In 1933, Adolf Hitler was appointed Chancellor of Germany through a legal electoral process. He had not seized power; he had been handed it, by conservative politicians who believed they could use him and control him. In the months that followed, the Weimar Republic's formal institutions were not destroyed by force. They were handed over, piece by piece, through legal mechanisms, by people who told themselves that each individual step was reasonable. The Reichstag Fire Decree suspended civil liberties. The Enabling Act transferred legislative power to the cabinet. Civil service laws excluded Jews from government positions. Each measure passed through existing legal channels. By the time the full nature of the transformation was undeniable, the institutional capacity to resist it had been dismantled.
This is the model that political scientists studying democratic collapse have returned to repeatedly over the past twenty years — not because it is the most common historical case, but because it captures the most dangerous contemporary pattern. For most of the mid-twentieth century, when political scientists studied how democracies ended, they studied coups: the general who announces on the radio that the government has been suspended, the tanks outside the presidential palace, the parliament dissolved by decree. Coups were dramatic, unmistakable, and clearly illegal. They could be resisted or reversed in ways that legal but systematically democracy-eroding policies could not.
What Steven Levitsky and Daniel Ziblatt documented in "How Democracies Die" (2018), synthesizing decades of comparative political research, was that the coup model had become largely obsolete in most parts of the world. Contemporary democratic erosion is gradual, incremental, technically legal, and carried out by elected leaders who use democratic mechanisms against the foundations of democracy itself. Understanding how this works is not merely an academic exercise. It is the precondition for recognizing it in real time — which is much harder than it sounds.
"The crucial issue is not whether a party expresses disdain for its opponents but whether it accepts their right to exist and compete for power. Rejecting the legitimacy of opponents is the first step on the road to authoritarianism." — Steven Levitsky and Daniel Ziblatt, How Democracies Die (2018)
Key Definitions
Democratic backsliding: The incremental erosion of democratic institutions, norms, and competitive fairness by elected leaders through legal or quasi-legal means, without formal declaration of authoritarian rule.
Executive aggrandizement (Bermeo): The dominant contemporary form of democratic breakdown: elected executives who incrementally expand their power and undermine checks through legal mechanisms, each step individually defensible.
Competitive authoritarianism: A regime that holds elections and maintains some formal democratic institutions but in which the playing field is so systematically skewed that genuine democratic competition is structurally impaired.
Mutual toleration (Levitsky/Ziblatt): The norm whereby political competitors accept the legitimacy of each other as rightful contestants for power, even while opposing them vigorously.
Institutional forbearance (Levitsky/Ziblatt): The norm of restraint in using legal powers to their fullest extent when doing so would be destabilizing; the unwritten rule of not maximizing advantage at the expense of the system.
Democratic recession (Diamond): The term for the post-2006 global pattern of more countries showing democratic decline than democratic improvement in successive years.
Polarization: The process by which partisan groups become increasingly ideologically distant and mutually hostile, reducing cross-partisan cooperation and making voters less willing to punish norm violations by co-partisan leaders.
Soft coup: A regime change accomplished through legal but anti-democratic manipulation of institutions, as opposed to military force.
How Democracies Used to End (And Why They Don't Anymore)
From the 1950s through the 1980s, if you wanted to predict which democracies would survive, you would have looked at military-civilian relations. The most common cause of democratic collapse was the coup d'etat: military officers who believed civilian governments were incompetent, corrupt, or ideologically dangerous, and who had both the organizational capacity and the confidence of important elites to act on that belief. Latin America's twentieth century was punctuated by such coups: Brazil (1964), Argentina (1966 and 1976), Chile (1973), Uruguay (1973). Greece (1967). The military stepped in, dissolved the elected government, and promised to restore order before eventually — sometimes — restoring civilian rule.
This model had several properties that, in retrospect, were almost advantages for democracy. Military coups were unmistakable: everyone understood that an illegal seizure of power had occurred, which provided a clear focal point around which democratic resistance could organize. International condemnation was straightforward because the violation was obvious and undeniable. And in many cases, the military's stated intention was eventually to hold new elections and return to civilian government — an implicit acknowledgment of democratic legitimacy that contained the seeds of subsequent democratization.
