Born with a silver spoon? Modes of transitions and democratic survival

ABSTRACT This study examines why regime survival rates vary across young democracies. The literature offers competing claims regarding the effect of the mode of transition on the duration of post-transitional democracy. This study reconciles these claims by proposing three modes of democratic transitions – military dominance (MD), popular sanction (PS), and consensual power transfer (CPT) – arguing that CPT leads the subsequent democracies to last longer than MD and PS. MD fails to incorporate the military into democratic systems, making it more likely for the ensuing democracies to suffer a coup, whereas PS enables regime insiders to change democratic rules without hindrance or outsiders to gain power through an organized armed conflict. CPT shapes a strong electoral performance by authoritarian successor parties, which provide checks and balances in post-authoritarian politics, thus decreasing the likelihood of collapse. This study verifies these hypotheses using data on nascent democracies between 1945 and 2022.


Introduction
Transition marks the interval between one political regime and another, and transition to democracy is delimited, on the one hand, by the launching of the process of dissolution of an authoritarian regime and, on the other hand, by the installation of some form of democracy. 1As a precondition to shaping institutions, party politics, and the social structure of an ensuing democracy, it is possible that the types of transition affect a country's prospect for democratic consolidation.To the best of my knowledge, there is a lack of a comprehensive and systematic framework to address how the mode of transition influences democratic duration.
The literature examines how the mode of democratic transitions can lead to different likelihoods of democratic survival. 2This could be built based on two major dimensionswhich actors control the process of transition and whether all actors are willing to seek consensus.They argue that the opposition teams up with the incumbent to create transitions that will help democracy survive, whereas elites-led transitions will hurt it.There are several points that merit further investigation.First, existing studies attribute democratic dysfunction to elite-led transitions, as the lingering power of old-regime elites and the absence of an effective counterbalance to the elites that oversees the transition impede democratic consolidation.The present study distinguishes elite-led transitions between those initiated by civilian and military elites, indicating that their willingness to accept democratic rule and power configurations in post-authoritarian politics are very different from each other.This distinction explains how democracy survives.Second, the existing research assumes that the transition to democracy starts from a process of the dissolution of a closed authoritarian regime, or a regime with restricted electoral competition.The present study identifies incumbent electoral defeat as a form of transition in competitive authoritarian regimes.Third, existing empirical studies on how transitions impact democratic survival do not distinguish between types of democratic breakdowns.The present study establishes multiple paths of democratic termination and analyses how different modes of transition affect the success and failure of a democracy.
This study proposes three modes of democratic transitionsmilitary dominance (MD), popular sanction (PS), and consensual power transfer (CPT).First, the proposed mode of transition distinguishes elite-led transitions between those initiated by civilian and military elites.It identifies incumbent electoral defeat as a form of democratic transition.Second, the proposed mode of transition encompasses two dimensions: actors' acceptance or rejection of democratic rules and power configuration in post-authoritarian politics, which can account for democratic survival.Finally, the proposed mode of transition explains the distinct ways in which democracies transition back to non-democracies.
Overall, it argues that CPT leads the subsequent democracies to last longer, when compared with MD and PS.MD fails to incorporate the military into democratic systems.The military can claim legitimacy in securing a country's institutional stability and intervene in politics and empower itself to alter a constitutional order with ease, thus making it more likely for ensuing democracies to suffer a coup.PS causes regime outsiders to see violence against incumbents as legitimate and gain power through an organized armed conflict.Furthermore, it can enable insiders within the power void to fashion a new constitutional order to cripple their opponents, which will pave the way for a gradual autocratization.CPT primarily creates a consensus on a liberal constitution and inclusive competition, ensuring that incumbent elites and opposition groups will not achieve power through undemocratic means.Second, it shapes a strong electoral performance by authoritarian successor parties (ASPs), which provide checks and balances in post-authoritarian politics, thus decreasing the likelihood of collapse.Using data covering all democracies between 1945 and 2022 and employing competing risk survival and mediation analyses with survival data, this study verifies this hypothesis.
The study addresses several studies on comparative politics.First, this study adds to the richness of existing literature on transition types 3 by distinguishing elite-led transitions and identifying incumbent electoral defeat as a form of transition in competitive authoritarian regimes.Second, this study adds to the literature on coups and democracy, 4 implying that, although the army installs competitive elections after a coup, succeeding democracies are more likely to experience a second military coup.Therefore, a coup-led democratization is not only associated with higher levels of state-sanctioned violence 5 but is also highly susceptible to a regime collapse.Third, this study distinguishes three distinct modes by which democracies revert to nondemocracies, arguing that transition types, rather than constitutional systems, economic growth, and regime crises, determine the different forms of regime change.It thus contributes to comparative research on modes of democratic breakdown. 6inally, this study emphasizes how transition types affect the performance of ASPs, which will exert influence on subsequent democratic survival, thus contributing to existing research on the dynamics of ASPs in democracies 7 and how they affect democratic survival. 8

Literature on modes of transition
The early literature on modes of transition emphasized which actors control the process of transition. 9They distinguish the relative power of incumbent elites and counterelites, thus differentiating the types of regime transition.Stepan establishes seven paths towards democracy, except for three paths in which war and conquest play a decisive part. 10These seven paths can be categorized into two based on whether the authoritarian elites themselves or the opposition forces initiate the move towards democratization.Mainwaring develops a three-fold typology based on the extent to which the incumbent authoritarian power dominates the transition process. 11However, it rapidly became clear that this single criterion did not capture more sophisticated transition dynamics.
