Debate: Why Democracies Survive

Issue Date October 2022
Volume 33
Issue 4
Page Numbers 133–49
file Print
arrow-down-thin Download from Project MUSE
external View Citation

Experts worr­­y that de facto single-person regimes in previous multiparty states (Russia, Turkey, Venezuela) and norm-defiance in existing democracies (Brazil, Hungary, the United States) signal a coming authoritarian age. Without examining the broader record, however, it is hard to know whether such tremors presage a global convulsion. A century’s worth of evidence (1920–2019) shows that contemporary democracies are sturdier than they look. Above all, high levels of economic development continue to sustain multipartism; OECD democracies have faced less risk than often intimated. Further, competition among political parties, regardless of national affluence, contains a momentum that even the most willful demagogues have had trouble stopping. These economic and institutional bulwarks help explain why democratic backsliding, which seems so portentous, has preceded democratic survival more often than breakdown. Even as executive aggrandizement and rancorous partisanship roil the world’s most venerable democracies, they are unlikely to produce new autocracies absent permissive material conditions.

Exactly a hundred years ago, in October 1922, Benito Mussolini began twisting Italy’s democracy into an icon of authoritarianism. The March on Rome that he and his Fascist Party staged that month convinced the king of Italy—fearing civil war—to hand the government to Mussolini, sending reverberations beyond Southern Europe. An early one was felt just a year later, when Adolf Hitler and his Nazi Party tried to seize control of Bavaria and start a “German revolution” against the Weimar Republic. Their bid failed, but the resulting trial made Hitler a national figure in German politics.

Over the next two decades, what Samuel P. Huntington would later call the “first reverse wave” of authoritarianism would see nearly two-thirds of the world’s democracies placed under the bootheels of dictators and occupying armies.1 The price of liberating at least some of these countries was a cataclysmic six-year conflict that killed more than seventy-million people, most of them civilians, and ended with the use of atomic weapons. For later generations, the titanic horror of the Second World War cast a grim light on the era leading up to it, when Mussolini, Hitler, and their followers had risen to power. The first reverse wave loomed as a dire testament to the ability of willful leaders to threaten freedom and peace.

About the Authors

Jason Brownlee

Jason Brownlee is professor of government at the University of Texas, Austin, and the author of Authoritarianism in an Age of Democratization (2007).

View all work by Jason Brownlee

Kenny Miao

Kenny Miao is a doctoral candidate in government at the University of Texas, Austin.

View all work by Kenny Miao

The specter of democratic collapse now haunts a new century. The 2010s began with 116 electoral democracies in the world, according to Freedom House, and ended with 115, a modest net decline that belied significant disruptions in a dozen earlier multiparty systems. Alarming numbers of elected leaders were using coercion to stay in power. A prime example was Turkish president Recep Tayyip Erdoğan, who exploited a clumsy July 2016 coup attempt to launch a sweeping “crackdown on the political opposition, academia, media and civil society.”2 Within two years, he would transform the once-ceremonial Turkish presidency into a potent executive office tailor-made for his brand of strongman rule. Other events of 2016 including the British vote to leave the European Union and the U.S. presidential election stunned professionals who made a living explaining world events.

Commentators and scholars soon placed Brexit, Donald J. Trump’s victory, and authoritarian leaders such as Erdoğan in a broader narrative of democratic decline.3 Two months after Trump was inaugurated as U.S. president, the title story in the Atlantic was “How to Build an Autocracy.” The cover depicted Trump addressing a throng that stretched beyond the horizon. The 14 May 2018 issue of Time magazine carried a lead story titled “Rise of the Strongman,” with a cover showing Russian president Vladimir Putin looming above images of Erdoğan, Hungarian premier Viktor Orbán, and Philippine president Rodrigo Duterte. A few months later, in October, the Atlantic asked readers “Is Democracy Dying?” Last December the magazine declared “The Bad Guys Are Winning,” with the illustration showing a black-suited Putin strutting, Reservoir Dogs–style, with Erdoğan as well as presidents Nicolás Maduro of Venezuela, Alyaksandr Lukashenka of Belarus, and Xi Jinping of China.

Talk of an antidemocratic Zeitgeist echoed in the halls of academe. Some scholars saw a “wave of autocratization” threatening to engulf the world’s most venerable polities.4 “If Trump wins this election in November,” said one political scientist in mid-2020, “democracy is gone.”5 Even after Trump lost, however, the menace to democracy did not disappear, as the former president and swaths of his base embraced the fiction that the election had been stolen and a usurper sat in the Oval Office.6

At the centennial of Mussolini’s putsch, it would be risible to ignore troubling parallels between the interwar epoch and our current era. Then and now, opportunists with no lasting fealty to democracy have weaponized elections to concentrate power in their own hands. Then and now, demagogues have stirred millions through naked appeals to prejudice and nativism. Then and now, office-seekers have celebrated the violence of the mob and the state to silence critics. No matter how much observers may disagree about the extent of these similarities, that the ugliest episodes of the 1920s and 1930s echo at all in the early twenty-first century is a cause for dismay and reflection.

Even as experts and laypersons reckon with these challenges, it is vital to note that historical analogies can mislead.7 Fascist Italy and Nazi Germany are notorious because they were terribly exceptional. Countless egotists have aspired to emulate the Duce and the Führer, but few have come close. The fates of Italy’s liberal Parliament and the Weimar Republic warn against complacency, but at the same time, the value of their examples for decisionmakers who must navigate today’s waters should not be overestimated.

The present essay takes seriously what was previously inconceivable: the prospect that, across the global North and South, a slew of democracies will fall in the twenty-first century. We step back from recent examples to consider an expanse of historical and contemporary evidence—including an analysis of democratic breakdown and survival between 1920 and 2019. We conclude that, while democracies today are certainly under stress, the prognosis is generally positive. Affluent democracies (in which national wealth derives from industry rather than unproductive rents such as may be charged for oil or mineral extraction) remain very likely to endure as democracies. Notwithstanding a small set of outliers such as Turkey, electoral democracy in wealthy states looks set to survive. This inference fits a storied research tradition regarding the link between socioeconomic modernization and the stability of democratic rule.8

A closer look at which democracies “backslid” in the early twenty-first century, and which cases broke down suggests that norm erosion, institutional gridlock, and other woes—while certainly troubling—are not portents of dictatorship. Democracy watchers should therefore keep their eyes on the socioeconomic conditions that have tamed the wild ambitions of would-be autocrats—or allowed their schemes to flourish.

Not With a Bang, but a Whimper

The 2000s and 2010s did not deliver a reverse wave of democratization, but the mode and setting of democratic breakdowns drew attention.9 Most recent democratic breakdowns traced back to politicians who had been elected to run the executive branch rather than to uniformed coup plotters. The mechanism of breakdown was death by a thousand cuts. Erdoğan also tightened his hold in a country where relative material prosperity was supposed to stymie authoritarianism.

From the 1950s through the 1980s, when democracies fell, odds were that generals or colonels were responsible. While traditional military coups continue to threaten civilian rule (as in the cases of Egypt 2013, Thailand 2014, Burma 2021, and Burkina Faso 2022), civilian-led takeovers have eclipsed them.10 Examples of other nonmilitary figures who have embedded themselves in power include Daniel Ortega, who is now in his fourth consecutive term as Nicaragua’s president, and Sheikh Hasina, who is enjoying her third successive term as the prime minister of Bangladesh.

While they have been behind many democratic breakdowns, civilian politicians have mostly eschewed overt autogolpes. The original “self-coup” artist was Louis-Napoléon Bonaparte, who in the early 1850s turned the French Second Republic into the French Second Empire, promoting himself from France’s president to its emperor in the process. The German Enabling Act of 23 March 1933, Ferdinand Marcos’s 1972 declaration of martial law in the Philippines, and Alberto Fujimori’s 1992 seizure of legislative and judicial power in Peru are other notable cases. One day citizens are enjoying multipartism and civil liberties; the next morning they awake under dictatorship. The autocratic premiers and presidents of the twenty-first century have generally steered clear of anything that jolting. Preferring to showcase continuity, they hold regular elections, keep legislatures open (even if manipulated and curbed), and permit at least some rival parties to operate.

Because the new authoritarianism arises in stages behind a republican frontispiece, the emergence of rule by a single party or person can be gradual enough to have a “dangerously deceptive” quality.11 The incumbent plays coy while spending years gradually shrinking the space for political competition. Observers both foreign and domestic may not reach consensus about what is happening until it is too late.12 Democratic decline is not unidirectional, but the downward slope of backsliding can be both slippery and steep. One study found that more than two-thirds of democracies that backslid eventually underwent full breakdown.13 Spotting such patterns in the past is far easier than tracking them in real time. Ortega bucked term limits to win reelection in 2011, with 62 percent of the vote, in a controversial but hotly contested race.14 He then garnered implausibly high supermajorities in later polls. Erdoğan pulled off a slow-motion executive takeover in a country where economic development was supposed to have advanced too far to let democracy be unraveled in this way.