The pattern began to shift in the late twentieth century. As international norms condemning military coups strengthened, as international organizations and trading partners began attaching democratic conditions to aid and membership, and as the failure of many military governments to deliver competent governance discredited the model, coups became rarer. Political scientists cheered this trend. What they were slower to recognize was that it did not mean authoritarianism had retreated; it meant authoritarianism had adapted. If the old-fashioned coup had become diplomatically and politically costly, ambitious leaders could achieve similar results more gradually, through legal mechanisms, maintaining enough democratic form to avoid triggering international sanction or domestic resistance.
The New Model: From Coups to Slow Erosion
Nancy Bermeo's 2016 Journal of Democracy article "On Democratic Backsliding" provided the most systematic typological account of this transformation. Bermeo distinguished between various forms of democratic breakdown: classic coups (sudden military overthrow), executive coups (elected leaders dissolving legislatures and suspending constitutions), and what she identified as the dominant contemporary form: executive aggrandizement. Executive aggrandizement involves elected executives incrementally expanding their power and eroding checks through a series of steps that are each individually defensible but collectively transformative.
The mechanism is worth tracing carefully, because it is designed to be individually defensible at each step. A newly elected government with a large majority finds the constitutional court has ruled against several of its key policies. It is, after all, legal to reform the judiciary — the governing party simply passes legislation expanding the number of court seats and fills the new seats with loyalists. The next election approaches, and the governing party believes that the opposition is using state media unfairly; it passes legislation requiring "balance" in public broadcasting in ways that functionally advantage the ruling party. Independent electoral administration agencies might be restructured to give the governing party more oversight. Civil society organizations receiving foreign funding might be required to register as "foreign agents." Nongovernmental organizations critical of the government might face enhanced tax scrutiny.
Each of these steps can be defended. Judiciary reform? Every government has the right to reform legal institutions. Media balance? Who could object to balanced reporting? Transparency about foreign funding? Obviously legitimate. Each generates resistance, but that resistance can be dismissed as special pleading by those whose unearned institutional privileges are being curtailed. The erosion is cumulative and directional — each step makes the next easier by reducing the institutional capacity to challenge it — but it appears piecemeal.
The Four Warning Signs
Levitsky and Ziblatt's research across multiple historical cases of democratic breakdown produced four behavioral warning signs — present across cases from Mussolini's Italy and Weimar Germany to Chavez's Venezuela and contemporary cases — that identify what they call "authoritarian predispositions" in political leaders.
The first is rejection of the democratic rules of the game: questioning the validity of future election results in advance, signaling willingness to operate outside constitutional constraints, associating with groups that advocate suspension of normal democratic procedures. The second is denial of the legitimacy of opponents: treating political rivals not as legitimate competitors for power but as criminals, traitors, enemy agents, or existential threats to the nation. This is distinct from vigorous opposition; it involves denying the opponent's right to participate in the political process at all.
The third warning sign is tolerance or encouragement of political violence: a leader who refuses to unambiguously condemn violence by supporters, who praises past episodes of political violence, who suggests that violence against political enemies might be justified, or who surrounds themselves with associates with records of political violence. The fourth is willingness to curtail civil liberties of opponents and critics: expressed support for restricting press freedom, using government power to investigate critics, punishing political opponents through legal mechanisms weaponized for that purpose.
Levitsky and Ziblatt are at pains to point out that historical gatekeepers — party establishments, major newspapers, business elites — have repeatedly failed to apply these criteria consistently, typically rationalizing each warning sign away. Hitler's violence was explained as youthful excess that responsibility would moderate. Mussolini's rejection of democratic rules was interpreted as campaign rhetoric. In the American context, political scientists applying this framework to contemporary leaders have found themselves in the uncomfortable position of making predictions that party loyalists on the relevant side consistently find implausible until they are vindicated.
Case Studies: Hungary, Turkey, Venezuela
Hungary under Viktor Orban provides the clearest case study in executive aggrandizement because it was carried out with unusual speed and completeness. Orban's Fidesz party won a two-thirds supermajority in the 2010 parliamentary elections, which under Hungary's constitution was sufficient to amend the constitution without opposition votes. Within two years, Fidesz had passed a new constitution (the "Fundamental Law"), expanded the constitutional court from 11 to 15 members and filled new seats with loyalists, lowered the mandatory retirement age for judges to force hundreds of experienced judges out and replace them, transferred regulatory authorities including the data protection ombudsman and electoral administration to bodies headed by Fidesz appointees, passed media laws requiring "balanced" reporting enforced by a regulatory authority stacked with Fidesz loyalists, and drawn new electoral boundaries for parliamentary constituencies.