An alternative perspective that distinguishes transitions is whether all actors are willing to seek consensus.Markovitz argues that we cannot assume that transitional democracies will automatically become consolidated. 12Rather, the consolidation depends on the extent to which major political actors in a society become willing to accept the institutionalization of a new regime type.This argument, in accordance with Linz and Stepan, highlights the need for transitions to reach an agreement on post-authoritarian institutional arrangements to produce a consolidated democracy. 13ombining the two perspectives, Karl and Schmitter use two dimensions that capture if elites or counterelites can unilaterally and effectively force a change, or must seek a compromise. 14Democracies that have endured for a respectable length of time appear to be those defined by relatively strong elite actors who engage in strategies of compromise, as contending forces agree to forego their capacity to harm each other by ensuring that each other's vital interests will be respected.Munck and Leff also depend on these two perspectives to distinguish the different transition processes into conservative reforms, reforms from below, through transaction, extrication, rupture, or from above, and social revolutions. 15Reforms through extrication make the steady progress towards democratic consolidation easier, as the incumbent and counterelites jointly drive the transition process, accept the need for elite contestation, negotiate on constitutional issues, and subsequently create a consensus in the new system for both old and new elites.Negotiated transitions are characterized by the balance of power. 16When opposition groups and incumbents are relatively equal in power, the transition tends to be characterized by bargaining and negotiation.Stradiotto and Guo argue that cooperative transitions are associated with the longest democratic survival, as reforms occur through joint cooperation between incumbents and the opposition that will protect mutual interests. 17Conversely, in countries in which elites of the ancien régime maintain a decisive power advantage, democracy dies. 18he existing research argues that elite-led transitions negatively affect democratic outcomes.When old regime elites retain their monopoly in the transition phase, opposition groups refuse to accept the new rules of governance because of a high level of dissatisfaction and mistrust regarding the process of building democratic institutions. 19Further, as the old regime elites are still able to get involved in the creation of new institutional rules there is an unclear break with the past.This makes an institutional legacy that hinders democratic consolidation more likely. 20This correlation between democracy and governance is valid only if authoritarian elites are incapable of imposing a constitution on a new democracy prior to their departure from power. 21Evidence shows that transitions fuelled by elites are associated with fewer substantial regime changes and lower democratic quality. 22here are four points that merit further investigation.First, scholars have proposed that the elites-led reform process may contribute to democratic decline. 23However, in many elites-led transitions, such as Bulgaria, Spain, or South Africa, the incumbent elites prefer to break with the old order, liberalize the regime, and set a more level playing field, thus sustaining democracy.It also indicates that elite-led transitions will pose a threat to democratic survival, if elites tend to change formal rule to further protect their interests or extend their influence in the post-authoritarian era.Consequently, it is necessary to distinguish elite-led transitions when explaining how democracy survives.
Second, elites-led transitions allow the incumbent elites to singlehandedly set their agenda in the transition process and likely suppress the post-transitional competition.Accordingly, the absence of an effective counterbalance to the elites that oversees the transition impedes democratic consolidation. 24However, allowing incumbent elites to impose their interest in institution building does not pave the way for authoritarian reversal, if there is a corresponding force that can effectively check their power in subsequent democracies.Unlike revolution, which generates a power void, elite-led transitions help incumbents occupy the leading position and impose their demand.In this way, the opposition is more likely to distinguish between "them" and "us," unite divergent social forces, and hold alliances together or build up a party against the common enemy.A balanced pattern of competition that provides checks and balances in postauthoritarian politics reduces the odds of non-coordinating parties, or a dominant power that hampers democratic consolidation.Consequently, it is necessary to examine the power configuration in post-authoritarian politics when investigating the impact of transitions on democratic survival.
