Tom Ginsburg is Leo Spitz Professor of International Law, Ludwig and Hilde Wolf Research Scholar, and professor of political science at the University of ChicagoAziz Huq is Frank and Bernice J. Greenberg Professor of Law and Mark Claster Mamolen Teaching Scholar at the University of Chicago. They are coauthors of the new book How to Save a Constitutional Democracy (University of Chicago Press, October 2018).


Recent and sharp deterioration in democracy in countries as far apart as Hungary, Venezuela, Russia, and Turkey has prompted intensive attention to “how democracies die.”1 Two lines of inquiry dominate the resulting debates. There is, first, vigorous disagreement about the magnitude and practical significance of the current democratic decline.2 Second, those who agree that a democratic recession is taking place have turned their attention to questions of causation: What economic, institutional, or social forces are doing the most to strain or undermine democratic structures? A great deal of important work, much of it in these pages, has closely examined recent instances of decline and fall.3 These studies have illuminated the interactions among institutional structures, choices by political actors, and larger economic or social developments in pressing democracy into retreat.

But the investigation of democratic decline should not begin and end with the instances in which democracy has ended up on its deathbed. An exclusive focus on such cases—no less than an exclusive focus on successful democratic transitions and consolidations—creates a risk of selection bias: By selecting on the dependent variable (democratic failure), we compromise our conclusions about the relative strength of the various forces in play when democracies comes under threat. A more robust understanding of democratic backsliding requires attention to both cases in which democracy fails and those in which it succeeds.4 At a minimum, researchers must take account of those instances in which [End Page 16] democracy survives major challenges in order to make strong causal claims about the conditions of democratic failure.

There are many cases in which democracy’s survival might be “explained” by the absence of any plausible threat to the democratic system in question. These cases might illuminate the general social and economic conditions in which democracy thrives, but they are not likely to speak to the factors that can repel a threat to participatory government once such a threat has arisen. Rather, we think that students of democratic design should pay special attention to a narrower class of democracy’s “near misses”—that is, the cases in which a democracy is exposed to social, political, or economic forces that could catalyze back-sliding, yet somehow overcomes those forces and regains its footing. The study of such near misses can yield lessons that are not apparent either from looking at outright democratic failures or from examining a large pool of heterogeneous cases in which democracy is only sometimes at risk. In particular, focusing on near misses helps bring to light the institutional mechanisms or situational conditions that redound to democracy’s benefit in a moment of crisis. It allows us to ascertain the countervailing social, institutional, or political factors that increase a democracy’s chances of survival under stress. In short, it can help us to understand not only why democracies die, but also why they persist under conditions of severe pressure.

Our hope is that by clarifying the concept of democratic “near misses” and identifying several core case studies (as well as a set of more debatable ones), we can stimulate a richer range of inquiries into democratic decline. We hence define a near miss for democracy as a case in which a country 1) experiences a deterioration in the quality of initially well-functioning democratic institutions, without fully sliding into authoritarianism, but then, 2) within a timeframe of a few years, at least partially recovers its high-quality democracy. Graphically, this can be imagined as a case in which the quality of democracy evinces a U-shaped dip over time. Depending on the depth and duration of the dip, this definition can encompass two discrete sets of countries: first, those in which democracy was under severe threat but survived intact; and second, those which may have fallen below a minimum threshold of democratic quality into a nascent competitive authoritarianism, albeit one that nevertheless was quickly reversed. We might call the former true near misses and the latter quick comebacks. Yet the boundaries between and beyond these categories can be fuzzy, and reasonable observers can disagree about how to characterize given cases. The key point is that democracy can survive under threat, and we need to understand how this happens.

Our definition admittedly is not fully inclusive insofar as it singles out those instances in which changes in the quality of democracy are visible. Countries might brush up against democratic failure without any such [End Page 17] manifest changes coming into view. For example, if international actors anticipate the risk of backsliding and intervene in a timely manner, visible deterioration in the quality of democratic institutions might never materialize. In post–World War II Europe, the United States provided military, political, and economic assistance under the aegis of the 1947 Truman Doctrine purportedly in order to “support free peoples” against “attempted subjugation by armed minorities or by outside pressures.”5 Of particular note, U.S. officials injected millions of dollars of aid into Greece, which had by then already slipped into civil war, while stepping in to restructure the country’s tax and budgetary policies. At the same time, U.S. military advisors supplied and retrained the Greek National Army. As a consequence, the Greek insurgency had subsided by 1949.

Between that year and 1952, Washington provided some US$12 billion in loans, grants, and technical assistance across various European countries, albeit without the degree of micromanagement employed in Greece. The trajectory of European democracy would, without question, have been substantially different had this Marshall Plan aid not been forthcoming. Its sheer magnitude might well have kept signs of democratic slippage from appearing in postwar European countries other than Greece that were nonetheless at risk of destabilization. Such cases would also involve the U-shaped dynamic of decline followed by revival that characterizes democratic near misses—but they are extremely hard to flush out. For the sake of clarity and tractability, therefore, we leave them to the side for now.

Understanding democracy’s near misses is of particular importance given the manner in which democratic backsliding now most often occurs. In previous work, we have argued that slow erosive processes have replaced fast collapses as the most common form of backsliding in recent decades.6 Erosion, in which democracy is undermined in a series of incremental steps, some of which can be defended in legalistic terms, is alluring for would-be autocrats because it can yield the same outcome as a frontal attack on democratic institutions, but without the same risks. By blurring the bright lines that mark the boundaries of permissible political contestation, erosion robs democracy’s defenders of clear focal points around which to organize. Like the proverbial boiling frog, the opposition often finds that by the time it notices the threat, it is already too late to coordinate an effective response.7

Nonetheless, erosion’s effects on the possibility of democratic retrenchment are complex: While the gradual nature of this process increases the coordination costs of mustering a prodemocratic response, the longer timeframe involved also creates windows during which some institutions and political forces can mobilize to mitigate or even undo backsliding. In contrast, when a democracy collapses because of a military coup or an antidemocratic revolution, no window of opportunity for countermobilization by parties and institutions will arise. Instead, if [End Page 18] democracy survives an attempted coup d’état, there is generally a simple explanation: The coupmakers lacked either the raw power or the internal coordination to succeed, or may have confronted an immediate and effective popular backlash. Under conditions of erosion, institutions and forces that are simply irrelevant in cases of democratic collapse may successfully militate against democratic failure—and a close study of the dynamics unfolding during a democratic near miss may shed light on these institutions and forces. Moreover, comparing these dynamics with the prodemocratic strategies that failed in countries such as Hungary, Turkey, and Venezuela could provide helpful clues for those in the United States and other established democracies who wish to resist global trends in democratic backsliding.

Identifying Near Misses

What counts as a true near miss or a quick comeback for democracy? We explore two ways of approaching this question. The first is quantitative, and draws on the existing indices used to compare the quality of democracy among nations and over time. Such standard measures of democratic quality, however, have limited value in pinpointing the marginal cases marked by swings away from and then back toward democracy. Different measures indicate very different subsets of cases as near misses. This striking divergence points toward two conclusions. First, for the task at hand, quantitative tools are not (yet) sufficiently precise that we can use them to identify the cases in which democracy is imperiled through erosion. Second, and more generally, our analysis suggests that the various standard indices of democracy are in fact measuring slightly different aspects of political systems’ performance. They should not be assumed to be interchangeable.

Given the limits of existing quantitative tools, we instead turn to a case-study method. That is, we examine a core group of clear near misses from different regions and time periods. We focus on three highly differentiated examples: interwar Finland, contemporary Sri Lanka, and contemporary Colombia. In Finland, the black-shirted Lapua movement that emerged in 1929 threatened to bring about a descent into authoritarianism of the sort that Germany and Austria were soon to experience. In Sri Lanka, the government of Mahinda Rajapaksa leveraged its 2009 victory over the Liberation Tigers of Tamil Eelam (LTTE) to start building a nepotistically organized, patronage-driven autocracy. In Colombia, President Alvaro Uribe first set aside a presidential term limit established by the constitution in 2004, then tried to do so again in 2009—part of a broader effort to centralize and entrench political authority. In each of these three cases, however, backsliding ceased, at least for a period, and a more robust species of democratic contestation reemerged. We recognize that these cases are not wholly uncontroversial; for example, some commentators [End Page 19] had raised doubts about the health of democracy in Sri Lanka and Colombia even prior to the events in question. While acknowledging such concerns, we think that in each of our case studies there is clear evidence of a deterioration in democratic quality followed by a partial recovery.

