Case selection
Student case writers selected from a list of countries that we identified as cases of potential democratic erosion based on data from the Varieties of Democracy (V-Dem) project. If, during the period of 2000 to the present, the country experienced a sufficient drop in V-Dem’s Liberal Democracy Index, it is eligible to be included.
Originally, we restricted the dataset to countries that were democracies (scoring at least 2 on the Regimes in the World Index during the year it began eroding). Beginning with version 5 of the dataset, we include countries that were autocracies (according to the Regimes in the World Index) during the period of erosion as well. While these instances of “autocratic consolidation” have much in common with democratic erosion, we develop a unique subset of event categories in the Autocratic Consolidation Events (ACE) codebook that apply mainly to these already authoritarian contexts.
Analytic framework
We classify all democratic erosion events as either precursors or symptoms. Among precursors, we further distinguish between civic, economic, political, institutional, and violent/security events. Symptoms include reductions in vertical accountability, reductions in horizontal accountability, or changes in societal norms.
The dataset focuses on “dynamic” events rather than “static” conditions. For example, economic inequality may be a precursor of democratic erosion, but we only code it if the level of economic inequality in a give country changes suddenly from one year to the next.
Finally, we code events in which citizens or other civil society actors engage in acts of resistance to democratic erosion. We summarize the events that fall into each of these categories in the table below.
| Precursor | Symptom | Resistance |
|---|---|---|
| Threats to horizontal accountability Delegitimizing or weakening the judiciary Delegitimizing or weakening the legislature Delegitimizing or weakening subnational units Manipulation of civil service Coup or regime collapse Horizontal corruption | Reduction in horizontal accountability Reduction in judicial independence Reduction in legislative oversight Weakened civil service or integrity institutions Suspension of laws or the constitution Relaxation of term limits Revision of the constitution Reduction autonomy of subnational units | Increase in horizontal accountability Check on executive by judiciary Check on executive by legislature Check on central power by subnational units Check on central power by civil service Post-democratic transition to new constitution |
| Threats to vertical accountability Co-optation of the opposition Mal-apportionment Electoral fraud Electoral violence Increasing control over civil society State-conducted violence or abuse Media bias Lack of legitimacy Polarization Extremist/populist parties Party weakness Vertical corruption Increased surveillance | Reduction in vertical accountability Repression of the opposition Systemic reduction in election freedom and fairness Curtailed civil liberties Media repression No-confidence votes or decreased voter turnout Foreign military action Discrimination against minorities | Increase in vertical accountability Coalitions or elite pacts Increase in electoral integrity Increase in civic capacity Nonviolent protest Violent protest Increase in media protections/media liberalization |
| Exogenous Risk Factors Non-state violence Refugee crisis External influence Economic shocks and health shocks Regional unrest spillover Border disputes Diaspora | Other Pressure from outside actor Exit of people or money State attempts to prevent backsliding |
Variables
The dataset includes the following information for each event:
- Event type (i.e. precursor, symptom, or resistance)
- Year(s) of the event
- Whether there was a reason to doubt the veracity of the evidence used to identify the event
- Whether the event related to horizontal accountability, vertical accountability, or other
- The precursor, symptom, and resistance subtype
- Any outstanding questions or concerns
- Severity of the event on a scale from 1 to 5
Our codebook describes all of these variables in further detail: