Chapter 5 Notes: Democracy, Dictatorship, and Measurement

Box 5.1: States, Governments, and Regimes

Distinction among state, government, and regime
- State: centralized political entity that uses coercion to rule a territory; includes institutions like bureaucracy, judiciary, military (e.g., “the state” of France, Iran, Malaysia, India).
- Government: the set of people who run the state or act on its behalf at a given time; the means through which state power is exercised (e.g., US, Saudi Arabia, Japan in Feb 2023).
- Regime: the set of rules, norms, or institutions that determine how the government is constituted, organized, and how major decisions are made; typically classified as democratic or dictatorial; focus of the chapter.
Democracies and dictatorships come in different types (e.g., presidential, parliamentary, semi-presidential; military, monarchy, civilian regimes).
Tunisian case as an illustration of regime evolution within a state
- Independence from France in 1956; initial civilian dictatorial regime (1956–1987) under Habib Bourguiba.
- 1987–2011: military dictatorial regime under Zine al-Abidine Ben Ali.
- January 2011: mass demonstrations during the Arab Spring forced Ben Ali to resign.
- October 2011: Constituent Assembly elected to draft a new constitution.
- January 2014: new constitution established a semi-presidential democratic regime.
- 2020: government led by Prime Minister Hichem Mechichi; Kais Saied as head of state.
- July 2021: Saied suspended parliament, dismissed the prime minister and cabinet, ruling via decree (described as a “self-coup”).
- July 2022: referendum passed a new constitution increasing presidential power.
- Post-2022: debate whether Tunisia remains a democracy; current government under Prime Minister Najla Bouden with President Saied as head of state.
Methodological point
- Political scientists must wrestle with how best to conceptualize and measure political phenomena (e.g., representation, government performance, economic well-being, culture, justice, accountability).
- Conceptualization and measurement issues in democracy/dictatorship apply similarly to other phenomena discussed in later chapters.
Thinking about measurement is essential for good theory testing; without good measures, empirical tests lose meaning.

Early Democracy vs Modern Democracy

Key historical idea (Stasavage, 2020): democracy existed long before modern, Western-style democracy; democracy arose independently in various places and times.
- Early democracy defined as a system where rulers sought consent of those governed and ruled jointly with assemblies or councils independent from the ruler.
- Early democracy tended to be direct democracy (no intermediaries) or with tight mandates to representatives; assemblies could bind or constrain rulers; decisions by assemblies sometimes not binding, and dissenters could ignore them.
Examples of early democratic-like institutions in various regions and eras:
- Europe: Ancient Greece Ekklesia (adult male citizens), Germanic things, Witan in Anglo-Saxon England, medieval Italian city-states.
- Outside Europe: Kingdom of Mari (Mesopotamia, 3000–2000 BCE), republican India (Himalayan foothills, 700–600 BCE), Huron and Iroquois in northeastern North America, Tlaxcala in pre-Columbian Mesoamerica, central Africa precolonial societies.
- Common features: assemblies with significant influence, regular assemblies, some geographic or social constraints on rulers.
Characteristics of early democracy:
- Typically small-scale, population centers permitting easy assembly attendance.
- More common in hunter-gatherer or highly mobile/remote communities where direct participation was feasible.
- Direct participation or tightly constrained representative practices; mandates on representatives were possible and binding.
Conditions shaping the emergence/disappearance of early democracy (in Stasavage’s framing):
- Emergence coincides with rulers needing revenue or manpower, driving bargaining with societal actors who had credible exit options.
- As societies grew and exit options diminished, or when the value of exit options fell, rulers could centralize power and move away from consent-based governance.
- Where conditions were favorable, rulers built centralized states capable of predation on society; where conditions were less favorable, rulers constrained themselves via assemblies.
Modern democracy
- Broadly: a political system where representatives are regularly chosen in competitive elections under universal suffrage.
- Modern democracy is representative, not direct: people elect representatives who make policy decisions on their behalf.
- The shift from direct to representative democracy was driven by the increasing size/complexity of societies, making direct democracy impractical.
- Elections are competitive when incumbents can lose power; non-competitive elections (incumbents ensured victory) do not constitute democracy.