We can draw from these examples several lessons concerning the elements of a democratic system that matter most in times of crisis. First, both the Finnish and the Sri Lankan cases show that party elites from within the political camp aligned with a potential authoritarian can play pivotal roles in forestalling backsliding. Second, the Sri Lankan and Colombian examples suggest that democratic crisis can be staved off by nonelected, nonmajoritarian institutions. The key players in the Sri Lankan case include the military and the bureaucracy. In the Colombian case, an unelected judiciary played a central role. Strikingly, in these initial case studies of near misses, democracy’s survival turned out to depend greatly on actors outside the ambit of democratic contestation (even if the judiciary is generally thought to be part of the architecture of liberal democracy). Rather than elected institutions, it was unrepresentative elites or unelected bureaucrats and judges who proved pivotal to democracy’s defense in its moment of peril.

How can we identify democracy’s near misses? Perhaps the most obvious approach would be to examine those instances in which there was a meaningful decline in a generally accepted measure of democratic quality followed by an uptick in the same variable. One would look, in short, for a U-shaped curve in a numerical measure of democracy; differences in the depth of the curve might then be used to distinguish between true near misses and quick comebacks. Several different metrics purport to capture the quality of a nation’s democracy at any particular point in time. We focus on the three most widely used measures: the Freedom House, Polity, and Varieties of Democracy (V-Dem) indices, which have the advantage of greater temporal and geographic coverage than other assessments (such as those produced by Bertelsmann or the Economist Intelligence Unit). If these measures reveal a common set of cases, then the study of democratic near misses should plainly begin with those.

We examined the data from these three indices to identify instances in which a democracy had deteriorated (but not crossed a threshold into competitive authoritarianism or the like) and then recovered. No common core of results emerged. Instead, different measures of democratic quality point in startlingly different directions when it comes to identifying near misses. This suggests that, at least in cases near the margin of democratic failure, the standard measures of democratic quality are capturing different aspects of political systems. It follows that efforts to distinguish quantitatively between true near misses and quick comebacks will be unavailing. [End Page 20]

We searched through all three datasets for instances in which a relatively stable democracy suffered a decline in its rating, but then reverted to the status quo ante after a relatively short interlude. The Freedom House measure ranks all countries since 1972 as Free, Partly Free, or Unfree. Examining all Freedom House data since 1972, we observe only a small number of countries that suffered a one-year category drop, only to see their previous rating restored the next year in an enduring way. If we limit our cases to countries that were rated Free for at least three years before declining, only three instances meet this criterion: Bolivia (1995), Guyana (2005), and Sri Lanka (1975).

Polity employs a 21-point scale of democracy. Using Polity scores of 6 or higher as the cutoff for considering a country democratic, we observe only four countries that experienced “interludes” between extended period of democracy. All these cases involve coups (if we include President Alberto Fujimori’s 1992 “self-coup” in Peru). Some entail what Ozan Varol has termed a “democratic coup d’état,”8 in which the military intervened to restore democracy, and then stepped quickly out of the way. Yet others appear to be instances of what Dan Slater calls “democratic careening,” in which a nation stumbles haphazardly from democratic to authoritarian regimes and back without either becoming fully institutionalized.9 Fiji, for example, experienced a dip in its Polity rating first between 1987 and 1991; once again after a coup in 2000; and a final time between 2007 and 2014.

Finally, if we consider the V-Dem “liberal democracy” index, there are relatively few cases where countries that had levels of democracy at or above the global mean suffered a drop of, say, 20 percent or more in the ratings, only to recover and enjoy a sustained period of democracy. India from 1975 to 1977 is more or less the only case that fits this pattern.

The fact that none of the indicators overlap with regard to particular cases implies that a quantitative approach to case selection will be of limited effectiveness. The Freedom House, Polity, and V-Dem indices offer different lenses for viewing a complex concept, and they do not generate consistent diagnoses of near misses. Rather, it would seem that these measures are sensitive to slightly different kinds of changes in the quality of democracy, reflecting their focus on subtly different elements of the overall political system.

Three Case Studies

If quantitative methods are of little avail in investigating the problem of democratic near misses, it seems appropriate to consider a more qualitative approach. We think it is useful to focus on a limited subset of cases in which 1) the risk of substantial democratic erosion had already become apparent, either through institutional deterioration or in the political agenda of an insurgent antisystem faction or leader, but 2) [End Page 21] the danger clearly and plainly receded with the defeat of the political force behind the threat. As we noted above, this definition encompasses only a subset of the near misses category; it excludes what may well be a larger class of cases in which external intervention or chance political factors prevented any observable erosion. We rely on country-specific scholarship and expertise to identify three cases that fit our criteria for near misses: Finland between World War I and World War II; Colombia in 2010 in the midst of a roiling civil war; and Sri Lanka in 2015, six years after the suppression of a violent insurgency that had raged on and off since 1983. All three countries had been relatively stable democracies in fragile environments. All confronted a significant threat to their democratic health. But all survived at least that crisis.