Political scientists have long recognized a powerful correlation between material development and democratic survival. “The more well-to-do a nation,” noted Seymour Martin Lipset in his seminal 1959 essay on the subject, “the greater the chances that it will sustain democracy.”15 Four decades later, Lipset’s idea was validated in the cross-national work of Adam Przeworski and Fernando Limongi, who wrote that “no democracy has ever been subverted, not during the period we studied nor ever before nor after . . . in a country with per capita income higher than that of Argentina in 1975.” Setting aside major fuel exporters (where rentier wealth has effects not found elsewhere), their analysis left “no doubt that democracy is stable in affluent countries.”16 Turkey stood as a glaring exception. By somewhere around 2016 to 2018, depending on whom you ask, the Turkish government no longer met the minimum procedural criteria for electoral democracy. In Purchasing Power Parity (PPP) terms, Turkey’s GDP per capita (GDPpc) makes it 70 percent richer than Argentina was in 1975.17

The Figure plots democratic breakdowns and GDPpc over time from the year 2000 through 2019, distinguishing between executive takeovers and executive removals (coups).18

As of this writing in September 2022, democracies above the historical “Argentina 1975” threshold continue to appear safe from military coups. (The May 2014 coup in Thailand came close to defying the trend.)19 Meanwhile, the stability that affluence was supposed to bestow did not extend to blocking takeovers by civilian chief executives.20 Even bracketed as an “outlier,” the case of Turkey, and an incident of breakdown in the small nation of the Maldives, still mean that there can be no deterministic relationship between development and democratic survival. At the same time, however, a potent example—or even a handful of such instances—does not constitute a trend. From 2000 through 2019, democratic breakdowns still clustered in the bottom quarter of the PPP GDPpc range (73 percent of them happened below $6,602, which is less than half the Argentina 1975 threshold). In the absence of a more methodical examination, it is difficult to know whether Turkey is an outlier or a bellwether.

Midrange studies underline the importance of not overgeneralizing from high-profile cases and potentially misreading the threat to democracy. Kurt Weyland found that among thirty “populist” leaders in the Americas and Europe, only six had “suffocated democracy.”21 When Robert Kaufman and Stephan Haggard studied the problem, they even set aside wealthy democracies, acknowledging that “outright reversions [to authoritarianism] are still virtually non-existent in developed countries.”22

Conversely, as illiberal politics spreads—but few democracies succumb—multiparty governments may prove more likely to recover from backsliding than plunge into autocracy. Just four years after the Atlantic mused about Trump building an undemocratic regime and Time placed Duterte in the company of dictators, both men were out of office. Both figures still commanded influence in national politics. Trump denied his defeat and raised expectations for a fierce run at the White House in 2024. Duterte’s daughter Sara was elected vice-president of the Philippines in 2022, while another prominent scion—Ferdinand Marcos, Jr.—won the presidency with nearly 60 percent of the vote. Nonetheless, neither Trump nor Duterte had exceeded his mandate, and the respective elections that replaced them evinced robust multipartism.

It is not hard to imagine prosaic outcomes for other elected executives who have flouted democratic norms: Brazil’s Jair Bolsonaro, Poland’s Jarosław Kaczyński, India’s Narendra Modi, and Hungary’s Viktor Orbán.23 These individuals may harbor fantasies of emulating Erdoğan, but such fearsome schemes remain circumscribed by the very political systems they seek to corrupt.

The Empirics of Democratic Breakdown

Is Erdoğan’s Turkey a black swan or a canary in a coal mine? Comparative study may help us to find out. Only then can we reliably assess the risks posed by elected leaders who may harbor ambitions similar to Erdoğan’s, but who find themselves in different contexts. Our full dataset (5,563 country-years and 235 spells from 1920 through 2019) allows us to test the degree to which economic development protected democracy during times of creeping authoritarianism. Our key findings come from our linear-probability model. (This type of model is among the most widely used in the social sciences; for our main regression table, see the online appendix at As we ran the model, certain patterns emerged repeatedly. Reported results are not idiosyncratic; they denote enduring trends. We tested a range of competing hypotheses, and we used multiple measures of democracy to make sure that no findings depended on judgments made by a single coder or team.

The topline result is that patterns of democracy in the early twenty-first century show as much continuity as change—and they offer as much evidence for optimism as cause for concern. First, productive wealth (nonrent GDPpc) stands as a strong predictor of democratic breakdown and survival. For the average electoral democracy of the early twenty-first century, the compounded predicted probability of surviving (avoiding breakdown) for a generation (twenty years) rises as GDPpc goes up. At a GDPpc level of $1,000 per year, democracy has a slightly better than one-in-three chance of surviving. That probability improves to 60 percent at $5,000 GDPpc, reaches 75 percent when it is $10,000 a year, and enters the realm of virtual certainty (97 percent) at $20,000.

The tests also showed how the perils of the time between the two world wars compare with the current era. The data underscored just how fraught were the 1920s and 1930s. The overall risk of breakdown (whether by army coup or civilian takeover) during the interwar period was a statistically significant 2 percent higher per year, which translates, over a twenty-year span, into a 50 percent greater chance of a democracy ending at some point during that two-decade interval. All later eras, including the opening decades of the present century, have been more conducive to democratic survival. Such a claim may sound puzzling, given the rise in executive takeovers, but the data offer an explanation.

