T407016 // Autumn 2026 // Université de Genève
Comparative
Politics
Session 02. States and state capacity.
Your groups
- Rahnia Sofia Kadri
- Aymeric Mathis Louis Montaz-Rosset
- Anouk Muriel Roth
- Dmitri Sergueev
- Xuan Tinh Catherine Trac
- Xué Mendez Cubillos
- Léa Mosquera Vazquez
- Younes Reda
- Valentine Salihu
- Sarah Nina Steinmayr
- Zoé Clara Dina Traversi
- Simeon Tucovic
- Fatine Zmama
- Hiba Naciri
- Tarciana Oliveira
- Darshan Lucien Rubischung
- Leyla Sahin
- Cherno Sarr
- Alexander Pottokaran
- Emanuela Lorenza Righi
- Mukund Saini
- Tezer Kemal Tezer
- Luana Wirz
Last week, this week
- Comparison is a method of control
- Lijphart's four methods and six types of case study
- Most similar and most different designs
- Three traps. Selecting on the outcome, cases that copy each other, stretched concepts
- Where states come from, with Tilly on war, taxes and protection
- How to measure what states can do, with Hanson and Sigman
- Two group exercises
Keep last week's tools at hand. We will use them to argue with both readings.
One question before the reading
What separates a state that taxes you from a gang that sells you protection?
What does each one give you in return?
Who sets the price, and can you refuse to pay?
At what point would you start calling the gang a government?
Protection, two ways
A shelter against danger, such as a powerful friend, a large insurance policy or a sturdy roof.
A racket in which a local strongman makes merchants pay to avoid damage he himself threatens to deliver.
"Someone who produces both the danger and, at a price, the shield against it is a racketeer."
When the threats a government guards against are imaginary or of its own making, the government runs a protection racket.
From bandits to kings
Kings commissioned privateers and hired bandits to raid their enemies. Demobilized troops became bandits and demobilized ships became pirates.
The Tudors broke up the lords' private armies and razed their fortresses. In France, Richelieu began the great disarmament in the 1620s.
No monarch could govern a large population by armed force alone. Rulers leaned on local magnates, who were also their likeliest rivals.
The line between legitimate and illegitimate violence came clear only slowly, as states made their armed forces unified and permanent.
Four uses of organized violence
Warmaking
- Aim
- Neutralize rivals outside the territory
- Leaves
- Armies and navies
Statemaking
- Aim
- Neutralize rivals inside the territory
- Leaves
- Surveillance and control
Protection
- Aim
- Neutralize the enemies of the rulers' clients
- Leaves
- Courts and assemblies
Extraction
- Aim
- Acquire the means for the other three
- Leaves
- Fiscal and accounting offices
Tilly's rule is that the costlier the activity, the larger the organization it leaves behind.
How war builds a state
- A lord makes war well enough to dominate a substantial territory.
- War needs men, arms, food and money, so he extracts them from the population.
- Extraction means neutralizing local rivals and building tax offices, police and courts. That is statemaking.
- Classes that lend money or help keep order receive protection in return.
National states grew as a by-product of rulers' pursuit of war, and none of those rulers foresaw it.
Protection as a business
- Taxes collected
- 75
- Cost of protection
- − 50
- Tribute to the prince
- 25
A merchant who pays his prince 10 for protection, while foreign rivals pay theirs 15, earns a protection rent of 5.
| If the government is owned by | Its managers will | It resembles |
|---|---|---|
| Citizens | Keep costs and tribute low | A Jeffersonian democracy |
| A monarch | Maximize tribute | A petty despotism |
| The managers | Pay themselves well and maximize tribute | A military junta |
| A dominant class | Tune tribute and protection rents to that class | Tilly's addition |
Swap princes for governments and the question stays current. Who captures the surplus the state produces?
War, debt and taxes
Interest Francis I paid the merchants of Paris on 200,000 francs, borrowed in 1522 for his war against Charles V
Of royal revenue. France's war debt reached about 3 billion francs by Louis XIV's death in 1715
England's national debt begins with William and Mary and grows with each major war
British taxes as a share of national income during the Napoleonic Wars, almost triple the French level
Tilly borrows Peacock and Wiseman's ratchet effect. Revenue raised for a war set a new floor once peace returned.
States that got their armies elsewhere
Follow signals from their patrons
Especially vulnerable to military coups
Wax and wane with world markets for their exports
Resemble the old European states
European rulers built armies by bargaining with their subjects, and the bargains constrained them. States that import their militaries skip that bargain. The examples date from 1982.
Argue with Tilly
Tilly notes that fewer states came out of each large war than went in. An account built on the survivors may be selecting on the outcome.
Rivals imitated the winners, as when Peter the Great rebuilt Russia on Dutch examples. How independent are these cases?
Tilly calls his essay history as a source of hypotheses. Where does that put it among Lijphart's six types?