These examples vary usefully across time and space—and, importantly, in the institutional dynamics that proved pivotal to averting further democratic collapse. They also illustrate a range of overlapping mechanisms whereby party elites and nonelected actors can shore up democratic continuity. Hence, Finland crisply illustrates the role of political elites, both those potentially aligned with an antisystem party and those in the opposing camp. The Colombian case shows how a nonelected institution designed ex ante to function as a check on majoritarian actors—here, a constitutional court—can play a critical role by insisting on respect for rule-of-law norms at key moments. Finally, the Sri Lankan case highlights the critical importance of both party elites, who are able to generate plausible electoral competition, and unelected bureaucrats in the election-administration and military spheres. In fact, these cases uniformly point toward the necessary role of unelected and nonmajoritarian actors in safeguarding and preserving democratic institutions.

We do not, to be clear, think that these examples are exhaustive. There may be other cases of near misses among the histories of Europe’s interwar democracies, whose varied trajectories have provided the basis for many earlier comparative studies of democratic robustness.10 One might argue that a more recent example of a near miss prevented by a nonmajoritarian institution occurred in December 2017, when the South African Constitutional Court issued a ruling requiring the national legislature to explore impeachment proceedings against President Jacob Zuma. Another nonelected “checking” institution—the independent watchdog agency known as the Public Protector’s Office—also featured in the complex process leading up to Zuma’s February 2018 ouster: The trigger for this process was a report on “state capture” compiled by Protector Thuli Madonsela, which exposed how Zuma’s family and allies had used a vast corruption-powered network to exert influence over the South African state.11 We leave to others a more comprehensive cataloging of democratic near misses. Here, we use our case studies to develop a more robust conception of a near miss and a better understanding of the mechanisms at work when such an event occurs. [End Page 22]

Finland 1930: In November 1929, red-shirted communist youth paraded in the small Finnish village of Lapua, located in the country’s religious and conservative southern Ostrobothnian region. An angry mob of local farmers attacked the parade, stripped the participants of their shirts, and began beating the unlucky leftists. That seemingly isolated and chance incident sparked a “a series of events which proved almost fatal to parliamentary government in Finland.”12 Unlike other European democracies that survived the interwar period, Finland had relatively young electoral institutions.13 Its democratic republic was born in 1918, when the country emerged from a civil war pitting socialists and communists against conservatives. Since 1919, a number of national elections had nevertheless been held and a succession of parliaments and presidents had peacefully competed for, exercised, and then relinquished power. Two parties, the Social Democrats and the center-right Agrarian Union, dominated national politics. By 1929, Finland appeared to be moving from the right toward the center—but not to be on the cusp of any violent convulsion.14

The spontaneous clash in Lapua cascaded quickly into a right-wing mass movement that drew support from larger farmers, professionals, and academics. Its leaders, first selected in early 1930, included bankers, well-known industrialists, and the chief editor of a leading Agrarian newspaper. In the summer of 1930, twelve-thousand Lapua members mimicked Mussolini’s March on Rome with a “Peasant’s March” on Helsinki in which they called for a blanket proscription of communist organizations. Conservatives, including President Lauri Kristian Relander (1925–31), gave them a sympathetic hearing. In June 1930, the Agrarian-Progressive government of Prime Minister Kyösti Kallio banned all communist newspapers, despite lacking any clear legal authority to do so.15 Communist deputies in the Eduskunta, the national legislature, were arrested on treason charges. Elections later in 1930 saw substantial gains for the right-wing National Coalition, and in the 1931 presidential contest the Lapua movement rallied around former prime minister Pehr Evind Svinhufvud. He ultimately prevailed in a closely divided Electoral College, but only after the leader of the civil guard warned the Agrarian Union’s leaders of violence should Svinhufvud lose at the polls. With these steps alone, Lapua seemed to have “gone far toward achieving a disintegration of the political system.”16

At the same time, the movement also embarked on a campaign of more generalized violence, kidnapping political opponents and dumping them on the other side of the Soviet border. Even Svinhufvud’s erst-while foe in the presidential election fell victim to this tactic. In February and March 1932, armed Lapua cadres flooded the small town of Mäntsälä, about sixty kilometers north of Helsinki, demanding a new “patriotic” government. A former army chief joined them. With both ordinary politicians and elite institutions such as the Electoral College [End Page 23] seemingly looking down the barrel of a gun, Finland appeared to be on the cusp of the sort of democratic erosion that was to engulf Germany and Austria soon thereafter.