Democratic breakdown did not become more likely in the 2000s and 2010s, but its modality morphed. The annual likelihood of a military coup dropped nearly 1 percent (0.9 percent), a slight but statistically significant change. Meanwhile executive takeovers became 0.6 percent more probable. For the most part, these countervailing patterns washed each other out. Yet the greater peril that elected incumbents posed to democracies did not quite match the increased security those same governments enjoyed from coups. On balance, an electoral democracy has annually been 0.4 percent safer, on average, since 2000.

In summary, our analysis found substantial evidence of political continuities between the recent era and the postwar twentieth century. Development continued to be a powerful predictor of a country’s ability to sustain democracy. Conversely, to the extent that new hazards for democracy have emerged, their net impact on democratic survival has been nil. To understand how democracy could have remained relatively sturdy while democratic backsliding proliferated, we complemented our quantitative analysis with qualitative attention to the latest cases.

Democratic backsliding is still an emerging concept in the social sciences. In her seminal contribution on the subject, Nancy Bermeo defined backsliding as “the state-led debilitation or elimination of any of the political institutions that sustain an existing democracy.”24 Subsequent researchers have been more successful at describing what backsliding looks like than they have been at establishing its relationship to regime changes, or even at agreeing on how to measure it.

Working with the same group of cases that we used in our long-range tests of democratic survival, we sought a measure of backsliding that would track incremental declines in the quality of democracy. To provide a continuous measure, we turned to data from the Varieties of Democracy (V-Dem) Institute. The V-Dem data have been cited to back up some of the most prominent claims about democratic fragility.25 By employing a V-Dem measure, we aimed to ensure that our analysis would not overlook instances of backsliding that a narrower method might miss.26 We then examined how the population of “backsliders” overlapped with cases of democratic survival and breakdown.27

Forty-seven electoral democracies (and fifty democratic spells) experienced breakdown or at least one period of backsliding under our measure (see the Table, which omits another 75 democratic countries that did not undergo breakdown or substantial backsliding during the two decades). By this count, countries with periods of backsliding slightly exceeded occurrences of democratic breakdown (30 to 28). What is remarkable, however, is how little the two phenomena coincided. In twenty cases (more than half of which were instances of executive takeover), democratic breakdown was not presaged by a backsliding period. By comparison, seven of the eight cases of the second column were executive takeovers, connoting the electoral road to single-person rule.

The last column solves the mystery of how multiple presidents and premiers can curtail local opposition but still not produce a global threat to democracy. As captivating it has been, the tale of backsliding preceding breakdown has been much less common than the story of backsliding—and recovery—within electoral democracies. Twenty-two countries had one or more democratic spells with backsliding but did not break down. This set of states accounts for more than two-thirds of the backsliders from 2000 through 2019. However portentous executive overreach and rancorous partisanship may have seemed at the time, they did not augur regime change.

Sources of Democratic Survival

Like most history buffs, students of democratic breakdown have been drawn to the most consequential figures and events. Confident conclusions also require studying the proverbial “dogs that did not bark,” the cases where the feared outcome failed to happen. For democracy this means considering periods of continuity, the “quiet times” which make no headlines but which are integral to any sound estimate of how often and under what conditions democracies end.

The preceding analysis supports three inferences: 1) Having national wealth that does not depend on oil rents strongly predicts democratic survival; 2) multipartism contains built-in defenses against authoritarianism; and 3) backsliding tends to roil electoral democracy without curtailing it. In sum, the main conditions that imperiled democracy after 1945 continue to pose the biggest risks, while the historic sources of democratic survival continue to offer the strongest safeguards.

The legacy of Lipset. The idea that higher national incomes sustain democratic governments remains one of the few lawlike regularities that comparativists have discovered in the political world. We uncovered hefty amounts of evidence supporting the insights of Lipset and his successors: For citizens of electoral democracies during the century after 1920, there was no more reliable guarantor of their political system than material prosperity.

These results indicate a robust probabilistic relationship between wealth and democratic survival. Electoral democracy has been nearly impregnable—although not invincible—in countries with GDPpc above the Argentina 1975 level. By comparison, the dynamics of lower-income states are less novel. In countries with less (non-oil) per capita wealth than Argentina had in 1975, democracy has been and still is highly susceptible to both coups and executive takeovers.

Hence, what sometimes appears to be a topsy-turvy dynamic on the surface displays profound continuities that accord with prior theory. The data strongly support the stabilizing effect that national wealth has on electoral democracy, even in the face of downturns in Moscow, Managua, and Dhaka. Meanwhile, the outlier, Erdoğan, has grabbed the spotlight but has proved a tough act to follow. In order to understand why, it is valuable to also consider the political dynamics that stop one party or person from monopolizing power.

The momentum of multipartism. For traditional coup-proofing, new democracies may have no better option than appeasing the military until civilian sovereignty becomes taken for granted. In the countries of postcolonial Africa, Asia, and Latin America, that process sometimes took generations, especially for fiscally constrained states that struggled to deliver essential goods while also placating the generals. When it comes to blocking executives from seizing total control, however, the situation is less bleak. Even the youngest electoral democracies have a potent defense mechanism: their own political pluralism, rooted in social constituencies and embodied by parties and candidates. This innate resistance to civilian autocracy is so fundamental that it often hides in plain sight.

