Old Autocracy or Something New?: Conceptualizing and Measuring Democratic Erosion and Overcoming It

Murat Somer, Alper Yılmaz2023 · DOI

Abstract

Existing measurements of democratic erosion (DE) do not necessarily start with adequate conceptual-theoretical scrutiny and insufficiently incorporate DE’s distinguishing features into operationalizing it. We review existing measurements, offer a definition based on the constitutive characteristics of DE and the roles of oppositions and institutions, and develop new measurement criteria. We show how this changes the population of DE cases in large- and small-N comparative analysis. We argue that recovery from DE should be conceptualized and measured separately from reversals and near-misses, and identify corresponding country episodes. Successful reversals of DE seem to be scarce, and none for countries that endured more than ten years of DE. Successful recoveries is also a small, if not null set. Exceptions are, technically but not necessarily substantively, Ecuador, and substantively but not yet technically, Moldova. DE is a new type of autocratization and polities may need time to learn how to counter it.

Introduction: The Challenge of Conceptual Fuzziness

In contemporary political science, terms like "backsliding," "regression," and "erosion" are frequently used interchangeably. This lack of terminological precision has led to significant inconsistencies in empirical research. As researchers attempt to identify why some democracies survive periods of decline while others collapse, they are met with a fragmented landscape of data. Depending on the specific study and its underlying definitions, the number of identified cases of democratic erosion in recent history fluctuates wildly, ranging from as few as 2 to as many as 44.

Comparison of case counts and process types across existing literature on democratic erosion and backsliding.

This "conceptual fuzziness" obscures the unique nature of Democratic Erosion (DE). Unlike traditional autocratic takeovers—such as sudden military coups—DE is a process of "death by a thousand cuts." It involves a gradual, legalistic hollowing out of democratic institutions from within, often led by democratically elected incumbents. To address this, a standardized "yardstick" is required to distinguish DE from other forms of regime change and to differentiate between a mere "reversal" (halting the decline) and a true "recovery" (restoring democratic health).

A New Framework for Measurement

To move beyond binary "democracy vs. autocracy" classifications, this research proposes a rigorous six-point filter to identify genuine DE episodes using the V-Dem (Varieties of Democracy) dataset:

  1. Democracy Criterion: The state must have maintained a stable democracy for at least five years prior to the decline to ensure a baseline of democratic health.
  2. Liberal Democracy Criterion: The framework utilizes the Liberal Democracy Index (LDI) rather than simple electoral scores to capture subtle infringements on civil liberties and media freedom.
  3. Minimum Length: The decline must persist for at least five continuous years to be classified as erosion rather than temporary volatility.
  4. Minimum Total Decline: A cumulative drop of at least 0.1 on the LDI scale is required to filter out statistical "noise."
  5. Maximum Annual Decline: To distinguish erosion from "sudden death" events like coups, the annual rate of decline must be less than 0.084 (the historical average impact of a coup).
  6. Recovery Period: A "recovery" is defined by five consecutive years of improvement and a return to a higher regime classification.

The 10-Year Tipping Point

Applying these criteria to global data reveals a sobering reality. The research identifies a "critical juncture" related to the duration of erosion. Strikingly, no country that has endured more than 10 years of continuous democratic erosion has successfully recovered.

This suggests that after a decade of decline, anti-democratic norms become "habituated," and power relations are restructured in ways that make the process functionally permanent. DE represents a novel 21st-century challenge; because it operates under the "veneer" of democratic institutions—utilizing laws, courts, and elections to dismantle the system—traditional methods of resistance often prove ineffective.

Key Results: Reversal vs. Recovery

The application of the six-point framework identifies 24 specific country episodes of true Democratic Erosion. The findings highlight a crucial distinction between stopping a decline and restoring democracy:

  • Successful Reversals: Only three cases—Bolivia (2021), North Macedonia (2015-2019), and South Korea (2015-2018)—showed signs of halting the erosion process. However, many of these remain precarious and have since experienced further slippage.
  • Successful Recoveries: The set of true recoveries is "small, if not null." While Ecuador technically meets the criteria for recovery, it remains substantively fragile and politically unstable. Moldova is identified as a promising "substantive" success story, currently four years into a recovery process, though it has yet to reach the five-year threshold required by the framework.

Conclusion and Future Directions

The rarity of democratic recovery underscores the urgent need for further research into the mechanisms of resistance. Future work should focus on the role of opposition actors and formal institutions in either facilitating or checking the erosion process. The "10-year hypothesis" remains a primary area for causal analysis, as understanding why this threshold is so difficult to cross could provide vital insights for protecting vulnerable democracies.

Summary updated

Abstract

Existing measurements of democratic erosion (DE) do not necessarily start with adequate conceptual-theoretical scrutiny and insufficiently incorporate DE’s distinguishing features into operationalizing it. We review existing measurements, offer a definition based on the constitutive characteristics of DE and the roles of oppositions and institutions, and develop new measurement criteria. We show how this changes the population of DE cases in large- and small-N comparative analysis. We argue that recovery from DE should be conceptualized and measured separately from reversals and near-misses, and identify corresponding country episodes. Successful reversals of DE seem to be scarce, and none for countries that endured more than ten years of DE. Successful recoveries is also a small, if not null set. Exceptions are, technically but not necessarily substantively, Ecuador, and substantively but not yet technically, Moldova. DE is a new type of autocratization and polities may need time to learn how to counter it.
BibTeX
@misc{somer2023old,
  author = {Murat Somer and Alper Yılmaz},
  title = {{Old Autocracy or Something New?: Conceptualizing and Measuring Democratic Erosion and Overcoming It}},
  year = {2023},
  doi = {10.33774/apsa-2023-sshrp},
}
Figure