Box 5.2: Why Bother with Elections?

Core idea: competitive elections are valuable for peaceful conflict resolution and accountability.
Przeworski (2020) coin-toss analogy
- Suppose outcomes in elections were determined by a coin toss; heads = incumbent re-elected, tails = incumbent removed.
- If losers have high chances to win next time, high costs of fighting, and not too large a loss from being ruled by the other side, they will prefer to comply with the coin toss rather than seek power by force.
- Winners moderate policies to avoid triggering violence by losers; they avoid abusing incumbency to undermine future electoral chances; this prevents bloodshed simply due to the expected turnover.
Why voting matters beyond turnover:
- Voting sanctions incumbents for behavior, enabling accountability.
- Election results provide information about the strength of each side and the likely costs of rejecting outcomes.
Conditions for successful peaceful competitive elections
- More likely in wealthy, more equally distributed societies (less high-stakes policy battles).
- More likely in societies with experience that elections do not entail catastrophic outcomes.
- Polarization reduces the likelihood of peaceful elections; high-stakes polarization increases democratic backsliding risk (e.g., Venezuela, United States cases cited).

England and America: Foundations of Modern Democracy

England (Stasavage, 2020): important for modern democracy’s origins
- Gradual decline of the mandate system that constrained representatives.
- Glorious Revolution (1688) established parliamentary supremacy over the monarch, leading to a centralized legislature and majority-rule decisions.
- This fostered competition for legislative majorities and a move toward representative government.
- Universal suffrage was not immediate; true universal suffrage would take centuries.
America: colonial assemblies foster broader political inclusion than Britain by the time of independence
- Colonies lacked a state apparatus; rulers depended on assemblies for governance and revenue.
- Land abundance in colonies meant abundant exit options for residents; labor was relatively scarce, increasing dependence on laborers.
- This dynamic pushed early democracy in the colonies to expand political inclusion to a broader set of people (especially in governing spheres) to sustain rule.
Labor scarcity and exit options (Box 5.3)
- Labor scarcity + land abundance in colonial America led elites to expand political rights for White populations while maintaining coercive labor systems (slavery for Africans).
- Indentured servants had exit options (could gain property after service); enslaved Africans had highly constrained exit options.
- This combination explains why early democracy emerged for Whites while slavery persisted systemically for Africans.
Transition to modern democracy in the United States
- U.S. Constitution (1787) marked a transition: representatives elected; centralized power in taxation and war; weakening local mandates; more competition for Congress.
- 1788: first national-level election with individual suffrage for some groups; full universal suffrage took much longer.
- 1870: Fifteenth Amendment (Black men could vote); 1920: Nineteenth Amendment (women’s right to vote); 1965: Voting Rights Act extended protections and access.
Broader historical point
- The universal suffrage pattern we see today is a relatively recent development; by 1900, only one country had fully universal suffrage; widespread universal suffrage became normative in the second half of the 20th century.

Box 5.3: Labor Scarcity, Exit Options, and Slavery in America

Economic logic: labor scarcity + land abundance creates incentives for coercive labor practices unless exit options are credible
- If elites control land and labor is scarce, they may need to constrain mobility to obtain labor, or offer coercive terms to attract and retain workers.
- In many parts of precolonial Africa and among Native American societies, similar dynamics are observed, with slavery or coercive labor arising in labor-scarce, land-abundant environments.
Colonial America specifics
- Land abundant, labor scarce; rulers depended on people’s labor for growth and revenue; broad suffrage for White colonists arose to maintain governance under collective consent.
- Slavery for Africans persisted due to limited exit options and visible racial hierarchies; indentured White servants had some exit options and potential upward mobility.
- The combination helped produce a political order that was inclusive for Whites but coercive and exclusionary for Black Africans.
Transition in the US: from colonies to a modern democracy with expanded suffrage and civil rights
- 1787–1788 transition and the long journey to universal suffrage for all citizens (eventually completed mid-20th century).

Box 5.4: Alternation in Power

Empirical puzzle: many multi-party democracies experience long-term dominance by a single party (Botswana, Japan, Malaysia, Mexico) before power is formally transferred
- Botswana: independence 1966; long-run rule by Botswana Democratic Party; no power alternation observed to date.
- Japan: Liberal Democratic Party dominated for decades; alternation occurred only after long incumbency.
- Malaysia: initial elections in 1957, 1959, 1964; a pivotal split in 1969; 1969–1990s situation involved emergency measures and a constitution that prolonged one-party rule.
- Mexico: Institutional Revolutionary Party (PRI) held power for many decades until 2000; the question is whether long incumbency indicates dictatorship or a dormant potential for peaceful alternation.
Implication in DD framework
- DD codes a country as a dictatorship if there is no evidence of power alternation under identical electoral rules.
- The presence of alternating power is a crucial condition for distinguishing democracy from dictatorship; long-term single-party dominance can still be coded as dictatorship by DD if there is no recorded alternation under the same rules.

Classifying Democracies and Dictatorships

Core problem: to understand why some countries are democracies and others dictatorships, one must measure democracy and classify regimes.
Theoretical underpinning: concepts are abstract; measurement requires operationalization (converting a concept into observable indicators).
Dahl’s view of democracy: two key dimensions
- Contestation: extent to which people can organize into competing blocs, form parties, express free speech/assembly, and elect leaders in free and fair elections.
- Inclusion: who gets to participate; breadth of the electorate; how inclusive the franchise is.
Dahl’s two-dimensional diagram (conceptual): top-left (high contestation, low inclusion) vs. top-right (high contestation, high inclusion) etc.; modern democracies generally in the top-right (high contestation and high inclusion).
Modern democracy can be viewed as a system where representatives are elected in competitive elections under universal suffrage, with a degree of inclusion and contestation that matches the concept.

Three Measures of Democracy and Dictatorship

Overview: three widely used measures – Democracy-Dictatorship (DD), Freedom House, Varieties of Democracy (V-Dem)
DD (Bormann & Golder, 2022)
- Annual measure for 190+ countries from 1946–2020.
- Conceptualization: democracies are regimes in which governmental offices are filled via contested elections; both chief executive and legislature must be elected; there must be more than one party; there must be an alternation in power under identical electoral rules.
- Operational rules:
- Democracy if and only if all four conditions hold:
 $C1: \text{Chief executive is elected}$ $C2: \text{Legislature is elected}$
 $C3: \text{There is more than one party}$ $C4: \text{An alternation in power under identical electoral rules}$
 $D=1 \text{ if } C1 \land C2 \land C3 \land C4, \text{ else } D=0.$ "
- Dictatorship if any condition fails.
- Key characteristics: minimalist, procedural, dichotomous (democracy vs dictatorship), ignores inclusion dimension.
- Strengths/weaknesses: simple, replicable; may miss nuances of inclusion and quality of democracy.
Freedom House
- Annual global freedom measure since 1972; 2022 coverage: 195 countries + 15 territories.
- Not strictly a democracy measure but widely used as a proxy for democracy/freedom.
- Two dimensions: Political Rights and Civil Rights.
- Political rights: 10 questions (0–4 points each) on electoral process, pluralism/participation, government functioning.
- Civil rights: 15 questions (0–4 points each) on freedom of expression, association, rule of law, personal autonomy.
- Scoring and scaling:
- Political rights score out of 40 → scaled to 1–7; Civil rights score out of 60 → scaled to 1–7.
- Overall Freedom House score = average of the two 7-point scales.
- Regime type continuum: often categorized as Free (Democratic), Partly Free (Mixed), Not Free (Dictatorship).
- Difference from DD: includes substantive outcomes and civil rights; treats democracy as a continuum rather than a strict dichotomy.
Varieties of Democracy (V-Dem): Electoral Democracy (Teorell et al. 2019)
- Annual measure for 182 countries from 1900 to today; multiple conceptions of democracy (electoral, liberal, participatory, deliberative, egalitarian).
- Focus here is on electoral democracy (procedural): captures contestation and inclusion through several components.
- Five core components:
- Freedom of Expression
- Freedom of Association
- Clean Elections
- Percentage of the Population with Suffrage
- Elected Officials
- Scoring and uncertainty:
- Each component score 0–1; component indicators feed into low-level indicators; measurement uncertainty is explicitly reported.
- Two-stage aggregation to a final score:
- Define five component scores: FE, FA, CE, SUF, EO.
- Multiplicative index:
  $MI = FE \times FA \times CE \times SUF \times EO$
- Additive index (weighted):
  $AI = (0.25 \times FE) + (0.25 \times FA) + (0.25 \times CE) + (0.125 \times SUF) + (0.125 \times EO)$
- Final V-Dem electoral democracy score:
  $V\text{-}Dem = 0.5 \times MI + 0.5 \times AI$
- Range: 0 to 1; includes uncertainty ranges for each component/score.
Cross-measure comparisons and notable patterns
- In 2021, classification shares differed by measure:
- DD: 62.9% democratic
- Freedom House: 42.3% democratic
- V-Dem: 51.7% democratic
- General agreement for clear cases (e.g., US, Australia, Sweden as democracies; China, Saudi Arabia, North Korea as dictatorships).
- Mixed regimes show substantial disagreements across measures (Africa and parts of South America).
- The choice of measure affects empirical findings about causes and consequences of democracy (Casper & Tufis 2003; Cheibub, Gandhi, and Vreeland 2010; Elkins 2000).
Bolivia classification example (Figure 5.5)
- Bolivia’s regime classification between 1978–1987 varied across DD, Freedom House, and V-Dem due to differing underlying concepts and aggregation rules.