Third, existing studies assume that the transition to democracy starts from the process of the dissolution of a closed authoritarian regime or a regime with restricted electoral competition.Thus, the relative power advantage of competing groups shapes the nature of the transition process and determines their willingness to negotiate institutional choices in a new democracy.However, competitive authoritarianism, a common form of autocracy after 1991, combines democratic rules with authoritarian governance.In competitive authoritarian regimes, elections are regularly held, and are open and competitive. 25A move towards democracy transpires when incumbents suffer electoral defeat and allow the winner of the election to take office.In other words, accepting the persistence of formal democratic rules is key to the transition from competitive authoritarian regimes to democracy.Consequently, it is necessary to identify such transitions, and examine the disposition of actors to accept or reject the rules of a democratic game, when investigating the impact of transitions on democratic survival.
Fourth, democracy faces multiple competing risks.Although the existing empirical studies on how transitions impact democratic survival do not make such a distinction, democratic breakdowns can be classified into several main groups. 26Consequently, it is necessary to distinguish multiple paths of democratic termination and understand how different modes of transition affect the success or failure of a democracy.
These four points mentioned previously indicate that we should differentiate among patterns of transition: (1) to distinguish elites-led transitions, and (2) to identify transitions where authoritarian incumbents accept electoral defeat.Further, the power configuration in post-authoritarian politics and actors' acceptance or rejection of the rules of the democratic game should be examined when accounting for how the mode of transition impacts the democratic survival.Finally, the patterns of transition should explain the distinct ways in which democracies transition back to nondemocracies.

Arguments
This study distinguished three modes of democratic transitionmilitary dominance (MD), popular sanction (PS), and consensual power transfer (CPT)using the implications mentioned in the previous section.MD is defined as a government ousted by members of the military of the regime, which subsequently holds a competitive election or hands over power to a transitional government for the purpose of overseeing and holding an election.PS refers to the situation under which opposition groups take the lead in toppling authoritarian incumbents and bringing about democracy through the use of unarmed popular uprising or organized armed conflict as insurgencies or revolutions.CPT indicates that the incumbents suffer electoral defeat and allow the winner of the election to take office, or the incumbents determine to change the rule under which all parties can participate, or suffrage is extended to most of the population.In other words, democratization is the result of the elites' active or passive (due to a powerful opposition) transfer of power in CPT.
The proposed mode of transition distinguishes elite-led transitions between those initiated by civilian and military elites.It also identifies incumbent electoral defeat as a form of democratic transition.
The proposed transition mode includes two dimensions: actors' acceptance or rejection of the rules of the democratic game and the power configuration in post-authoritarian politics, which can account for democratic survival.When analysing whether actors accept or reject the rules of the democratic game, PS causes incumbent elites not to accept a new constitutional order.This is because opposition groups are able to control the transition, and the institutions or norms are built with little constraint exerted by the incumbent elite, making it likely for old regime elites to see this violent form of transition as legitimate.Similarly, MD tilts the balance of power towards the military during the transition and is thus assumed to help the military's refusal to relegate itself to a subordinate position within the civilian constitution, rendering the constitution a toothless tiger.Under CPT, by contrast, the regime elites are able to agree on a compromise, where they relinquish power in return for an inclusive competition, or to allow opposition groups to challenge their authority in subsequent elections.Opposition groups accept and see the form of power transfer as legitimate, as they do not have to resort to violence against the government's unwillingness to hand over power or push for open competition, and the incumbents are able to see their survival as not being threatened by a regime change.
Power configuration in post-authoritarian politics is very important.The military exerts the most influence in politics owing to its monopoly of power in the transition phase under MD, by creating a new constitutional order that protects its interests and reduces the power of civilian leaders.Under PS, opposition groups in a power void are expected to design post-transitional institutions to their advantage, thus facilitating the emergence of a dominant power after democratization.There are two circumstances in CPT.One is that opposition groups are powerful enough to defeat old regime elites in elections, or force the incumbent government to transition the state.The temporary exit from power and peaceful power transfer mark an opportunity for incumbent elites to exert the resources they inherit to compete in subsequent elections.It thus generates a powerful parliamentary opposition engaging in serious bargaining, with policies or constitutional issues, or likely returning to power.The other is a dominant incumbent regime that can actively push for democratization and still hold the upper hand in determining the rules that are accordingly adopted in the new democracy.Such civilian elite-led democratization offers incumbent elites a leading place at the table in a new democracy, which will encourage a divided opposition that unites to pursue a shared objective, such as targeting the legitimacy of the incumbent's reign and challenging their capacity to govern.In either case, CPT, compared with MD and PS, contributes to diminishing power disparities between incumbents and their challengers.Thus, CPT reduces the risk of democratic failure, owing to the actors' acceptance of the rules of democratic game, and provide a more balanced pattern of competition in post-authoritarian politics.
The proposed mode of transition explains the distinct ways in which democracies can revert to nondemocracies.The transition through MD causes a new democracy to have a higher chance of reverting to authoritarianism through coups.One possible reason is that the military can claim its ability to save the country in a time of crisis, 27 thus making it more likely for an unconstitutional interruption of the democratic order.When a coup brings about a change from dictatorship to democracy, they often announce their intention to hold and oversee democratic elections in the near future, offer a timeline for the transition, and even allow the candidate or party that won to take office.In this way, the military can establish a kind of justification that it ousted illegitimate regimes and chose an assumed legitimate government under the rules of the democratic game. 28Such a form of political legitimacy, however, offers the military a justification to intervene and rule in subsequent democracies.The military, as an impartial political arbiter, can denounce that the democratic government has lost the legitimacy to rule by accusing it of mismanagement or its inability to tackle some dangerous forcessuch as communists, terrorists, or separatist rebelsand use these as a pretext to seize power. 29In Thailand, the reason for the continued legitimation of military power is social tensions and border problems, which rationalized the need for stronger armed forces for internal security.Meanwhile, a civilian rule rife with corruption and abuse of power rationalized the coup. 30lthough military leaders rarely compete in post-authoritarian elections, they play an important role in major democratic transitions, for example, by fashioning the Constitution to strengthen their power and annul the unwanted influence of civilian leaders.Once their goal cannot be achieved, however, they may grab power through unconstitutional means and seek to consolidate their power through lasting new political arrangements.The postcoup Thailand's Constitution of 2007, drafted under military sponsorship, bears anti-democratic traits, including a fragmentation of the party system, the appointed half of the Senate being filled with military surrogates, and an executive branch with much-reduced authority. 31After discovering that the current political system had failed to erode the power of officials seen as loyal to former premier Thaksin Shinawatra, 32 the military used public disorder as a pretext for its intervention in 2014.In contrast to the 2006 coup that transferred most of the power to a civilian government, the 2014 coup kept tight control, and redesigned the constitution to restructure the country's political system. 33Therefore, MD may contribute to a higher likelihood of democratic failure through coups.
PS is associated with a higher likelihood of democratic collapse through organized armed conflict or autocratization.The reason for democratic failure is the absence of a compromise between incumbent elites and opposition groups regarding the rules of a new democratic game that advances mutual interest and thus reduces the possibility of a seizure of power through undemocratic means.In this way, political actors who were violently removed see this type of power transfer as legitimate, thus resorting to armed forces that topple the current government.In the Republic of the Congo, the process of democratic transition through a popular uprising caused rival partiesone led by President Pascal Lissouba and the other by the antecedent dictator Denis Sassou-Nguesso to recruit armed wings or private militias to halt the election campaign of the other side.In fact, the opposition gained control of the 1991 National Conference to chart Congo's political future, rewrote the Constitution, and stripped then-President Sassou-Nguesso of most of his powers, without his involvement.The short-lived democratic regime ended with a series of clashes and violent overthrows of the Lissouba government by Sassou-Nguesso's forces. 34In fact, existing research argues that when a regime is ousted by popular uprising, insurgencies, or revolutions, the subsequent regime is often associated with an increased likelihood of conflict and with possible escalation and war involvement. 35ransition through PS often creates a greater power void, under which opposition groups or major candidates and parties endorsed by opposition forces dominate the transition process.Actors who are in an advantageous position tend to fashion a new constitutional order to cripple their opponents and secure continued rule, paving the way for the gradual autocratization in post-transition politics.In Pakistan, violent demonstrations deposed Yahya Khan, and Zulfikar Ali Bhutto assumed office after his party won an absolute parliamentary majority.Owing to the lack of a strong opposition that effectively checked the government's power, Bhutto easily amended the Constitution and the laws concerning the use of emergency powers and preventive detention.Under these amendments, the major opposition party was dissolved, and its leaders were arrested. 36Further, it is likely that after the violent overthrow of a dictator, the power void leads opposition movements to appear divided and struggle over the distribution of power, which escalates into an armed conflict against each other.
CPT offers the greatest promise of democratic success, and one reason for this is its ability to create a peaceful transition where incumbents and opposition groups work together.It may be that incumbents lose an election and allow the winner of the election to take office in exchange for an inclusive competition that ensures that both sides will not come to power through undemocratic means.It may also be that incumbents decide to change formal or informal rules under which all parties can participate, or suffrage is extended to most of the population.By accommodating all actors in a liberal constitution and a competitive election, winners or losers are more likely to cultivate a tolerant attitude.This is important because those excluded from political participation cannot get the recognition they want, but instead seek to oust the regime through extra-institutional means, while those in power resort to more violent actions against opposition groups.In contrast, inclusive competition offers opposition groups a means to find a place in the new government and protects the interests of the incumbents once they become political minorities.For example, after its defeat in the 2002 Kenyan elections, President Daniel Arap Moi expressed his willingness to step down because Mwai Kibaki, the winner of the presidential election, promised that he would not seek to investigate him or ban his party Kenya African National Union (KANU).Kibaki, instead, brought several high-level incumbents from KANU into his government. 37This made Kenya's democratic consolidation more likely.Therefore, CPT reduces the risk of democratic failure through outsiders' armed conflicts.