The same context of multiparty competition that vested the chief executive with authority then becomes a built-in impediment to the incumbent’s ability to hold power indefinitely. Even a minimal procedural democracy will have leaders whose path to office was a competitive struggle for the people’s vote. (If no such struggle took place, then the regime is something other than a democracy.) A chief executive intent on installing authoritarianism must be not only willing but able to stop future electoral competition and stay in power even after a plurality of voters has soured on the executive’s rule.

Consolidating a single-party state can be a tall order in civilian-dominated political systems where large-scale state repression is off the table and voting remains the principal mechanism of assigning political power. In effect, the incumbent executive needs to build a coalition of elite and mass supporters large and potent enough to defeat the broad resistance that indefinite single-party rule is sure to provoke. The resisters will include not only rival parties, but also former backers of the incumbents who may once have cheered but who now balk at a full-on plunge into the authoritarian unknown.

The protean nature of “competitive authoritarianism” is a testament to the inertial power of multipartism. Christopher Carothers noted in 2018 that competitive authoritarianism tends to be a passing political situation—it moves toward pluralism or else autocracy—rather than a stable regime type.28 Elections can be hotly contested even on a skewed playing field, and incumbents will struggle to stop the ballot from mandating the ebb and flow of power to and from the various parties. At the time of Carothers’s essay, more than half the original thirty-five competitive authoritarian regimes had moved out of that category, into electoral democracy, while less than a fifth (including Belarus, Cambodia, Cameroon, Gabon, Nicaragua, and Russia) established de facto single-party rule under an individual dictator.

In sum, when civilian politicians keep the army in its barracks—eschewing mass armed repression and preventing outright coups—these politicians will likely have no choice but to keep tussling in elections and partisan politics, without one figure or faction ever achieving permanent predominance. The potential for multipartism to provide a self-sustaining equilibrium carries a related implication for concerns about democratic backsliding.

Backsliding without breakdown. It is prudent to take note of attacks on democratic rules and norms, even when the overall political system is not immediately endangered. Yet close attention to scattered cases of backsliding may suggest causal or predictive relationships that broader evidence does not support. A train of abuses may precede a democratic breakdown in a way that cannot be written off as pure coincidence, but the question of how much backsliding contributes to breakdown remains unanswered until one considers backsliding’s relationship to democratic survival as well.

The data revealed that the road from backsliding to breakdown may be less traveled than previously assumed. Whereas speculation about a democratic crisis has centered on illiberal elites and their loyal zealots, the statistics point to the continuing power of structural conditions to hem in those actors and uphold democracy. Today’s would-be electoral autocrats choose their strategies for cementing power, but they do not choose the circumstances in which they pursue those strategies. When antecedent conditions include a competitive struggle for the people’s vote, bolstered by a robust economy, incumbents will find themselves hard-pressed to turn incremental gains into indefinite mandates.

We would not assert that backsliding and breakdown are orthogonal, but the predictive relationship is so weak that students of democracy should exercise caution before treating signs of democratic deterioration as portents of dictatorship. As the Table indicated, backsliding has been a problem within electoral democracies, but it is not associated with regime breakdown. If the patterns we observed hold over the coming decades, it may become commonplace to see backsliding and forward progress in equal measure as governments oscillate along the spectrum of electoral democracy. If commentators want to opine about the lifespan of a democracy, they would do well to speak of backsliding as a generally nonfatal ailment, from which electoral democracies often recover. And they will find a country’s material status to be the clearest guide to its long-term prognosis.

An Agenda for Action

If recent observers have underrated economic development’s power to block dictatorship, it would be a mistake to hew to approaches that boosted democracy in the past but may not help much anymore. Since the end of the Cold War, the challenge facing multipartism has changed.

The predominant reaction to the political tumult of the 2010s has been to issue general warnings about power-hungry firebrands endangering democracies the world over. Vigilance is fine; it does not require alarmism. When discussions of a democratic crisis slip into lurid historical analogies and unempirical speculation, they cloud matters rather than clarify them.

A dozen established electoral democracies were interrupted in the 2010s, but close attention to the ways in which democracy broke down and the larger contexts of events shows that our planet is not in the midst of an interwar-style reverse wave of democratization, much less the rise of an “authoritarian international.”

Knowing the specifics of where and how democracy is endangered is crucial. The most likely countries to experience breakdown are low-income and often relatively young democracies in the global South. These countries have not featured prominently in popular coverage, perhaps because they lack geopolitical heft or because their instability seems less novel. But aiming attention and resources at places where breakdown is not likely will not help the many millions of people who live in places where it is. A discussion of the practical means for preventing democratic breakdown is beyond our present scope, but there are some general lessons to keep in mind.