Comparing Dahl and DD vs Continuous Measures

Dahl’s conceptualization (procedural): two dimensions – contestation and inclusion
- Contestation: freedom to organize into competing blocs, form parties, free speech/assembly, free and fair elections.
- Inclusion: who gets to participate (voting rights, naturalization, etc.).
- Dahl’s framework aligns with a minimalist, procedural view of democracy.
DD vs Dahl
- DD adopts a minimalist, purely procedural/dichotomous view (democracy vs dictatorship) with contestation as the main driving dimension; largely ignores Dahl’s inclusion axis.
- DD treats regime type as a dichotomy rather than a continuum, arguing that regimes with uncontested offices are not democracies.
- Dahl’s view would treat some regimes as democracies if they meet inclusive criteria, even if contestation is limited; DD would not.
The debate on continuous vs dichotomous conceptualization
- Some scholars (Bollen & Jackman, 1989) argue democracy is a matter of degree; continuous measures capture nuances.
- Others argue regimes cannot be half-democratic; a clear-cut break between democracies and dictatorships is more meaningful for some analyses.
- The appropriate approach depends on research question: transitions vs backsliding vs effects of democracy on outcomes.
Practical implications
- The three measures (DD, Freedom House, V-Dem) are highly correlated for clear cases but diverge on mixed regimes.
- Different measures can lead to different empirical conclusions; researchers should match the measure to their theoretical concept and research question.

Validity, Reliability, and Replicability of Democracy Measures

What validity means in this context
- Validity: the extent to which a measure accurately represents the concept it is intended to reflect.
- Challenges include: attributes included, aggregation method, measurement level, and whether the measure’s development aligns with the theoretical concept.
Attributes and aggregation
- The choice of attributes must reflect the underlying concept without over- or under-inclusion.
- DD uses a multiplicative aggregation of necessary conditions; if any condition fails, democracy is not achieved.
- Freedom House uses additive aggregation; attributes are considered interchangeable; a country can compensate a low score in one attribute with a high score in another.
- V-Dem uses a hybrid approach (multiplicative and additive) and weights some attributes differently; this can reflect a more nuanced view of democracy.
Measurement levels
- DD: nominal (democracy vs dictatorship) because it is a dichotomous classification based on necessary conditions.
- Freedom House and V-Dem: interval or ratio-like in practice; provide continuous scores (and V-Dem provides a true 0–1 continuum with underpinning uncertainty).
Reliability
- Reliability refers to the consistency of a measure across time and coders.
- DD is highly reliable due to observable, objective criteria (elections, parties, alternation).
- Freedom House and V-Dem rely on expert coders; subjectivity can reduce reliability, particularly for earlier years.
- V-Dem mitigates some reliability concerns by employing multiple coders and reporting uncertainty ranges.
Replicability and data transparency
- Replicability: ability of others to reproduce the measure’s construction and results.
- King emphasizes the need for full data access and transparent coding rules for replication.
- DD provides clear coding rules; Freedom House provides limited disaggregation; V-Dem disaggregates data and offers extensive metadata.
- DA-RT (data access, production transparency, analytic transparency) movement advocates for public data, transparent production, and transparent analyses to improve reliability and trust.
Implications for using measures
- All measures involve simplification; rely on judgments and data availability.
- The chosen measure should align with the research question; complexity of democracy often necessitates minimalist measures for comparisons or targeted questions about transitions or causal mechanisms.
- Researchers should be explicit about limitations and consider multiple measures to assess robustness of findings.