Yet Finnish democracy prevailed, for two reasons. First, unelected actors stepped up in democracy’s defense. Key military personnel, and in particular the army’s commander-in-chief Aarne Sihvo, opposed the moves that might have led to erosion. While some individual members of the armed services joined the 1932 rebellion, the leadership of the civil guard did not lend its support. Judges, too, meted out harsh penalties for Lapua’s use of violence. The Mäntsälä uprising fizzled quickly when the conservative Svinhufvud declared a state of emergency, demanded the arrest of Lapua’s leaders, and made a personal appeal by radio urging the movement’s rank-and-file members to disperse and renounce arms.17 Leaders of his Conservative party also increasingly saw Lapua as more of a threat than an asset.18

Second, starting in 1930, an ideologically diverse array of political parties (including the Agrarian Union and the Social Democrats) formed a “lawfulness front” to explicitly distance themselves from the Lapua movement. Crucially, conservative politicians in Finland did not make the same choices as their Italian and German contemporaries. They chose to marginalize and limit Lapua, rather than letting it coopt and neutralize them. To be sure, this led to fissures within the Agrarian Union, which saw schisms and the emergence of new far-right breakaway parties.19 These formations, however, never gained the mass footing that Lapua had briefly achieved. By March 1937, a new center-left coalition of Progressive, Social Democrat, and Agrarian interests was securely in power, and the threat to Finnish democracy had receded.

Colombia 2010: President Alvaro Uribe was wildly popular in his country while in office (2002–10) and he remains so to this day, in large part for his successful prosecution of the long-running civil war against groups such as the Revolutionary Armed Forces of Colombia (FARC). From the early days of his presidential career, however, Uribe harbored ambitions of extending his rule beyond its constitutionally mandated expiration date. The presidential term limits set out in the Colombian constitution were the immediate barrier to that ambition, as they have been in many other cases where a strong president has driven the erosion of democratic norms.20 The Constitution of 1991 had been carefully designed around a single-term presidency to ensure rotation in office, a particular concern in Colombia since the president has the power to appoint other officeholders. During Uribe’s time in power, conflict over those restrictions took center stage in the struggle that would decide the future of Colombia’s democracy. Perhaps because the barrier to democratic erosion was a constitutional one (as opposed to a political [End Page 24] or social impediment), this contestation was ultimately resolved by the nation’s Constitutional Court.

Flush with victories against insurgent groups, Uribe first successfully pushed Congress to amend the constitution’s term-limit provisions in 2004. A challenge to that amendment before the Constitutional Court did not succeed. But the Court’s decision did include some cautionary language suggesting that there might be limits to the degree of constitutional change that it would accept. After he won reelection, Uribe consolidated power further. A major scandal erupted in 2006, when investigators discovered a computer belonging to a paramilitary leader that indicated the latter’s close relationships with various politicians. This “parapolitics scandal” implicated many members of the Colombian Congress. Leveraging the crisis, Uribe proposed a novel “justice-reform bill” that would have removed cases involving politicians to a new special court. Ultimately, he failed to secure sufficient support to pass this measure. Nevertheless, other indicators of democratic erosion emerged. A member of Congress was found to have taken bribes to support the constitutional amendment allowing Uribe’s reelection. Supreme Court deliberations were wiretapped, and the Court as well as journalists became targets of a campaign of harassment. Moreover, the sheer fact of Uribe’s lengthened tenure meant that he was able to make more appointments to the courts and the elections body.21

Six years after first amending the constitution, Uribe sought a new amendment that would increase the permitted number of presidential terms from two to three. This time, the Constitutional Court balked. Even though four out of its nine members had been appointed by Uribe, the Court rejected on procedural grounds a plan to hold a referendum on the proposed amendment. Its February 2010 ruling held that a second extension of the presidential term would effectively be an unconstitutional replacement of the constitutional scheme as a whole (a “substitution”), since it would allow the president to “name members of the central bank, the attorney general, the ombudsman, the chief prosecutor, and many members of the Constitutional Court.”22 The Court also expressed concern about media dominance by a three-term president. In other words, the Court found that a third presidential term would represent a threat to Colombia’s democracy. This second decision arguably forestalled “a significant erosion of democracy by preventing a strong president from holding onto power indefinitely.”23 Twelve consecutive years in power would have vested the president with tremendous power over various institutions of state, not least those charged with checking him.