Economic development is the big factor. As a predictor of democratic survival, it eclipses all other variables. The magic number is a GDPpc (apart from oil rents) of 16,374 constant PPP 2017 international dollars per year. Above that level, the odds of democratic survival become favorable, and no other variable can compete. Were South Korea as poor as Brazil, for example, then the chance of South Korean democracy surviving for the next twenty years would drop by 25 percent on that token alone. Conversely, if Zambia were as affluent as Brazil, the likelihood of Zambian electoral democracy lasting for at least a generation would go from barely better than a coin flip (52 percent) to a solid prospect that nearly any wagering person would take (87 percent).

The road to development can be bitterly long, unfortunately. If GDPpc grew at 3 percent per year (the mean growth rate from 2000 through 2019), it would take the median case of democratic breakdown (Bangladesh in 2013, the year before breakdown) 53.6 years to exceed the level of 1975 Argentina. This is a daunting time horizon, but at least the data show that national material gains—slowly made though they may be—do yield the surest guarantee that, once a people have achieved their republic, they can keep it.

While political economy’s tectonic plates grind away in the background, policymakers and practical political scientists will focus on the nearer term. They have long sought ways to strengthen democratic institutions and norms, and nothing in our analysis argues for abandoning that search. We do reason, however, that a country’s macroeconomic profile communicates important information for interpreting its internal politics.

When it comes to holding accountable state officials who flout the law, suppress public debate, or plunder the treasury, development level should not figure. But where a country stands in terms of material affluence—and the baseline risk of democratic breakdown—does signify how seriously such malfeasance jeopardizes an entire political system. In economically advanced multiparty states, the abuse of power by elected leaders and their subordinates has not historically posed the level of existential threat that equivalent behaviors have brought to developing democratic states. These robust patterns indicate that speculation about antidemocratic regime change in many countries may be misplaced even—and perhaps especially—when concerns about discrete political failures are justified.

As it happens, there are even strong signs of long-term coherence, despite recurring challenges, in medium- and low-income democracies. Take Ecuador, a country that experienced a coup in 1997 and another in 2000.29 It would be facile to call Ecuador’s democracy dictator-proof. In the first half of 2013, left-wing populist and thrice-elected president Rafael Correa appeared to be taking Ecuador “down an increasingly authoritarian path.”30 As his personal control over national institutions grew, Correa “looked like a sure bet to become president-for-life.”31 Yet prior politics suggested—and later events confirmed—that Correa would struggle at erecting a one-person autocracy. By 2017, the political and economic winds had turned against Correa; he had been forced to scupper his reelection plans and back an ostensible surrogate, Lenín Moreno. Moreno won the presidency in a runoff, then reversed much of Correa’s intended legacy and served only a single term. He did not run in the 2021 election, and in May of that year peacefully passed power to the democratically elected opposition candidate.32

Symptoms of backsliding may be just that, signs that the machinery of government and the mores of society are falling behind expectations and historic standards. These signs are not negligible problems that will vanish if left alone, but neither do they point to a collapse of multiparty politics. In Ecuador, the Philippines, and other periodically volatile regimes, some form of multiparty civilian-driven pluralism has generally cohered even when the times have seemed to favor despotism. This is because when aspiring autocrats try to concentrate power, the very systems of political competition that they threaten also impede them. In light of this, the best response to backsliding will not be high-level forecasts of democracy descending into autocracy, but more granular work that seeks to tune up the gears of the state and cultivate a greater openness to differences among citizens.

A Gallery of Isolated Rogues

Surprising events have left observers of politics grasping for answers. Recent literature on the weakening and collapse of democracy focuses on important cases where authoritarianism took hold (Nicaragua, Russia, Turkey, Venezuela) or illiberal leaders who challenged sacrosanct democratic norms and rules in their respective countries (Brazil, Hungary, India, the United States). These countries are few in number but heavy in implication. They have attracted close academic and media attention under the suspicion that they point to a global trend. Still, it has been difficult to gauge the utility of historical comparisons or the scope of today’s antidemocratic threats. Without complacency, students of democracy must be on guard as well against an evidence-resistant “tyrannophobia” that treats every bumptious and overweening executive as the next Mussolini.33

Our inquiry into democratic breakdown over a hundred-year span allows us to consider the most alarming recent setbacks for democracy, the calamitous fascist wave of the 1920s and 1930s, and numerous other events that have attracted less attention. Statistical tests on more than two-hundred democratic spells show that recent problems of executive takeover and democratic breakdown are unlikely to amount to anything like the authoritarian cascade that followed the First World War. Fortunately, in most of the world where conditions for democratic stability (especially productive national wealth) appeared strong twenty or thirty years ago, they continue to bode well. Conversely, where these conditions have been scarce, countries have not miraculously solidified their democratic institutions in the historical equivalent of a fortnight. Both types of democratic breakdown can be expected to continue to bedevil states as impoverished as Haiti, Nepal, and Niger.