Conclusion: Measurement Trade-offs and Best Practices

Measuring abstract political phenomena like democracy is inherently difficult and requires trade-offs between simplicity and richness.
Minimalist/procedural measures (e.g., DD) are valuable for studying transitions and regime-type dynamics, particularly when you want a clear, replicable dichotomy.
More substantive measures (Freedom House) capture quality of democracy and civil/political rights, but may introduce circularity if used to predict those same outcomes.
Continuous measures (V-Dem) provide nuanced gradations of democracy and explicitly incorporate measurement uncertainty, offering flexibility for different research questions.
Practical guidance for researchers
- Align measure choice with research question and theory: transitions and regime type vs. quality/outcomes of democracy.
- Be explicit about aggregation, attributes, and measurement level; justify attribute inclusion and weighting schemes.
- Consider the trade-offs between reliability and validity; acknowledge potential biases in expert-coded measures, especially for historical years.
- Use multiple measures to check robustness; report uncertainty ranges (as in V-Dem) and ensure replication by sharing data and coding rules (DA-RT principles).
- Be cautious in interpreting “mixed regimes”; cross-measure differences can reflect conceptual differences rather than measurement errors.
Final philosophical note
- Democracy is multifaceted and context-dependent; the goal of measurement is to enable theory testing, not to claim perfect, universal truth about every regime.
- The field should continue developing better measures while recognizing the value of simpler, well-justified proxies for specific research purposes.

Appendix: Key Formulas and Definitions

Democracy (DD) criterion (aggregated, dichotomous):
- Democracy if and only if all four conditions hold:
 $\begin{cases} ext{C}1:\ ext{Chief executive is elected}[2mm] ext{C}2:\ ext{Legislature is elected}[2mm] ext{C}3:\ ext{There is more than one party}[2mm] ext{C}4:\ ext{An alternation in power under identical electoral rules} \ ext{Democracy} = 1 ext{ if } C1 \land C2 \land C3 \land C4; \ 0 ext{ otherwise} ext{(dichotomous)} \end{cases}$
V-Dem: five components and final score
- Let FE = Freedom of Expression, FA = Freedom of Association, CE = Clean Elections, SUF = Suffrage, EO = Elected Officials.
- Multiplicative index:
  $MI = FE \times FA \times CE \times SUF \times EO$
- Additive index:
  $AI = (0.25 \times FE) + (0.25 \times FA) + (0.25 \times CE) + (0.125 \times SUF) + (0.125 \times EO)$
- Final V-Dem score:
  $V\text{-}Dem = 0.5 \times MI + 0.5 \times AI$
Freedom House score (conceptual, not a single equation here): two dimensions (Political Rights, Civil Rights) aggregated to a 1–7 scale per dimension; overall score is the average of the two 1–7 scales; categories often used: Free, Partly Free, Not Free.
Conceptual dimensions (Dahl):
- Contestation: freedom to organize, form parties, elections, etc.
- Inclusion: who is allowed to participate (suffrage, naturalization, etc.).

Notes

The content above captures the major and many minor points raised in the transcript, including historical context, theoretical distinctions, three major democracy measures, their strengths/weaknesses, and methodological considerations for validity, reliability, and replication.
Throughout, explicit numerical references are included where they appeared (dates, numbers of countries, scales, and specific year ranges).