"If summarizing European warmaking and statemaking is like skating on thin ice, extrapolating that summary to the contemporary world resembles walking on water."
The capacity to do what?
Stateness
Only the monopoly on the legitimate use of force. It misses much of what modern states are expected to do.
Core functions
Defense against external threats, internal order, basic infrastructure for the economy, and revenue extraction.
Every function
Justice, public services, economic management and more. It risks measuring political priorities instead of ability.
Capacity is kept apart from regime type and from policy choices. A highly capable state can still choose to do little.
Three dimensions
Raising revenue. It takes reach into the population, good records, trustworthy agents and enforcement.
- Measured by
- Taxes as a share of GDP, income and trade taxes as shares of all taxes
Holding enough force to defend borders, keep order and enforce the law across the territory.
- Measured by
- Military spending and personnel, police, territory under state authority
Making policy, delivering services and regulating commerce through capable agents and reliable information.
- Measured by
- Bureaucratic quality, census frequency, statistical capacity
Tilly's extraction and warmaking reappear here as dimensions you can measure.
Many indicators, one latent variable
Nobody can observe state capacity directly. A Bayesian latent variable model treats each indicator as a noisy reading of the same hidden quantity.
Indicators from many sources
Countries in a year, at most
Annual estimates
Data points in total
Models with two or three dimensions failed to converge. Only one dimension held up, which the authors read as the three being entwined in practice.
What moves with Capacity
Customs duties are easy to collect at the border and need little administration. States that lean on them tend to score low.
Soldiers and police per person barely track Capacity. Weak states often expand their forces in response to weakness.
Does the measure work?
In 2015 Singapore ranks in the top 25, so the measure picks up more than democracy. Somalia, Yemen and the Central African Republic sit near the bottom.
Capacity correlates at .91 with the World Bank's government effectiveness index and at −.88 with the fragile states index.
Its correlation with log GDP per capita is .79. The two are close, and still distinct.
Researchers sent undeliverable letters to 159 countries. One point more Capacity means about 21 percentage points more letters returned, and 69 days faster.
Capacity in 1960 also predicts 2010. One point higher meant about 12 fewer infant deaths per 1,000 and 5.3 more years of life expectancy, holding income, democracy and taxes constant.
Who gained and who lost
Most countries that scored high in 1975 still scored high in 2015, and Capacity rose in most countries.
In 1982 Tilly picked Iraq as his example of an oil-exporting merchant state. Does its fall bear out his warning?
Argue with the measure
Only the one-dimensional model converged. That fits dimensions that are entwined, and it also fits indicators too noisy to pull apart.
Several indicators are expert scores. The authors concede that bureaucratic quality ratings may reflect analysts' views of economic outcomes.
One score per country and year hides variation inside a state, such as a capital that works and a periphery that does not.
Measurement validity returns at the 27 October complement.
Two ways to study the state
| Tilly 1985 | Hanson and Sigman 2021 | |
|---|---|---|
| Question | Where did national states come from? | How much capacity does each state have? |
| Cases | Early modern Europe, a handful of named states | Up to 163 countries a year, 1960 to 2015 |
| Method | Comparative history that generates hypotheses | Statistical measurement with a latent variable |
| Evidence | "Few illustrations, and no evidence worthy of the name," in his own words | 21 indicators and four kinds of validity test |
| Last week's axis | Depth | Breadth |
Hanson and Sigman cite Tilly's 1990 book for the idea that the dimensions of capacity grow together. Their single dimension is one way to check it.
A finance ministry sends your group one question
Our new trade agreement phases out most tariffs over ten years. Can the state replace the lost customs revenue?
- Tax revenue
- 11% of GDP
- Trade taxes
- 38% of taxes
- Income taxes
- 14% of taxes
- Last census
- 17 years ago
- Territory under state authority
- 80%
- Military spending
- Rising after a border dispute
- Rate the state's extractive, coercive and administrative capacity as low, medium or high, with one reason each.
- Use Tilly. Where could new revenue come from, and whom would the state have to bargain with or coerce to get it?
- Write a three-sentence memo that answers the ministry and names the one indicator you would watch over the ten years.
The profile is invented and describes no real country.
Before Thursday 1 October
Week 03. Regimes and how we measure them.
Munck, Gerardo L., and Jay Verkuilen. 2002. "Conceptualizing and Measuring Democracy. Evaluating Alternative Indices." Comparative Political Studies 35(1), 5-34.
Levitsky, Steven, and Lucan A. Way. 2002. "The Rise of Competitive Authoritarianism." Journal of Democracy 13(2), 51-65.
- Quiz 1 opens the classUni Mail M R030
- Week 4 classM2170
- No class
Progress saved
Questions?
Office hours Thursdays 14:30 to 16:00 in M4328
Rate each group member's participation in today's exercises. Required before you leave the room.