At the other end of the economic-development continuum, high (non-oil) GDPpc remains the single most consistent predictor of survival in electoral democracies. The case of Turkey notwithstanding, high levels of development continue to provide one of the strongest defenses against autocracy. In wealthy countries not dependent on oil rents, even the most power-hungry leaders are likely to hit obstacles if they try to smother dissent.

More often than not, mass electoral contests have proved a bulwark against would-be tyrants. Multipartism forces civilian politicians to vie for and share power, without the armed forces picking winners. Even when that system suffers from ills such as polarization and majoritarianism, it stands apart from the small number of erstwhile democracies that are now led by executives who have managed to install themselves for life.

Worrying about civil-liberties violations and electoral malfeasance remains reasonable: Alternation in power does not prevent incumbents and oppositionists from flouting norms and institutions, sometimes with force. Elected leaders are likely to continue challenging the soft guardrails of norms and the hard limits of the law. They will not pursue such schemes in self-selected circumstances. In this respect, their fate lies as much in their stars as in themselves.

If every power-hungry leader could mimic Putin or Erdoğan, today’s gallery of autocratic rogues would fill many magazine covers. Instead, these figures stand not as “vanguards of a global trend,” but as an isolated batch.34 While no democracy is invincible, the countries that have been practicing some form of competitive electoral politics are well positioned to preserve that tradition.



1. Samuel P. Huntington, The Third Wave: Democratization in the Late Twentieth Century (Norman: Oklahoma University Press, 1991), 14–15.

2. Zafer Yılmaz and Bryan S. Turner. “Turkey’s Deepening Authoritarianism and the Fall of Electoral Democracy,” British Journal of Middle East Studies 46 (December 2019): 691.

3. Aziz Huq and Tom Ginsburg, “How to Lose a Constitutional Democracy,” UCLA Law Review 65 (February 2018): 78–169; Adam Przeworski, Crises of Democracy (New York: Cambridge University Press, 2019); Larry Diamond, Ill Winds: Saving Democracy from Russian Rage, Chinese Ambition, and American Complacency (New York: Penguin, 2019); Timothy Snyder, The Road to Unfreedom: Russia, Europe, America (New York: Tim Duggan, 2018); Steven Levitsky and Daniel Ziblatt, How Democracies Die (New York: Broadway Books, 2018); Ozan O. Varol, “Stealth Authoritarianism,” Iowa Law Review 100 (May 2015): 1673–1742; Emily Holland and Hadas Aron, “We Don’t Know How Democracies Die,” London School of Economics Phelan U.S. Centre blog, 8 February 2018; Steven Levitsky and Lucan Ahmad Way, “The Myth of Democratic Recession,” Journal of Democracy 26 (January 2015): 45–58; Chris Maisano, “Democracy’s Morbid Symptoms,” Catalyst: A Journal of Theory and Strategy 3 (Summer 2019): 181–207; Daniel Treisman, “Is Democracy Really in Danger? The Picture Is Not as Dire as You Think,” Washington Post, 19 June 2018.

4. Anna Lührmann and Staffan I. Lindberg, “A Third Wave of Autocratization Is Here: What Is New About It?” Democratization 26 (October 2019):1095–1113.

5. Christopher Ingraham, “The United States Is Backsliding into Autocracy Under Trump, Scholars Warn,” Washington Post, 18 September 2020.

6. Barton Gellman, “Trump’s Next Coup Has Already Begun,” Atlantic, 6 December 2021.

7. Dylan Matthews, “Is Trump a Fascist? 8 Experts Weigh In,” Vox, 23 October 2020.

8. Seymour Martin Lipset, “Some Social Requisites of Democracy: Economic Development and Political Legitimacy,” American Political Science Review 53 (March 1959): 69–105; Adam Przeworski et al., Democracy and Development: Political Institutions and Well-Being in the World, 1950–1990(New York: Cambridge University Press, 2000).

9. In the 1920s and 1930s, nearly two-thirds of the world’s existing democracies collapsed; between 1960 and 1975, over a third of electoral democracies broke down. Huntington, Third Wave, 14–21.

10. Levitsky and Ziblatt, How Democracies Die; Milan W. Svolik, “When Polarization Trumps Civic Virtue: Partisan Conflict and the Subversion of Democracy by Incumbents,” Quarterly Journal of Political Science 15 (January 2020): 3–31.

11. Levitsky and Ziblatt, How Democracies Die,5.

12. Nancy Bermeo, “On Democratic Backsliding,” Journal of Democracy 27 (January 2016): 5–19.

13. Vanessa Alexandra Boese et al., “Deterring Dictatorship: Explaining Democratic Resilience Since 1900,” V-Dem Working Paper 101 (May 2020): 12.

