Civics & Geopolitics: The Spectrum of Political Systems

Rather than a binary "democracy vs. dictatorship" switch, picture regimes along a spectrum, located by three measurable dimensions: ① are elections free and fair; ② is power checked by independent institutions (courts, media, legislature); ③ can rulers be replaced peacefully. Strong on all three → near liberal democracy; weak on all three → near closed authoritarianism; mixed → a middle "hybrid regime." The underlying problem this design solves: how to balance "letting power work effectively" against "preventing power from running out of control."

• Electoral vs. liberal democracy: the former just requires "competitive elections"; the latter adds the rule of law, rights protections, and separation of powers — a ballot box alone is not a complete democracy.
• Competitive authoritarianism (coined by Levitsky and Way): multiparty elections are held regularly, but "the playing field is uneven" — incumbents control the media and abuse state resources; the opposition can run but rarely wins. Neither democratic nor fully closed.
• One-party-dominant systems: aim for centralized decisions, long-term planning, and execution capacity; the cost is a weaker competitive error-correction mechanism — course correction relies more on internal oversight than external turnover.

The procedural view: democracy is simply the set of procedures that let you "swap rulers peacefully via ballots" — verifiable, and we shouldn't smuggle value preferences into the definition. The substantive view counters: without rule of law and rights, elections can become a tool for the majority to oppress the minority; correct procedure isn't the same as freedom. Each sacrifices something — the procedural view may wave through "elected autocracy," while the substantive view risks turning the definition into "only what I like counts as democracy."

"Elections mean democracy" — elections are necessary but not sufficient; whether the field is fair and power is checked matters just as much. Conversely, "no Western-style elections = no governance" is also a misreading: governance quality and regime form are two different questions.

This is a design about how executive and legislative power are coupled. Presidential: the chief executive (president) and the legislature are elected separately, each with a fixed term, neither easily able to dismiss the other — like two independent power supplies, relying on checks to prevent either from overloading. Parliamentary: the chief executive (prime minister) emerges from the legislative majority and survives on its "confidence"; the legislature can topple the government by a no-confidence vote — executive and legislature are wired in series.

Political scientist Linz's famous critique ("The Perils of Presidentialism"): both president and legislature are directly elected and both claim to embody the popular will, so once they clash no democratic principle can decide who must yield — the "dual legitimacy" dilemma; combined with the rigidity of fixed terms and the zero-sum nature of presidential elections, deadlock can escalate. Defenders of presidentialism stress: direct election gives clear personal accountability, fixed terms give stable expectations, and voters choose a person directly. Parliamentary systems are more flexible (a failing government can be swapped anytime), but the cost is potential instability — frequent turnover when many parties fragment the chamber.

"A directly elected president makes presidentialism more democratic and stronger" — direct election is just a different mode of mandate; a prime minister backed by the legislative majority isn't necessarily weaker. The real difference is in how conflict is resolved: presidentialism relies on checks (which can also seize up), parliamentarism on reshuffling (which can also be unstable).

The first two cards covered the "horizontal" division of power; this one covers the vertical — how center and locality share power. Federal: subunits (states/provinces) hold powers that are constitutionally guaranteed and cannot be unilaterally stripped by the center; sovereignty is split between center and units. Unitary: sovereignty belongs wholly to the center, and local powers are "granted" by it — in principle retractable or adjustable. The key difference is not "whether power is shared" but whether local power is locked in by the constitution and untouchable.

• Federal (USA, Germany): states/provinces have their own constitutions, legislative and judicial space; upsides are local fit, protection of diversity, and serving as "policy laboratories"; costs are high coordination overhead, regional inequality, and gridlock-prone veto points.
• Unitary (France, Japan, China): unified policy, clear chains of execution; but over-centralization can ignore real local needs. Note: unitary states can also devolve heavily — Nordic countries have strong local self-government yet remain unitary.
• Devolution (UK): within a unitary frame, the center hands some powers to Scotland, Wales, etc. — a flexible arrangement between the two.

Federalists argue: push power to where it's closest to the problem (subsidiarity), so decisions fit the ground better, accommodate internal diversity within a large country, and let regions try different approaches. Centralizers reply: federalism can create a fragmented "different fates in one country," entrench gaps between rich and poor regions, and let local veto points stall major reform. Each path sacrifices something — federalism sacrifices unity and speed; unitary sacrifices flexibility and local fit.

"Federal = loose, unitary = centralized" is the most common misreading. Regime form and degree of centralization are two independent axes: unitary states can be highly decentralized (Nordics), federal ones quite centralized. What actually determines local power is usually fiscal — who holds the purse — not the nominal structure.

The first three cards point to one underlying fact: institutional design is "trade-off," not "optimization." Just as system architecture has no "best architecture," only "trade-offs against specific constraints," political systems keep choosing among three tensions: accountability vs. efficiency, stability vs. flexibility, broad representation vs. decisive action. Strengthen one end and you almost inevitably weaken the other — this is not a design flaw but the nature of constraints.

Political scientist Lijphart divided democracies into two types that perfectly illustrate this trade-off:

• Majoritarian model (Westminster, UK): power concentrated in whoever wins the majority; fast decisions, clear responsibility; the cost is that minority voices are marginalized and policy can swing sharply with each party turnover.
• Consensus model (Switzerland, Netherlands): through proportional representation, coalition governments, and multiple veto points, it includes as many groups as possible; the cost is slow decisions, diffuse responsibility, and difficulty pivoting quickly.

One wants "fast and clear," the other "stable and inclusive" — neither is better for all societies.

This leads to the dispute over institutional transplantation: can a good system simply be copied? Optimists hold that excellent designs have universal logic worth borrowing; the cautious stress path dependence — the same system moved into different historical legacies, social structures, and cultural soil may grow entirely different fruit. A finely crafted constitutional text won't automatically run where the supporting conditions are missing.

"There exists one optimal system valid everywhere" — this is the deepest misconception. A system's performance depends heavily on the social conditions it's embedded in; asking "which system is best" apart from its soil is like asking "which tool is best" apart from the task.