Probably because the Court itself was a very popular and respected institution, and also because Uribe thought he could install a placeholder president as his successor, the president accepted the Court’s decision and withdrew his candidacy for the 2010 presidential race. His defense [End Page 25] minister, Juan Manuel Santos, ran in his stead. Once in office, however, Santos distanced himself from Uribe—engineering, for instance, a reversal of the constitutional amendment allowing two presidential terms (though only after his own reelection in 2014).

The Colombian case illustrates the potential importance of the national judiciary in checking democratic erosion. Both the Constitutional Court and the Supreme Court—which oversaw an aggressive prosecution of the parapolitics scandal—served as important venues for maintaining accountability. The first preserved electoral accountability and forestalled Uribe’s capture of state power; the second inhibited the use of corruption to stifle other aspects of democratic competition, and gave standing legal rules new vigor. Judicial interventions created a measure of breathing space for politics to “recover” from its brush with back-sliding, and thus made it possible for Santos to retreat from Uribe’s autocratic style and tactics. It is important, however, to recognize the element of contingency at work in these events. Uribe, who may have considered noncompliance, ultimately respected the 2010 judicial decision, given the Court’s status in Colombia and a belief that he could work around the ruling. Further, Santos’s willingness to “defect” and restore the institutional status quo ante was not foreordained, let alone specified in any law. Prudent choices by specific elite politicians, as in the Finnish case, likely played a decisive role in the outcome.

The Colombian case, like the Finnish one, thus underscores the importance of elite actors within nonmajoritarian institutions (the military or the judiciary) and within established political parties. To the extent that popular sentiment or even electoral outcomes played a role, their effect was distinctly secondary. But there is some countervailing evidence, again from Latin America, that both elite dynamics and mass politics (as crystallized in electoral outcomes) can matter. In Bolivia, at roughly the same time that Uribe was struggling to retain power, President Evo Morales pushed through a new constitution, and he is now on his way to what may become a fourth term in office. Somewhat similarly, Ecuadorean president Rafael Correa persuaded his nation’s legislature in December 2015 to lift term limits, reassuring legislators with the promise that he would not run again immediately. Yet after Correa left office in May 2017, his handpicked successor, Lenín Moreno, moved to put political distance between himself and Correa. He pressed for a referendum to restore the repealed term limits—a measure that passed in February 2018, effectively precluding Correa’s return to power, at least for the time being. Elite politics in Ecuador hence interacted with mass politics in a manner that shored up democratic stability, yielding an outcome that resembles the fate of Colombia after Uribe more than recent events in Venezuela or Bolivia.

Sri Lanka 2015: Sri Lanka illustrates a situation in which democracy’s erosion had progressed further than in the other two cases. Since [End Page 26] the early 1980s, Sri Lanka has been classified by Freedom House as only Partly Free, even as it has consistently held elections and gone through several turnovers of power. Sri Lanka’s flawed democracy came under severe threat with the rise of Mahinda Rajapaksa, who was first elected to the presidency in 2005. His win at the polls was alleged to have involved collusion with the Tamil Tigers (LTTE)—an armed separatist group drawing its support from elements of Sri Lanka’s Tamil ethnic minority—to suppress votes in the northeast of the country. Upon taking office, however, Rajapaska restarted the longstanding civil war with the LTTE and then prosecuted it with new ferocity; in 2009, he declared an unqualified victory. He won a second presidential term in January 2010 by defeating his former commander-in-chief, General Sarath Fonseka—whom he subsequently had imprisoned, taking his most potent electoral rival out of play.