The result was that in 2014, Fidesz won 44.5 percent of the popular vote — not a majority — but received two-thirds of parliamentary seats. In 2022, the opposition coalition won a majority of votes in several measures, but won only a fraction of parliamentary seats. The formal architecture of democracy — elections, parliament, courts, a constitution — remained intact. The competitive fairness that gives democracy its substance had been comprehensively dismantled.
Turkey's trajectory under Recep Tayyip Erdogan followed a similar pattern with different sequencing. Venezuela under Hugo Chavez and then Nicolas Maduro began with more genuine popular support for redistribution and proceeded to hollow out independent institutions more completely. In each case, the pattern is recognizable: an elected leader with a strong popular mandate who uses that mandate not merely to implement policies but to change the institutional rules in ways that entrenched their advantage and disabled future challenge.
Institutional Guardrails: What Actually Protects Democracy
Levitsky and Ziblatt's most original theoretical contribution is their identification of two unwritten norms — mutual toleration and institutional forbearance — as the essential guardrails of democratic stability. These norms are not legal requirements; they are implicit understandings about how competitive politics should be conducted that make formal institutions function as intended.
Mutual toleration means accepting the legitimacy of political opponents as rightful contestants for power. This does not mean agreeing with opponents or refraining from vigorous opposition. It means not treating them as criminals, traitors, or existential threats whose very participation in the political process endangers the nation. The rhetorical shift from "my opponents are wrong" to "my opponents are illegitimate" is precisely the shift that makes democratic competition deteriorate into zero-sum conflict where any cost of defeating the opponent becomes acceptable.
Institutional forbearance means refraining from using legal powers to their fullest extent when doing so would be destabilizing. The United States Senate filibuster was for most of its history a norm, not a rule: senators could theoretically talk legislation to death but chose to exercise this option only rarely and for major legislation. The gradual erosion of this norm through the Obama and Trump administrations — as each side escalated in response to the other's escalation — illustrates how norms deteriorate through reciprocal defection. President Franklin Roosevelt's attempt to pack the Supreme Court in 1937 failed not because it was clearly unconstitutional (it was legal) but because enough senators of his own party refused to participate in what they recognized as a dangerous violation of institutional forbearance.
The practical implication is sobering. Formal institutional rules alone cannot protect democracy; they depend for their effectiveness on informal norms that powerful enough actors can choose to violate. Courts can be packed legally. Filibusters can be eliminated. Term limits can be extended by constitutional amendment. The formal rules are only as strong as the political will to observe them, and that political will is precisely what erodes under the conditions of polarization and competitive anxiety that authoritarian leaders deliberately cultivate.
Polarization: Democracy's Solvent
The relationship between political polarization and democratic erosion is one of the most important findings in recent political science, and one of the most alarming for democracies where polarization has been intensifying. Milan Svolik and Matthew Graham's 2019 American Political Science Review study tested it directly in an American context.
Their experimental design asked survey respondents to choose between two candidates who differed in partisan identity, policy positions, and commitment to democratic norms. The key manipulation was whether one candidate had expressed a willingness to engage in norm-violating but technically legal behavior — gerrymandering opponents into electoral districts with no competitive chance, restricting press freedom, or limiting oversight of the executive. The key question was whether respondents would sacrifice their preference for the co-partisan candidate in order to protect democratic norms.
The results were striking. Most respondents were willing to do so — but only weakly, and the willingness declined sharply with partisan polarization. Respondents who showed higher levels of partisan identity and hostility toward the opposing party were substantially more willing to support a co-partisan norm violator than were less polarized respondents. The democratic constraint on norm violation depends on voters being willing to punish co-partisan leaders for anti-democratic behavior. Polarization systematically erodes this constraint by making the perceived threat from the opposing party large enough to justify tolerating norm violations by one's own side.
This finding has a disturbing implication: polarization and democratic backsliding create a positive feedback loop. As leaders engage in norm-violating behavior, they increase polarization by framing opponents as existential threats. Increased polarization makes voters less willing to punish norm violations, enabling further backsliding. And further backsliding produces more polarization. The cycle can be self-reinforcing, which is one reason democratic erosion tends to accelerate once it begins. For related analysis, see why-political-polarization-increases.
Economic Conditions and Democratic Stability
The economic underpinnings of democracy have generated more research than perhaps any other question in comparative politics. Seymour Martin Lipset's 1959 paper established what became known as "modernization theory": as societies become wealthier, more educated, and more urbanized, they become more likely to be democratic. The proposed mechanisms included the growth of a middle class with independent economic resources and civic capacity, the development of civil society organizations that provide organizational infrastructure for political participation, and the spread of values and cognitive skills that support democratic participation.
The Lipset correlation — wealthier societies are more democratic — is robust and widely replicated. The causal interpretation remains contested. Daron Acemoglu and colleagues' influential 2008 American Economic Review paper argued that much of the apparent relationship between income and democracy reflects common causes — colonial history, geographical factors, institutional legacies — rather than income's direct effect on democracy. When these confounders are addressed through better research designs, the estimated causal effect of income becomes much smaller. Their own preferred hypothesis emphasizes the causal primacy of institutions: societies with property rights-protecting, contract-enforcing institutions both grow richer and become more democratic, but the institutions are the common cause.
Carles Boix's alternative framework emphasizes inequality rather than absolute wealth. His argument in "Democracy and Redistribution" (2003) is that economic elites resist democracy because they fear redistribution: a democratic majority is likely to vote for policies that transfer resources from the wealthy to the less wealthy. The more unequal the society, the higher the expected transfer, and the stronger the elite incentive to prevent or undermine democracy. The middle class plays a pivotal role not through modernization mechanisms but through distributive politics: its members have too little to fear redistribution severely while also too much to depend on the kind of clientelistic patronage that can substitute for democratic accountability among the very poor.
Economic crises consistently appear as precipitants of democratic breakdown in case studies. The Weimar Republic's fragility owed much to the hyperinflation of 1923 and the depression of the early 1930s, which delegitimized the incumbent democratic government and made extreme alternatives — including Hitler's NSDAP, whose electoral support surged with unemployment — electorally attractive to desperate voters. Venezuela's democratic erosion under Chavez had complex economic dimensions; Bolivia's democratic stress under Morales did too. Economic security is not sufficient to guarantee democratic resilience — wealthy established democracies have shown vulnerability — but economic crisis is a consistent amplifier of whatever authoritarian tendencies are present in a political system.
What Saves Democracies
Understanding what causes democratic erosion also reveals, by implication, what can prevent it. Levitsky and Ziblatt's comparative work identifies several conditions associated with democratic resilience: opposition parties that maintain organizational strength and the capacity to mobilize support even when out of power; courts that maintain genuine independence and the capacity to enforce rulings against the executive; military and security forces that stay out of partisan politics; civil society — independent media, professional associations, civic organizations — that provides alternative channels of political information and coordination; and international community pressure that raises the cost of norm violations.
Their historical comparisons are instructive. Costa Rica in the mid-twentieth century faced conditions — political polarization, economic crisis, a tradition of political violence — that had produced democratic collapse elsewhere in Latin America. It survived through a combination of party system structure, the abolition of the military (which removed the coup option), and the development of strong electoral administration institutions. Post-World War II Germany and Japan rebuilt democratic institutions explicitly designed with safeguards against the pathologies that had destroyed Weimar — five percent electoral thresholds, constructive votes of no confidence, strong constitutional courts — though both also benefited from extraordinary international support and unusually capable political leadership.
What Levitsky and Ziblatt are less clear about is whether democracies can recover from advanced backsliding without a genuine political crisis — an economic failure, an international humiliation, an internal fracture of the dominant coalition — that shifts the political incentives. Hungary suggests that once the incumbent advantage has been sufficiently entrenched, electoral reversal becomes extremely difficult even with majority public support for change. Poland's 2023 election, in which the Law and Justice party lost its parliamentary majority despite significant institutional advantages, provides a cautiously more optimistic precedent — though the subsequent fight to reclaim independent institutions illustrated how much damage had been done and how slowly it could be undone.
For analysis of the psychological mechanisms by which ordinary people become complicit in institutional erosion, see why-good-people-do-bad-things. For broader context on how power concentrates and maintains itself, see what-is-power.
