarization relates to a new normal, in which low levels of electoral structuration and high levels of fluidity prove long lasting (and recurrent) given politicians serial inability to structure societal discontent via legitimate political representation. Two series of factors might underpin the durability of disjointed polarization: self-reinforcing factors that are endogenous to the type of disjointed polarization observed in Chile, and exogenous constraints on politicians.

In terms of endogenous or self-reinforcement mechanisms, disjointed polarization contingently aligns supply and demand by default in the context of elections that structure access to power. Anti-identities provide a basis of support for different camps, but they also create low-ceiling effects. For that reason, those who win elections essentially do so by default. Once in office, they cannot sustain their strategy while exercising power, and thus become premature lame ducks opposed by new and vociferant aspirants to office. This makes for polarization at the elite level, but such polarization does not match the structure of preferences increasingly displayed by a majority of voters who rapidly grow alienated from those in power. While this cycle repeats itself, disjointed polarization endures.

A generalized lack of political legitimacy is a defining trait of disjointed polarization. Politicians have legitimacy when they can credibly “buy time” to solve citizens’ most salient needs. Constant political turnover and the compression of newly appointed incumbents’ honeymoons mean that democratic leaderships are serially “out of sync.” These leaders, thus, cannot create socially legitimate order and prove unable to govern and implement public policies, once they arrive in office.

Recurring discontent and political scandals undermine leadership legitimacy, thus rapidly defeating attempts at successfully synchronizing subjective and objective temporalities. Compressed temporalities also shape politicians’ strategic choices in a context where traditional political parties have exhausted their symbolic and organizational resources (Rosenblatt Reference Rosenblatt2018), and new electoral vehicles have mushroomed, but very few instances of successful party-building are observed. In this context, while non-partisan ideological identities persist and continue to structure the electoral behavior of more politicized voters, the personalization of politics has increased. Today, successful electoral mobilization pivots around “non-political” figures (with privileged access to economic resources or popularity cultivated outside of politics) who self-select into politics and seek to make quick headway. However, even if they succeed at such attempts, they face rapid illegitimacy.

In institutionalized party systems, aspiring candidates had to work their way up through the system over a long period. The payoff was that those who had a successful career could eventually govern. Candidates operating in contexts such as Chile’s disjointed representation have a much easier time gaining office at the local and national levels. To gain office, they can polarize affectively and ideologically, but they often resort to anti-partisan and anti-establishment platforms, as well. The personally wealthy and those who cultivated a reputation and some name recognition outside politics (i.e., those with successful careers in sports and/or media personalities; social movement leaders) can now successfully enter the political arena. Rapid declines in legitimacy and weaker governance coalitions are the counterpart of rapid entry opportunities for outsiders.

In terms of exogenous mechanisms that might prolong disjointed polarization, it is worth considering that the political regimes and systems of interest representation we seek to understand are embedded in specific social and state structures, which have fundamentally changed in the last few decades (see Roberts Reference Roberts2014; Handlin Reference Handlin2017; Silva and Rossi Reference Silva and Rossi2018; Collier and Handlin Reference Collier and Handlin2015). By stressing too greatly the autonomy of politics and excessively focusing on elite-centered analyses, we have often downplayed the role of state and social structures in shaping relevant outcomes at the party system level. Weak state capacity is an overarching, primordial factor behind such outcomes, and one that is pervasive across Latin America (see Mainwaring Reference Mainwaring2006; and Mainwaring and Pérez-Liñan Reference Mainwaring and Pérez-Liñán2023 for a similar argument).Footnote ¹⁶

For parties to connect with citizens and consolidate their role as legitimate mediators between the state and society, it is fundamental that when parties win elections and have a chance to govern, they can deliver public goods. After all, the whole idea of party programs is that the winning party will implement their programs and produce results that their supporters want instead of just making statements or passing laws that do not translate into citizens’ everyday realities. Yet, the chronic weakness of Latin American states, along with increasing electoral fragmentation (i.e., another relevant endogenous mechanism), hinders the formation of programmatic party organizations. State weakness also drives the entrenchment of corruption and criminal capture of institutions.

If we rely on the conventional distinction in the social sciences between politics and administration (Mazzuca Reference Mazzuca2010), this deficit is, strictly speaking, not a problem of the supply side of politics. Politicians can come to office with ambitious programmatic agendas and be extremely apt. However, they must work with a public administration developed in the past and they pay a political cost for past failings in state building (Mazzuca and Munck Reference Mazzuca and Munck2020).

Deficits in state capacity frequently undermine politicians’ ability to bring about tangible changes. And the failure to deliver the results citizens expect weakens the link between citizens and politicians and undermines the prospects for building and entrenching parties in society. More often than we admit, in weak states, those constraints might also undermine populist leaders’ capacity to consolidate their following once in office.

The weakening role of states might seem welcome in terms of curtailing opportunities for clientelism and patronage. However, we also often overlook the fact that clientelism and patronage were instrumental in entrenching the traditional political organizations we now yearn for (for Chile see Valenzuela Reference Valenzuela1977.) Moreover, in a context in which parties are weak, state campaign funding meager, and campaigns increasingly expensive, businessmen and organized crime organizations have become better able to advance their interests via campaign financing. When socially visible (often through scandals), this new reality further weakens popular allegiance to political leadership.

More generally, the failure to deliver on campaign promises and produce results leads, rather predictably, to the questioning of politicians. Indeed, a common pattern in Chile (and arguably elsewhere in Latin America) is that expectations regarding the performance of governments are dashed and that public discontent with officeholders, and the parties they belong to, rapidly ensues. In sum, state weaknesses, shallow and segregated public good provision, and state capture exacerbate political discontent with incumbents (traditional politicians or their electorally successful challengers).

Socio-structural change is also massive in Chile, as well as in other contemporary Latin American societies. The atomization of society along functional and territorial lines that do not scale up creates parallel universes, which politicians often fail to structure around a handful of encompassing divides. If democracy held out (and partially fulfilled) the promise of instituting civil, political, and a modicum of social rights, the atomization of the experience with democracy, the state, and social incorporation complicates efforts at aggregating social preferences (Luna and Medel Reference Luna and Medel2023).

State-society interactions are atomized across multiple territorial and functional arenas while engaging different societal actors and social groups across space. The dynamics of social protest provide an interesting illustration of the consolidation of parallel universes. Either we have intense, locally oriented protest activities that do not aggregate at the national level in a consistent movement, as observed in the contestation of neoextractivist activities, or massive spasms of nationwide protest that, no matter how disruptive they may be, ultimately fail to articulate politically beyond the streets.

Like never before, the aggregation of parallel universes (both across localities and functional arenas) is a fundamental challenge that politicians can only partially overcome. Bridging those parallel universes (particularly aggregating interests from the local to the regional and national levels as well as forming public policy coalitions based on compromises between different functional groups) has become increasingly difficult for the handful of territorially and functionally encompassing political organizations that are still electorally viable. Parallel universes thus complicate, in particular ways, the structuring of political representation.

Luck and economic voting may entrench some leaders a bit longer while forcing others out of office more promptly (Campello and Zucco Reference Campello and Zucco2020). This was the case, for instance, during the region’s commodity boom (2005–2015). In general, however, new leaders come and go more frequently than in the past. Usually, these leaders can run successful political marketing campaigns, which eventually produce evanescent election-night electoral coalitions. Electoral mobilization “technologies” that allow for a successful aggregation across local settings and social groups in elections are pivotal in explaining electoral outcomes and their eventual limitations for subsequent government action.