14. “The November 2011 Elections in Nicaragua: A Study Mission Report,” Carter Center,

15. Lipset, “Some Social Requisites of Democracy,” 75.

16. Adam Przeworski and Fernando Limongi, “Modernization: Theories and Facts,” World Politics 49 (January 1997): 165.

17. In constant 2010 U.S. dollars, the GDPpc of Turkey in 2016 was 23 percent larger than that of Argentina in 1975. All economic data come from the Gapminder project, which compiles information from multiple sources, including the World Bank, the Maddison Project, and Penn World Tables,

18. This analysis includes democratic spells of at least three continuous years. Periods of democracy and incidents of breakdown come from the Lexical Index of Electoral Democracy (LIED), version 5.3. See Svend-Erik Skaaning, John Gerring, and Henrikas Bartusevičius, “A Lexical Index of Electoral Democracy,” Comparative Political Studies 48 (October 2015): 1491–1525. The LIED sorts democratic breakdowns into four categories: gradual regressions induced by incumbents, coups, foreign occupations, and self-coups (incumbents close parliament and take full control). We coded incumbent-led self-coup and gradual regression as “executive takeover” and coup and foreign occupation as “executive removal.” The general pattern did not substantially change under slightly different codings from a range of leading democracy datasets.

19. The Maldives skirts the line between executive takeover and executive removal. In February 2012, opposition forces irregularly deposed President Mohamed Nasheed amid a power struggle that had already ruptured the country’s short-lived electoral democracy. See Fathima Musthaq, “Shifting Tides in South Asia: Tumult in the Maldives,” Journal of Democracy 25 (April 2014): 166.

20. Milan W. Svolik, “Which Democracies Will Last? Coups, Incumbent Takeovers, and the Dynamic of Democratic Consolidation,” British Journal of Political Science 45 (October 2015): 725–27.

21. Kurt Weyland, “Populism’s Threat to Democracy: Comparative Lessons for the United States,” Perspectives on Politics 18 (June 2020): 397–98.

22. Robert R. Kaufman and Stephan Haggard, “Democratic Decline in the United States: What Can We Learn from Middle-Income Backsliding?” Perspectives on Politics 17 (June 2019): 417.

23. Orbán has left no doubt that he sees Russia and Turkey as lodestars, but only one major research team has labeled Hungary an “electoral autocracy.”See “Full Text of Viktor Orbán’s Speech at Băile Tuşnad (Tushnádfürdő) of 26 July 2014,” Budapest Beacon,29 July 2014; V-Dem Institute, “Autocratization Turns Viral: Democracy Report 2021,” 22–23.

24. Bermeo, “On Democratic Backsliding,” 5.

25. Kaufman and Haggard, “Democratic Decline in the United States”; Vanessa Alexandra Boesse et al., “How Democracies Prevail: Democratic Resilience as a Two-Stage Process,” Democratization 28 (July 2021): 885–907; Max Fisher, “Is the World Really Falling Apart, or Does It Just Feel That Way?” New York Times, 12 July 2022.

26. Moreover, V-Dem’s “electoral democracy index” (v2x_polyarchy) offers a continuous variable (ranging from 0 to 1) that takes in electoral fairness, freedoms of the press and association, and other basics that establish regime type. See Michael Coppedge et al., “V-Dem Codebook v11.1,” V-Dem Project, 43.

27. Looking at all countries from 2000 through 2019, we identified “backsliding periods” by looking for a substantial (0.1) net decline in the v2x_polyarchy score over some portion of that period—a method based on Kaufman and Haggard, “Democratic Decline in the United States,” 418. This analysis covered countries with democratic spells of at least three continuous years. Use of a lower threshold (0.08) yielded one additional case of a breakdown preceded by a backsliding period (Zambia 2016), and four additional cases of survival preceded by backsliding periods (Argentina, the Czech Republic, Mexico, and Papua New Guinea).

28. Christopher Carothers, “The Surprising Instability of Competitive Authoritarianism,” Journal of Democracy 29 (October 2018): 129–35.

29. Dieter Nohlen and Simon Pachano, “Ecuador,” in Dieter Nohlen, ed., Elections in the Americas A Data Handbook: Volume II South America (Oxford: Oxford University Press, 2005), 368.

30. Carlos de la Torre, “Latin America’s Authoritarian Drift: Technocratic Populism in Ecuador,” Journal of Democracy 24 (July 2013): 35.

31. Carlos de la Torre, “Latin America’s Shifting Politics: Ecuador After Correa,” Journal of Democracy 29 (October 2018): 80.

32. John Polga-Hecimovichand Francisco Sánchez, “Latin America Erupts: Ecuador’s Return to the Past,” Journal of Democracy 32 (July 2021): 5–18.

33. Eric A. Posner and Adrian Vermeule, The Executive Unbound: After the Madisonian Republic (Oxford: Oxford University Press, 2011).

34. Fisher, “Is the World Really Falling Apart?”


Copyright © 2022 National Endowment for Democracy and Johns Hopkins University Press

Image Credit: Gguy/