The Deeper Problem
The most unsettling conclusion of the comparative political science on democratic backsliding is that democracy is structurally more fragile than most of its citizens typically assume. Democratic institutions are not self-enforcing mechanisms that automatically produce democratic outcomes regardless of what actors do with them. They are tools whose effectiveness depends on the willingness of powerful actors to observe the norms that make them function as intended. Courts cannot protect themselves from being packed. Electoral systems cannot prevent themselves from being gerrymandered. Press freedom cannot secure itself against legal harassment. Each institutional safeguard ultimately depends on political will — on leaders, parties, and citizens who value democratic competition sufficiently to accept their losses and constrain their own use of power.
This dependence on norms and will, rather than on automatic institutional machinery, is what makes democracy permanently susceptible to the kind of erosion Levitsky and Ziblatt document. It also means that democracy's defense is ultimately a political project rather than a purely institutional one: it requires parties and leaders who refuse to violate norms even when doing so would be advantageous, and citizens who are willing to punish norm violations regardless of which side commits them. In a polarized environment, both conditions are structurally difficult to maintain. The challenge is not a new one — every generation of democratic citizens has faced some version of it — but the sophistication with which contemporary backsliders have learned to exploit democratic forms while hollowing out democratic substance makes it particularly acute.
Comparison: Coup-Based vs. Backsliding-Based Democratic Erosion
| Feature | Classic Military Coup | Executive Aggrandizement |
|---|---|---|
| Speed | Sudden, within hours or days | Gradual, over years or decades |
| Legal status | Clearly illegal | Each step technically legal |
| Visibility | Unmistakable, public announcement | Ambiguous, each step deniable |
| Initiator | Military officers | Elected civilian leaders |
| International response | Clear grounds for condemnation | Diplomatic ambiguity |
| Electoral system | Suspended | Maintained but distorted |
| Point of no return | The coup itself | Gradual, often recognized too late |
| Primary examples | Chile 1973, Argentina 1976 | Hungary 2010-, Turkey 2013-, Venezuela |
References
- Levitsky, S., & Ziblatt, D. (2018). How Democracies Die. Crown.
- Bermeo, N. (2016). On democratic backsliding. Journal of Democracy, 27(1), 5–19. https://doi.org/10.1353/jod.2016.0012
- Graham, M. H., & Svolik, M. W. (2019). Democracy in America? Partisanship, polarization, and the robustness of support for democracy. American Political Science Review, 113(2), 392–409. https://doi.org/10.1017/S0003055418000910
- Lipset, S. M. (1959). Some social requisites of democracy. American Political Science Review, 53(1), 69–105. https://doi.org/10.2307/1951731
- Diamond, L. (2015). Facing up to the democratic recession. Journal of Democracy, 26(1), 141–155. https://doi.org/10.1353/jod.2015.0009
- Acemoglu, D., Johnson, S., Robinson, J. A., & Yared, P. (2008). Income and democracy. American Economic Review, 98(3), 808–842. https://doi.org/10.1257/aer.98.3.808
- Boix, C. (2003). Democracy and Redistribution. Cambridge University Press.
- Linz, J. J. (1978). The Breakdown of Democratic Regimes: Crisis, Breakdown, and Reequilibration. Johns Hopkins University Press.
- Przeworski, A. (1991). Democracy and the Market: Political and Economic Reforms in Eastern Europe and Latin America. Cambridge University Press.
- Freedom House. (2024). Freedom in the World 2024. Freedom House.
Frequently Asked Questions
How do modern democracies typically fail?
The pattern by which democracies fail has changed dramatically over the past half-century. In the 1960s and 1970s, most democratic collapses came through military coups: soldiers in tanks rolled into capital cities, dissolved parliaments, and assumed direct control of governments. This model, though still occurring in isolated cases, has become relatively rare. Steven Levitsky and Daniel Ziblatt's 'How Democracies Die' (2018) documents the shift: most contemporary democratic erosion occurs from within, carried out by elected leaders who use the tools of democracy against democracy itself. The pattern involves an elected leader who, once in power, systematically undermines the institutions and norms that constrain executive power — packing courts with loyalists, delegitimizing independent media, changing electoral rules to disadvantage opponents, politicizing previously independent bureaucratic agencies, and using legal mechanisms to harass the opposition. Hungary under Viktor Orban since 2010 is the paradigmatic case: within two years of his supermajority electoral victory, Orban had packed the constitutional court, rewritten the electoral system to heavily advantage his party, brought media under oligarchic control sympathetic to his government, and used constitutional amendments to entrench his changes. Each individual step was legal; collectively they transformed a functioning democracy into what political scientists call a 'competitive authoritarian regime' — one that holds elections but in which the playing field is so tilted that genuine democratic competition is structurally impaired. The gradual, incremental nature of this process is precisely what makes it so difficult to recognize and resist in real time.
What are the warning signs that a democracy is eroding?
Levitsky and Ziblatt identify four key behavioral indicators — present in historical cases from Mussolini and Hitler to Chavez and Orban — that a leader poses an authoritarian risk. The first is rejection of democratic rules of the game: a leader who questions the validity of election results in advance, signals willingness to operate outside constitutional constraints, or whose associates suggest that extraordinary circumstances justify suspension of normal democratic procedures. The second is denial of the legitimacy of political opponents: treating political rivals not as legitimate competitors but as criminals, agents of foreign powers, or enemies of the nation whose very participation in political life is illegitimate. Third is tolerance of or encouragement of political violence: a leader who refuses to unambiguously condemn violence by their supporters, praises past episodes of political violence, or indicates that violence against political enemies might be acceptable. Fourth is willingness to curtail civil liberties of opponents, ethnic minorities, journalists, or critics: expressed support for restricting press freedom, investigating journalists, or punishing citizens who criticize the government. These warning signs appear consistently in historical cases of democratic erosion and identify what political scientists call 'authoritarian predispositions' — a behavioral profile that predicts subsequent erosion of democratic norms. Levitsky and Ziblatt stress that none of these warning signs alone guarantees democratic collapse, but their combination substantially raises the probability. They also emphasize that these signals are often dismissed as campaign rhetoric, exaggerated claims, or political theater — a pattern of rationalization that consistently proves mistaken.
What is democratic backsliding and how is it different from a coup?
Democratic backsliding is the incremental erosion of democratic institutions, norms, and competitive fairness by elected leaders using legal means. Nancy Bermeo's 2016 typology distinguishes it from earlier forms of democratic breakdown. Classic coups — the sudden, violent overthrow of an elected government by the military — were the dominant form of democratic collapse in the mid-twentieth century. Executive coups — elected leaders using executive power to close legislatures and suspend constitutions — were common through the 1970s and 1980s. What Bermeo calls 'executive aggrandizement' — the dominant contemporary form — involves incremental self-empowerment by elected executives who stay within formal legal bounds at each step, even while collectively transforming the regime. The distinction matters enormously for how democracies can defend themselves. Against a military coup, the response is relatively clear: civilian resistance, international pressure, and the illegality of the takeover provide focal points for opposition. Against executive aggrandizement, each individual step can be — and typically is — defended as a legal exercise of normal executive or legislative power. Who, after all, can object to a government that wins an election reforming the judiciary? Or filling administrative vacancies? Or passing new media regulation? The defense of democracy against backsliding requires recognizing that the whole is more dangerous than the sum of its parts, and that the long-term trajectory matters as much as the legality of any individual action. This recognition typically comes too late, after enough institutional change has occurred that the political costs of reversal have become prohibitive for any opponent.
What economic conditions sustain or undermine democracy?
The relationship between economic development and democracy has generated more empirical research and more controversy than almost any other question in comparative politics. Seymour Martin Lipset's 'modernization theory' (1959) proposed the most influential early version of the argument: wealthier, more educated, more urbanized societies are more likely to be democratic. The mechanism he proposed involved the growth of a middle class with the resources, education, and independence to resist authoritarian domination, and the development of civic associations and political organizations that make sustained popular pressure for democratic rights possible. Lipset's correlation between national wealth and democracy is real and robust across the data. The causal interpretation is more contested. Daron Acemoglu and colleagues' 2008 paper in the American Economic Review argued that the correlation largely reflects common causes — colonial history, geography, cultural and institutional legacies — rather than a direct causal effect of income on democracy. Once these confounders are controlled for, the estimated causal effect of income on democracy becomes much smaller. Carles Boix's 'Democracy and Redistribution' (2003) proposed a different economic mechanism: the key variable is not absolute wealth but inequality. In highly unequal societies, economic elites face higher expected costs from democracy (which might redistribute their assets) and therefore have stronger incentives to prevent or undermine it. The middle class is democracy-sustaining not merely because of education or civic culture but because its members simultaneously lack the elite's stake in inequality and the poor's dependence on patronage. Economic crisis consistently appears in case studies as a precipitant of democratic breakdown — by delegitimizing incumbent governments, polarizing society, and creating conditions under which authoritarian 'solutions' become electorally attractive.