Another positive effect of CPT on democratic survival is its facilitation of authoritarian successor parties (ASPs) which can decrease the risk of political instability associated with extra-institutional force-led regime change.As ASPs are still veto players, there is no room for new ruling parties to be divided.For instance, after its defeat in the 1990 Nicaraguan elections, in addition to organizational reformation, the long-ruling Sandinista National Liberation Front (FSLN) mobilized its grassroots organizations to protest against the new government.To deter the FSLN from returning to power, Arnoldo Alemán marshalled the divided liberal forces to establish a highly organized Constitutionalist Liberal Party and coalesced with three other parties to form the Liberal Alliance.The alliance became an electorally strong anti-Sandinista force in the late 1990s. 38As strong contenders in the first and subsequent democratic elections, the FSLN created incentives for the new ruling coalition to remain united.
ASPs may encourage cooperation among the opposition if they win the elections.For example, after the National Salvation Front (FSN), the successor to the Romanian Communist Party, won a significant victory, the Christian Democratic National Peasants' Party (PNT-CD) decided to ally with centre-right parties to oppose the FSN. 39In the 1990s, the two primary political forces in Romania were the FSN and the PNT-CD.Thus, whether in the opposition or government, a strong electoral performance by ASPs reduces the likelihood of non-coordinating parties, which helps stabilize subsequent democracies and reduces the risk of intervention by extra-institutional forces.
ASPs also reduce the potential for new ruling parties to change the formal rules that lead to autocratization.The electoral success of ASPs assures the electorate that if they are unsatisfied with the performance of the government, they have a clear alternative.In this scenario, ASPs find it easier to mobilize public support when they criticize the government's policies, condemn its economic or democratic failure, or propose bills to curtail the government's power, while incumbents find it difficult to alter the electoral outcome through extensive vote buying, by imposing a ban on other parties, or by beating, jailing, or killing opposition leaders or supporters.Conversely, the electoral failure of ASPs often encourages new ruling parties to cross the line from democracy to authoritarianism.This is because actors who are in an advantageous position tend to fashion a new constitutional order to cripple their opponents and secure continued rule, paving the way for the gradual autocratization of post-transition politics.For instance, in Guinea-Bissau, the electoral weakness of the ASP, the African Party for the Independence of Guinea and Cape Verde, enabled the new president Kumba Ialá to refuse to promulgate the 2001 Constitution, instead sending it back to parliament with recommendations for increased presidential powers.President Ialá even dissolved the parliament in November 2002, repeatedly postponed elections, and sought to manipulate the law to detain members of opposition parties without charge, ensuring that he would remain in power. 40hese potential mechanisms indicate that a strong electoral performance by ASPs creates a balanced pattern of competition that provides checks and balances in postauthoritarian politics.It reduces the likelihood of non-coordinating parties, thus inoculating young democracies against political instability that could lead to regime collapse or the emergence of a dominant power hindering democratic consolidation.In other words, ASP's superior electoral performance results in opposition collaboration, forming an electorally competitive party that can present itself as a compelling alternative government.This regular and robust interparty competition implies that ASPs and their rival parties exhibit better electoral performance for an extended period, thereby sustaining democracy for a considerable time.Furthermore, both sides may become more familiar with how the rules of the democratic game work, making them more likely to adhere to these rules.The long-term positive outcomes of CPT on the survival of democracy are not solely contingent on inclusive competition that ensures neither side attains power through undemocratic means but also on the repeated utilization of competition.
CPT allows ASPs to perform well in post-authoritarian politics.First, a peaceful power transfer to the new incumbent, due to electoral defeat, marks the former ruler's willingness to maintain inclusive competition.When inclusive competitions are respected, antecedent ruling parties can exert the old regime's resources, including a party brand, clientelistic networks, and sources of party finance, addressing the government's poor performance for their comebacks.In Mexico, the Institutional Revolutionary Party (PRI) graciously accepted defeat in the 2000 presidential election, as inclusive elections were endorsed by all major political parties.The peaceful turnover of power ensured an opportunity for the PRI to utilize the old regime's clientelistic structures, business community allies, and the government's mediocre economic performance for its return to presidency in 2012. 41In Taiwan, the Kuomintang (KMT) accepted its defeat in the 2000 presidential election, as securing inclusive competition provided the KMT with an indispensable safety net against a catastrophic free fall. 42nce the rules of the democratic game were maintained, its immense organizational network at the grassroots level and hundreds of real estate properties and business enterprises enabled the KMT to survive and resurge.