Throughout both his first and second terms, Rajapaksa’s rule was marked by nepotism, corruption, and a degradation of rule-of-law institutions such as courts, prosecutors, and the police. Rajapaksa used his family to seize the levers of economic power, appointing his three brothers to key cabinet positions, and presided over a period distinguished by both rapid economic growth and deepening corruption scandals. As Rajapaksa increasingly promoted his own cult of personality, independent-minded journalists came under severe attack; many were killed or jailed. After having the constitution amended to allow him to run for a third term, Rajapaksa called a snap election for January 2015, with a campaign period of only six weeks. Sri Lanka seemed on the brink of seeing its democracy totally degraded.24

Rajapaksa’s effort to consolidate power went off the rails only when Maithripala Sirisena, the former minister of health in Rajapaksa’s government, split with the president and declared his own candidacy a day after the snap-election announcement. Bolstered by an electoral coalition including disaffected voters from the Sinhalese ethnic majority, Tamils, and members of small Muslim minority groups, Sirisena won at the polls with the assistance of almost fifty political parties and civil society groups that united around a common opposition platform. As Neil DeVotta has documented in these pages, Rajapaksa initially considered declaring a state of emergency and annulling the vote.25 He was stopped from doing so, however, by resistance from army and police leaders, as well as opposition from the attorney-general. Moreover, the head of the Department of Elections, Mahinda Deshapriya, successfully oversaw the polling that did occur, prevented violence and intimidation, and thereby maintained what we have called the “administrative rule of law” in respect to the 2015 polls.26

Upon his election, Sirisena immediately began undoing some of the measures introduced by Rajapaska that had undercut democratic and rule-of-law institutions. Parliamentary prerogatives were restored. [End Page 27] A new constitutional amendment was adopted, reversing a prior one that had envisaged allowing three consecutive presidential terms. To be sure, Sirisena’s repudiation of authoritarianism was not completely effective: As of this writing in September 2018, Rajapaksa remains in Parliament and is actively seeking a comeback. Yet even if he succeeds in regaining high office, the measures adopted since his previous term have made it less likely that he will once again be able to take over the entire system.

Sri Lanka’s comeback from the brink underscores the themes that we have already noted in the Finnish and the Colombian cases: Although there was a role for popular sentiment (here, the willingness of the Tamil and Muslim populations to turn out at the polls, and of some Sinhalese to cross over and support Sirisena), the conditions under which it could matter came about only due to the decisions of political elites and nonelected institutions. As in the Finnish case, the choices made by key leaders within the armed services were apparently crucial in turning events down the path they finally took in 2015. The Sri Lankan experience also shows the importance of neutral electoral machinery staffed by officials able to resist political capture: In the face of partisan pressures, autonomous electoral administrators grounded in professional norms can be key to preventing fraud. Sri Lanka’s democracy, like those of Finland and Colombia, thus survived because a large number of unelected elite actors committed themselves to democracy.

And again, as in the Finnish case, a broad coalition bringing together erstwhile political opponents proved key to resisting democratic erosion. In this vein, the Sri Lankan experience highlights the importance of intraparty competition. Sirisena was for quite some time an ally, and likely a beneficiary, of the Rajapaksa political machine. Yet he defected at the crucial moment in an act of undoubted personal courage. Even though his candidacy at first seemed to have virtually no chance of success, he took on the long odds—and won. The potential significance of elite defection, then, is another lesson of the Sri Lankan case. At the same time, a note of caution is warranted: Such a move might well have become impossible had Sirisena waited until after Rajapaksa’s third term, by which time the overbearing leader might well have been able to bring under his control a coercive machinery so extensive and so deeply rooted that no one within his party could have dared launch a challenge. Timing, as well as courage, matters.

There is no single “magic institution” that can be adopted to prevent [End Page 28] democratic backsliding or to arrest it once it has begun. Instead, the case studies suggest some broader principles and dynamics worth keeping in mind. Paradoxically, the experiences of democratic near misses that we have explored underscore the role of political elites and nonelected institutions—courts, military commanders, and election administrators—in decisively repudiating authoritarian leaders bent on democratic erosion.

A democracy under threat depends critically on support from unelected and nonmajoritarian actors. Such support serves to slow down erosion, giving political parties and public movements time to regroup and reorganize in the face of threats. Of course, no institution is foolproof, and much will depend on contingent choices made by political actors—including mainstream forces that may decide to ally with antisystem actors for tactical reasons. Sustained antidemocratic mobilization is hard to defeat, but a well-timed decision by judges, generals, civil servants, or party elites can make all the difference between a near miss and a fatal blow.



1. Steven Levitsky and Daniel Ziblatt, How Democracies Die (New York: Crown, 2018).

2. For differing perspectives, see Tom Ginsburg and Aziz Huq, “Defining and Tracking the Trajectory of Liberal Constitutional Democracy,” in Mark Graber et al., eds., Constitutional Democracy in Crisis?(New York: Oxford University Press, 2018), and Valeriya Mechkova, Anna Lührmann, and Staffan I. Lindberg, “How Much Democratic Backsliding?” Journal of Democracy 28 (October 2017): 162–69.