How does political polarization threaten democracy?
Political polarization — the sorting of citizens into increasingly opposed partisan camps with growing hostility toward outpartisans — creates structural conditions that make democratic backsliding more likely and harder to reverse. Milan Svolik and Matthew Graham's landmark 2019 American Political Science Review study directly tested the relationship between polarization and democratic erosion. They presented American voters with pairs of hypothetical candidates: one who shared the respondent's partisan identity but had expressed a commitment to an authoritarian policy (gerrymandering to disadvantage opponents, limiting press freedom, etc.), and one from the opposing party who had not. The key question: would voters punish the co-partisan candidate for the authoritarian stance? The answer was: only very weakly, and the effect was substantially moderated by the degree of partisan polarization. The more polarized a respondent's partisan identity, the more willing they were to vote for the co-partisan despite the authoritarian policy. This finding is extremely important because it identifies the precise mechanism by which polarization enables backsliding: it makes voters willing to tolerate norm violations by their side's leader in order to defeat the opposing side. The democratic constraint on executive overreach depends critically on the willingness of co-partisan voters to punish norm-violating leaders — to prioritize democratic rules over partisan victory. Polarization systematically erodes this constraint. When politics is experienced as an existential conflict rather than a competition between legitimate alternatives, the perceived cost of losing becomes high enough that voters are willing to accept norm violations by their side that they would not accept from the other side. This creates the electoral environment in which authoritarian backsliding is most likely to succeed.
Is global democracy really in decline?
Freedom House's annual 'Freedom in the World' reports have recorded continuous decline in global democracy scores since 2006 — approximately two decades of consecutive years in which more countries deteriorated than improved. Larry Diamond's influential 2015 Journal of Democracy article coined the term 'democratic recession' to describe this pattern. By 2023, Freedom House data showed that countries with declining freedom had outnumbered those with improving freedom for eighteen consecutive years, and that the number of people living in countries classified as 'Not Free' had grown substantially. The pattern of decline includes democratic backsliding in countries that were previously considered consolidated democracies — Hungary, Turkey, India, Poland (partially reversed after the 2023 election), Israel (ongoing as of 2024) — as well as democratic stagnation or reversal in many developing countries. This is qualitatively different from the democratic recession of the 1970s, which was primarily a phenomenon of the Global South. The decline in democratic quality in established democracies is a new and alarming development. However, important caveats apply. The measurement of democracy is itself contested: different indices (Freedom House, Varieties of Democracy V-Dem, Polity V) emphasize different dimensions and sometimes produce different trend assessments. The 'democratic recession' claim is strongest if one focuses on declining quality of existing democracies rather than on raw counts of democratic vs. authoritarian regimes. The number of outright democratic collapses is not dramatically elevated by historical standards. The concern is less about sudden reversals than about the cumulative erosion of democratic quality in countries that formally remain democratic.
What institutional structures make democracies more resilient?
Comparative political science has identified several institutional and social features that appear to make democracies more resilient to backsliding. Independent courts with genuine authority to constrain executive action and credible enforcement mechanisms are consistently identified as crucial: when courts capitulate to executive pressure early, as they did in Hungary and Venezuela, the loss of judicial independence accelerates subsequent erosion. Electoral systems matter: proportional representation systems tend to force broader coalition-building and may limit the extent to which any single party can concentrate power, though they also create coalition instability that can be exploited. Parliamentary systems with strong party discipline may be more vulnerable to rapid transformation when a single party achieves a majority than presidential systems with stronger separation of powers — though the U.S. case complicates simple generalizations. Levitsky and Ziblatt emphasize what they call 'mutual toleration' and 'institutional forbearance' as the crucial unwritten norms: mutual toleration means accepting the legitimacy of political opponents as rightful contestants for power, even if you disagree deeply with them; institutional forbearance means refraining from using legal powers to their maximum extent when doing so would destabilize the system. These norms cannot be legislated; they depend on political culture and on the strategic calculations of elites who recognize that restraint today is the condition of fair competition tomorrow. Strong civil society — independent media, active civic associations, professional organizations with ethical norms — provides redundant constraints on executive overreach and maintains alternative sources of political information and coordination. International community pressure has historically played a significant role in some reversals, particularly in Latin America under the framework of the Inter-American Democratic Charter.