Second, authoritarian incumbents are likely to opt for democratization when regimes possess substantial antecedent political strengths and resource advantages. 43n other words, the decision to concede requires ruling parties to gain confidence that democratic politics will secure their continued dominance, or at least will not bring the party's electoral demise.Resources determine incumbents' antecedent strengths, and a return of preceding rulers is likely if they have developed a party that has an institutionalized mechanism, a territorially encompassing infrastructure, and sources of party finance.For example, the Bulgarian Communist Party (BKP) liberalized the political system in the presence of a relatively weak opposition, which helped it to secure some favourable decisions such as using a mixed electoral system to prevent a disastrous electoral performance.The organizational continuity and enormous property of the BKP, that the newly renamed Bulgarian Socialist Party chose to maintain and inherit, allowed them to come back to power even after its defeat in the 1997 and 2001 elections. 44able 1 summarizes the comparison of the three modes of transitions.The study made the following hypothesis: H1: CPT results in the greatest survival rate after democratization compared to other transitional modes.H2: Compared to other transitional modes, MD contributes to the democratic decline through coups, while PS results in democratic collapse through organized armed conflict or autocratization.H3: The electoral performance of ASPs mediates the effect of CPT on the longevity of democracies.

Empirical analyses
First, this study defines the terms it studies.The unit of analysis is the duration of nascent democracies between 1945 and 2022."Democracies" are defined either as regimes where the executive achieves power through a direct competitive election, in which at least 10% of the total population is eligible to vote, all major parties are permitted to compete, and neither fraud nor violence determine the election outcome; or as regimes where power is won through an indirect election by a body, at least 60% of which is elected through direct competitive elections. 45A "nascent democracy" referred to a country that was ruled by a dictator for some years after 1945, but subsequently transitioned to a democratic political regime.Based on these criteria, 125 nascent democracies are analysed.Appendix A contains the list of nascent democracies (1945-2022). 46he dependent variable is the failure or survival time of a democratic regime in years.A democratic regime is considered to have ended under the following conditions: (1) if an executive achieves power through undemocratic means; (2) if the government achieves power through democratic means, but subsequently changes formal or informal rules; or (3) if competitive elections are held to choose the government, but one or more parties for which substantial numbers of citizens will be expected to vote are banned. 47The data therefore cover 125 democracies, including 2254 pooled time series and cross-sectional observations between 1945 and 2022, using data from the Geddes, Wright and Frantz (GWF) Autocratic Regimes dataset (1945-2010, for democratic regimes after 2010, see Appendix D-III).
As for independent variables in modes of transition, this study creates a dummy variable where 1 = CPT and 0 = Non-CPT, in which MD and PS are included, using data from the codebook of the GWF dataset, which provides the narratives of the start and end events for each regime (1945-2010, for democratic regimes after 2010, see Appendix D-III).Of the 125 regimes, 44 achieved democracy through CPT, whereas 40 experienced regime transitions towards democracy under MD.PS led 36 dictatorships to convert to democracy, and another five built democracy by assassination, or foreign intervention or invasion. 48Appendix B contains the list of the modes of transition for these nascent democracies (1945-2022).
A list of control variables that influence the duration of new democracies has been added.These include constitutional design, economic development, economic growth, ethnic and religious fractionalization, former British and French colonies, the timing of democratization, prior experiences with democracy, and a dummy variable to indicate whether a country was governed by a single communist party guided by Marxism-Leninism (See Appendix C for full details).
To confirm my theoretical expectations, this study uses the Cox PH model with time-variant covariates. 49The Cox PH model is robust, ensuring that the results obtained closely resemble those from the correct parametric model.The Cox model is a safe choice because users do not need to worry about selecting the wrong parametric model.It considers survival times and censoring, in addition to a (0, 1) outcome, which makes it preferable to the logistic model.To test for robustness, this study first tests if, relative to MD or PS, CPT reduces the risk of democratic collapse by dividing non-CPT into the two modes of transition.Second, this study investigates how transition types influence the survival of a democratic regime in the initial 20 years (see Supplementary Figure 1), as democracies face a greater risk of reversal into authoritarian regimes in the few years immediately after democratization. 50Third, transitional similarities in post-Soviet countries show that democratic survival may have been influenced by the lasting effects of the antecedent USSR rule. 51Instead of former communist states, this study controls for countries governed by the USSR dictatorship during the Cold War. 52Fourth, a prolonged experience with democratic rule should have a strong and positive influence on democratic survival.For example, this may be seen through economic growth 53 or enhanced human development. 54The risk of democratic breakdown may also increase with a larger number of regime collapses in the past. 55Rather than using the number of democratic transitions, this study controls for both the number of years of democracy since 1945, and the number of past regime changes.Finally, the survival time is assumed to follow a specified distribution; thus, parametric models are adopted. 56This study also considers frailty as an unobserved individual factor.Because regimes are more likely to encounter higher risks of collapse within some countries, this study uses the shared frailty model to capture these common but unobserved characteristics.
A glance at the figure reveals the impact of modes of transition on the survival curves of democratic regimes (see Supplementary Figure 2).CPT leads to subsequent democracies that last longer than their counterparts achieving democracy through MD or PS.However, the survival curves of democratic regimes with prior modes of transition, such as MD and PS, are not significantly different.Table 2 presents how the threat of the democratic regime collapse varies according to modes of transition, and other variables.Model 1 is a Cox PH model, with time-variant and time-invariant covariates.CPT lowers the risk by 64% compared to MD and PS.Model 2 is an extended Cox model used to correct variables from Model 1, in which the PH assumptions are not met.The model demonstrates that the CPT lowers the risk of abrupt regime collapse in new democracies by 68%.Finally, a series of robustness checks substantiate my theoretical expectations (Models 3-6).Similar results are found in the parametric model with shared frailty (see Supplementary Table 1).
Table 2 further confirms the impact of certain control variables on the duration of new democracies.Higher levels of GDP per capita and positive economic growth both reduce risk of democratic collapse.This shows that economic conditions can still affect how long a democratic regime lasts.Furthermore, countries that enter democratization later tend to have longer democratic survival.This indicates that they may gain insights from the lessons learned from failed democracies in the past.However, even when these factors are considered, we still observe the independent effects of the transition types.
Overall, this study observes that: CPT will have the greatest survival rate after democratization compared to other transitional modes (H1).

Causal mechanisms
This study examines the mechanisms by which MD and PS are likely to lead to democratic collapse.This study considers survival data in which each democratic regime could experience only one of the different types of failures over follow-up.The dependent variable is the failure or survival time of a democratic regime in  There are at least two ways in which a democratic regime can fail, but only one such failure type can occur in each incident of failure.Thus, this study uses competing risk survival analysis.It is based on the sub-distribution hazard (SDH) model and models the SDH function derived from a cumulative incidence curve.This gives the hazard rate for a cause-specific event at time t, based on the risk set that remains at time t, after accounting for all the previously occurring failure types.
Figure 1 reveals how the variability in the hazard of democratic collapse through a coup is determined by its mode of transition.In the presence of competing risks, MD increases the risk of a democratic regime being toppled through a coup (3.8-fold, left panel; 6.1-fold, right panel, compared to CPT). Figure 2 presents how modes of transition relate to the risk of autocratization and armed conflict.Compared to other transition types, transitions through PS are associated with a higher risk of failure through autocratization and armed conflict (2.6-fold, left panel; 2.6-fold, right panel, compared to CPT).My theoretical expectations are confirmed through robust and multivariate statistical methods (see Supplementary Table 2-3).Overall, this study observes that: MD makes it more likely for ensuing democracies to suffer coups, while PS enables regime insiders to change democratic rules without hindrance or outsiders to gain power through an organized armed conflict (H2).
This study examines the mechanisms by which CPT, compared to MD and PS, promotes regime survival.If my theoretical expectations are true, CPT shapes a strong electoral performance by ASPs, which provide checks and balances in post-authoritarian politics, thus decreasing the likelihood of a collapse.This study identifies a list of ASPs between 1945 and 2022 according to Loxton's definition. 57It uses data from the literature, conventional country reports, newspaper articles, and academic blog articles. 58A dummy variable is created to indicate whether ASPs perform well in a new democracy, with 1 = ASPs that have a majority of seats in the legislature or whose party members are appointed prime minister or win the presidential elections, and 0 = otherwise.Appendix E contains a list of ASPs with superior electoral performance.Similarly, this study also controls for the ten variables mentioned previously.This study uses mediation analysis with survival data to explain the mechanism that links the transition types and electoral performance of ASPs to the duration of a democratic regime. 59For robustness tests, the analysis uses the seat share of ASPs as an alternative measure of their electoral performance, and examines the succeeding democratic survival in the initial 20 years after democratization and includes different sets of control variables.
Following Baron and Kenny's steps for mediation, the following three steps confirm a mediation effect. 60First, both the mode of transition and ASP electoral performance affect democratic survival.Table 3 shows that CPT reduces the risk of democratic failure by 64% compared to MD and PS (Model 1), while the strong electoral performance of ASPs is associated with a decreased risk of regime collapse (Model 2: -83%).Second, the mode of transition is associated with ASP electoral performance; CPT is likely to promote ASPs with the potential to achieve significant electoral success in post-authoritarian politics (see Supplementary Figure 4). 61Finally, the effect of the mode of transition on democratic survival disappears (or at least weakens) when ASP electoral performance is included in the regression.Table 3 shows that the effect of modes of transition on democratic survival is reduced when the explanatory variable and the mediator are included (Model 3: -55% compared to Model 1: -64%).Furthermore, a series of robustness checks confirm this mediating effect (see Supplementary Figure 5).Overall, this confirms that transitions through CPT lead to the strong electoral performance of ASPs and thus have a positive impact on the longevity of democracies (H3).

Conclusion and discussion
The literature offers competing claims with regard to the mode of democratic transitions and their effects on subsequent democratic survival, which largely focus on single-case research and small-N comparative studies.This study reconciles these claims by proposing three modes of democratic transition, and employs cross-national statistical analyses to confirm that CPT leads subsequent democracies to last longer than MD and PS.
There is a growing body of literature focusing on autocratic regime types to understand subsequent democratic survival, highlighting competitive electoral experience, 62 regime openness, 63 and power-sharing between dictators and elites 64 to distinguish dictatorships.It is possible that authoritarian regime type is associated with the mode of transition.When a regime allows limited pluralism, holding multiparty elections where opposition groups can challenge the incumbent, the ruler may suffer electoral setbacks.The ruler can also co-opt the opposition and employ divide-and-rule strategies to avoid widespread class-based challenges; therefore, elite-led regime change is more likely. 65Further, when a regime establishes legislatures, parties, and the judiciary to share power with the ruling elites, it can accelerate the liberalization of the political system.This makes it more likely that elite-led regime changes will become the dominant form of transition. 66This suggests that distinct autocratic rules may be responsible for both the mode of transition and democratic survival but also suggests a causal pathway that links competitive authoritarianism and/or dominant-party dictatorships and CPT to result in a durable democracy.Future studies should investigate this relationship by identifying suitable instrumental variables.
The analysis shows that MD makes it more likely for subsequent democracies to suffer a coup, whereas PS enables regime insiders to change democratic rules without hindrance or outsiders to gain power through an organized armed conflict.However, this study fails to consider the impact of foreign intervention on the duration of democracy.First, foreign military intervention often ousts rulers, similar to PS, which can create a power void in the transition process that will raise the likelihood of a democratic breakdown.Second, as a regime change is unilaterally imposed by dominant foreign power, not all domestic actors are given the opportunity to peacefully hand over power, thus reducing their incentives to foster consensus during the transitional period.However, democracy may proceed smoothly if such interventions aim to topple rulers who resort to electoral fraud, or suffer electoral defeat but reject abdicating power, as is the case in Panama and Gambia.This is because opposition groups, instead of subverting state power, seek to abide by democratic rules in competitive authoritarian regimes, and consequently accept institutional arrangements after democratization to some extent.There were very few observations.In future research, we will test theoretical expectations when the sample size increases.
The analysis shows that CPT leads ASPs to perform well in post-authoritarian politics, which provide checks and balances in post-authoritarian politics, thus offering the greatest promise of democratic success.There is mixed evidence regarding how ASPs influence democratic outcomes. 67Given this opposing evidence, this study examines a full sample of ASPs and their effects on democratic survival.To my knowledge, this is the most comprehensive quantitative study, to date, of ASPs in new democracies.Existing research confirms that ASPs performing well in post-authoritarian politics are often those that have inherited positive resources. 68The results indicate that CPT ensures an opportunity for ASPs to utilize the old regime's resources, and devise strategies to offset the liabilities of an authoritarian past for their rise, like a phoenix from the ashes.Future research should consider the interplay between resource abundance and opportunity structures.
There are at least two policy implications that warrant consideration.First, efforts to guide the autocrat to push democratization, by changing the rule under which all parties can participate might be more conductive to the survival of democracy, if the use of such strategies is conditional on the existence of ASPs that perform well in post-transitional politics.Second, scholars of transitions and democratic consolidation could pay close attention to institutional factors such as the existence of multiparty elections that raise the likelihood of incumbent defeat and thus facilitate the emergence of CPT.

Notes
DEMOCRATIZATIONyears. Further, multiple risks faced by a democratic regime are distinguished: coup, autocratization, and organized armed conflict.A military coup is defined as an ouster by the military of the regime in power.An organized armed conflict refers to an insurgency, revolution, or civil war, leading up to the ouster.Autocratization occurs when an executive changes the rules of democracy, such that competition in subsequent elections is limited.The study uses data from the literature, conventional country reports, newspaper articles, academic blog articles, and the codebook of the GWF dataset, which provides narratives of the start and end events for each regime (1945-2010; for the list of autocratization and organized armed conflict between 1945 and 2022, see Appendices D-I and D-II).Of the 125 democracies, 36 were overthrown by a coup, whereas four were ousted by an organized armed conflict.20 more regimes failed due to autocratization, while 65 still remained democracies at the end of 31 December 2022.Supplementary Figure3shows the frequency of the types of democratic failure, based on the three modes of transition.

Table 1 .
Comparison of three modes of democratic transitions.
Source: The author.