3. See, inter alia, János Kornai, “Hungary’s U-Turn: Retreating from Democracy,” Journal of Democracy 26 (July 2015): 34–48; Javier Corrales, “The Authoritarian Resurgence: Autocratic Legalism in Venezuela,” Journal of Democracy 26 (April 2015): 37–51.

4. For an excellent example, see Agnes Cornell, Jørgen Møller, and Svend-Erik Skaaning, “The Real Lessons of the Interwar Years,” Journal of Democracy 28 (July 2017): 14–28.

5. Quoted in Dennis Merrill, “The Truman Doctrine: Containing Communism and Modernity,” Presidential Studies Quarterly 36 (March 2006): 27.

6. Aziz Huq and Tom Ginsburg, “How to Lose a Constitutional Democracy,” UCLA Law Review 65 (February 2018): 78–165.

7. Zhaotian Luo and Adam Przeworski, “Subversion by Stealth: Elementary Dynamics of Democratic Backsliding,” (unpublished manuscript, New York University Department of Politics, 27 March 2018); Adam Przeworski, Crises of Democracy (forthcoming from Cambridge University Press).

8. Ozan Varol, The Democratic Coup d’État (New York: Oxford University Press, 2017).

9. Dan Slater, “Democratic Careening,” World Politics 65 (October 2013): 729–63.

10. Examples include Cornell, Møller, and Skaaning, “Real Lessons”; Frank Aarebrot and Sten Berglund, “Statehood, Secularization, Cooptation: Explaining Democratic Survival in Inter-War Europe—Stein Rokkan’s Conceptual Map Revisited,” Historical Social Research/Historische Sozialforschung 20, no. 2 (1995): 210–25.

11. Economic Freedom Fighters and Others v Speaker of the National Assembly and Another, 47 (ZACC, 2017),; “South Africa’s Public Protector Finds ‘State Capture’ by the President’s Pals,” Economist, 5 November 2016,

12. Marvin Rintala, Three Generations: The Extreme Right Wing in Finnish Politics (Bloomington: Indiana University Press, 1962), 164–99.

13. Cornell and colleagues underscore the age of interwar democracies as a predictor of their ultimate stability (“Real Lessons,” 26). Finland is, as they concede, an exception.

14. Alan Siaroff, “Democratic Breakdown and Democratic Stability: A Comparison of Interwar Estonia and Finland,” Canadian Journal of Political Science 32 (March 1999): 103–24.

15. Giovanni Capoccia, “Defending Democracy: Reactions to Political Extremism in Inter-War Europe,” European Journal of Political Research 39 (June 2001): 431–60.

16. Risto Alapuro and Erik Allardt, “The Lapua Movement: The Threat of Rightist Takeover in Finland, 1930–32,” in Juan J. Linz and Alfred Stepan, eds., The Breakdown of Democratic Regimes: Europe (Baltimore: Johns Hopkins University Press, 1978).

17. Rintala, 162.

18. Alapuro and Allardt, 133.

19. Siaroff, 120.

20. Tom Ginsburg, James Melton, and Zachary Elkins, “On the Evasion of Executive Term Limits,” William and Mary Law Review 52 (May 2011): 1807–72.

21. Harvey F. Kline, Fighting Monsters in the Abyss: The Second Administration of Colombian President Álvaro Uribe Vélez, 2006–2010 (Tuscaloosa: University of Alabama Press, 2015).

22. Rosalind Dixon and Samuel Issacharoff, “Living to Fight Another Day: Judicial Deferral in Defense of Democracy,” Wisconsin Law Review 2016 (November 2016), 718.

23. David Landau, “Abusive Constitutionalism,” UC Davis Law Review 47 (November 2013): 203.

24. Neil DeVotta, “Sri Lanka: From Turmoil to Dynasty,” Journal of Democracy 22 (April 2011): 130–44.

25. Neil DeVotta, “A Win for Democracy in Sri Lanka,” Journal of Democracy 27 (January 2016): 152–66.

26. Huq and Ginsburg, “How to Lose a Constitutional Democracy,” 83.


Copyright © 2018 National Endowment for Democracy and Johns Hopkins University Press

Issue Date: 
October 2018
Page Numbers: 

Subject Headings: