Comparative Politics

Lecture 13: Corruption, Clientelism, and Accountability

Bogdan G. Popescu

bogdan.popescu@johncabot.edu

John Cabot University

I. The Puzzle

Two Countries, One Starting Point

• 1960s: Both deeply clientelistic
• Philippines was richer per capita
• Both US-aligned Cold War allies
• Both ran on patronage networks
• By 2000: radically different outcomes

Seoul today. Source: Wikimedia Commons.

The Divergence

Figure 1: GDP per capita divergence. Source: World Bank (2023).

What Explains the Gap?

• Not geography — both Pacific-facing
• Not colonial legacy alone
• Key: one broke the loop; one deepened it
• Park insulated technocratic agencies
• Marcos entrenched patronage oligarchy

The Lecture Question

• What is clientelism, and why persist?
• How does it trap states?
• When and how do countries escape?

Manila, Philippines. Source: Wikimedia Commons.

II. Conceptual Foundations

Three Concepts, Often Conflated

Source: Adapted from Kitschelt & Wilkinson (2007).

Concept	Definition	Key feature
Corruption	Abuse of office for private gain	A transaction
Clientelism	Benefits exchanged for political support	A relationship
Patronage	Discretionary allocation of public jobs	An organizational resource

Why the Distinctions Matter

• Corruption is one-off; clientelism is ongoing
• Clientelism involves obligation and monitoring
• Patronage is supply infrastructure for clientelism
• Bribe is impersonal; patron-client tie is social
• Conflating them leads to wrong prescriptions

Clientelism as a Relationship

%%{init: {'theme': 'base', 'themeVariables': {'fontSize': '20px', 'primaryColor': '#ffffff', 'primaryTextColor': '#1e293b', 'lineColor': '#334155', 'primaryBorderColor': '#334155'}, 'flowchart': {'useMaxWidth': true, 'nodeSpacing': 50, 'rankSpacing': 70}}}%%
flowchart LR
    A["<b>Patron</b><br/>Controls resources"] ---|"Material benefits<br/>(jobs, transfers, access)"| B["<b>Client</b><br/>Needs goods/services"]
    B ---|"Political support<br/>(votes, loyalty, turnout)"| A
    C["<b>Broker</b><br/>Monitors & enforces"] --- A
    C --- B
    style A fill:#ffffff,stroke:#4a7c6f,stroke-width:2px
    style B fill:#ffffff,stroke:#334155,stroke-width:2px
    style C fill:#ffffff,stroke:#64748b,stroke-width:2px

Source: Author’s illustration based on Scott (1972).

What Makes It Sticky?

• Asymmetric power: patron controls access
• Social embeddedness: ties are personal
• Iterative exchange: repeated interaction builds norms
• Monitoring: brokers track compliance locally

III. Why Does Clientelism Persist?

The Demand Side

• Poor voters prefer certain small transfers now
• Programmatic promises are uncertain, delayed
• Clientelism provides insurance (Stokes, 2005)
• Income volatility makes bird-in-hand rational
• Not ignorance — rational risk management

The Supply Side

• Parties with weak programmatic credibility
• Strong local networks but no policy brand
• Targeted benefits solve mobilization problem
• Cheaper than programmatic infrastructure
• Equilibrium serves both parties and voters

The Clientelistic Equilibrium

%%{init: {'theme': 'base', 'themeVariables': {'fontSize': '20px', 'primaryColor': '#4a7c6f', 'primaryTextColor': '#1e293b', 'lineColor': '#334155', 'primaryBorderColor': '#334155'}, 'flowchart': {'useMaxWidth': true, 'nodeSpacing': 50, 'rankSpacing': 70}}}%%
flowchart LR
    A["<b>Poverty &<br/>income volatility</b>"] --> B["<b>Voters prefer<br/>certain transfers</b>"]
    B --> C["<b>Parties invest in<br/>targeted delivery</b>"]
    C --> D["<b>No programmatic<br/>infrastructure built</b>"]
    D --> A
    style A fill:#fbe9e7,stroke:#b44527,stroke-width:2px
    style B fill:#fff3e0,stroke:#b7943a,stroke-width:2px
    style C fill:#fff3e0,stroke:#b7943a,stroke-width:2px
    style D fill:#fbe9e7,stroke:#b44527,stroke-width:2px

Source: Author’s illustration based on Stokes (2005).

The Enforcement Problem

• Secret ballot means can’t verify votes
• So how does vote-buying work?
• Nichter (2008): key distinction
• Not vote-buying but turnout-buying
• Pay supporters to show up

G. C. Bingham, The County Election (1852). Source: Wikimedia Commons.

Vote-Buying vs. Turnout-Buying

Source: Adapted from Nichter (2008).

	Vote-buying	Turnout-buying
Target	Opponents / swing voters	Own supporters
Goal	Preference change	Mobilization
Verification	Very hard	Easier (showed up?)
Defection risk	High	Low
Broker role	Persuade	Monitor attendance

Brokers as Enforcement Technology

• Local brokers monitor neighborhoods
• Track who attended rallies, who voted
• Social pressure in small communities
• Brokers are the infrastructure, not cash
• Remove brokers and clientelism collapses

The Korea/Philippines Anchor

• Marcos: turnout-buying via barangay captains
• Local brokers monitored, rewarded, punished
• Park: dismantled broker networks
• Replaced with bureaucratic service delivery
• Removed the enforcement infrastructure

Exercise 1

Demand or Supply? (5 min)

Two scenarios:

• Country A: Universal cash transfer ($50/month). Clientelistic party loses 15% vote share.
• Country B: Same program via local party offices. Party vote share increases.

Which is demand-side disruption vs. supply-side capture? Where is the binding constraint?

IV. Measurement

The Fundamental Problem

• Corruption is hidden by definition
• Participants actively conceal it
• Must infer from indirect evidence
• Three approaches, each with limitations

Approach 1: Surveys

• Afrobarometer, LAPOP: “Were you bribed?”
• Captures reported experience
• Problem: social desirability bias
• Thresholds vary across cultures
• Useful for patterns, not precise estimates

Approach 2: Perception Indices

• TI Corruption Perceptions Index
• World Bank WGI
• Aggregated expert/business surveys
• Measure reputation, not corruption
• Treat as noisy signal

Clean (80+) 60–79 40–59 20–39 Corrupt (0–19)
Source: Transparency International via Wikimedia Commons.

Corruption Perceptions Index, 2023

Figure 2: CPI scores for selected countries. Source: Transparency International (2023).

What the CPI Cannot Tell You

• Cannot distinguish corruption types
• Conflates petty and grand corruption
• Expert perceptions lag actual reform
• Rich countries: corruption may be legal (lobbying)
• Use with caution as dependent variable

Approach 3: Behavioral Audits

• Ferraz & Finan (2008): Brazil’s audit lottery
• Random municipalities selected for audit
• Quasi-experimental design
• Audited mayors punished at ballot box
• Clean identification of accountability

Ferraz & Finan: The Design

%%{init: {'theme': 'base', 'themeVariables': {'fontSize': '20px', 'primaryColor': '#4a7c6f', 'primaryTextColor': '#1e293b', 'lineColor': '#334155', 'primaryBorderColor': '#334155'}, 'flowchart': {'useMaxWidth': true, 'nodeSpacing': 50, 'rankSpacing': 70}}}%%
flowchart LR
    A["<b>Random audit<br/>lottery</b>"] --> B["<b>Some municipalities<br/>audited</b>"]
    B --> C["<b>Corruption<br/>revealed</b>"]
    C --> D["<b>Information<br/>disseminated</b>"]
    D --> E["<b>Voters punish<br/>corrupt mayors</b>"]
    style A fill:#e8f5e9,stroke:#4a7c6f,stroke-width:2px
    style B fill:#e8f5e9,stroke:#4a7c6f,stroke-width:2px
    style C fill:#fbe9e7,stroke:#b44527,stroke-width:2px
    style D fill:#fff3e0,stroke:#b7943a,stroke-width:2px
    style E fill:#e8f5e9,stroke:#4a7c6f,stroke-width:2px

Source: Author’s illustration based on Ferraz & Finan (2008).

The Key Finding

• Disclosed corruption led to a drop in incumbent’s likelihood of re-election by approximately 20 percent
• Effect strongest with local media coverage
• Without information, no electoral punishment
• Accountability = information + elections
• Corruption is punishable — if voters know

V. The Low-Capacity Trap

The Core Logic

• Weak states can’t deliver goods impersonally
• Citizens turn to patrons for access
• Patrons resist bureaucratic reform
• Clientelism undermines state capacity further
• Mutual reinforcement creates stable trap

The Low-Capacity Trap

%%{init: {'theme': 'base', 'themeVariables': {'fontSize': '20px', 'primaryColor': '#4a7c6f', 'primaryTextColor': '#1e293b', 'lineColor': '#334155', 'primaryBorderColor': '#334155'}, 'flowchart': {'useMaxWidth': true, 'nodeSpacing': 50, 'rankSpacing': 70}}}%%
flowchart LR
    A["<b>Weak state<br/>capacity</b>"] --> B["<b>Citizens rely<br/>on patrons</b>"]
    B --> C["<b>Patrons gain<br/>political power</b>"]
    C --> D["<b>Patrons block<br/>bureaucratic reform</b>"]
    D --> A
    style A fill:#fbe9e7,stroke:#b44527,stroke-width:2px
    style B fill:#fff3e0,stroke:#b7943a,stroke-width:2px
    style C fill:#fff3e0,stroke:#b7943a,stroke-width:2px
    style D fill:#fbe9e7,stroke:#b44527,stroke-width:2px

Source: Author’s illustration.

Government Effectiveness and CPI

Figure 3: State capacity and corruption perceptions. Source: World Bank WGI & TI (2023).

Korea: Breaking the Trap

• Insulated Economic Planning Board
• Merit-based technocrat recruitment
• Cold War pressure for developmental results
• Method was authoritarian repression, not reform
• Raises question: can democracies replicate this?

Park Chung-hee, President of South Korea (1963–1979). Source: Wikimedia Commons.

Philippines: Deepening the Trap

• Had equivalent formal institutions
• Chose not to insulate them
• Patronage served regime survival
• Barangay captains = electoral infrastructure
• Oligarchic families captured local state
• No external pressure to reform

Ferdinand Marcos, President of the Philippines (1965–1986). Source: Wikimedia Commons.

VI. When Do Systems Reform?

Three Exit Mechanisms

• External shocks change elite incentives
  • Korea: US pressure, export competition
• Competition makes programmatic policy credible
  • Argentina (Weitz-Shapiro, 2012)
• Rising incomes erode the insurance logic

The Programmatic Turn

• As incomes rise, voters need patrons less
• Programmatic promises become more credible
• Parties shift from targeted to universal goods
• Kitschelt & Wilkinson (2007): broad cross-national pattern
• But the transition is not automatic or linear

Why Reform Is Hard

• Incumbents benefit from clientelistic system
• They control the levers of reform
• Collective action problem among reformers
• Brokers resist merit-based recruitment
• Trap is stable; escape requires disruption

Weitz-Shapiro: Competition as Catalyst

• Argentine mayors in competitive districts
• Shifted to programmatic delivery
• Competition made clientelism costly
• Programmatic commitment signals credibility
• Competition can substitute for external shock

Exercise 2

Diagnosing the Trap (5 min)

Korea vs. Philippines comparison:

• Structural factors (income, geography) or political agency (elite choices)?
• Could Ferraz & Finan’s audit logic work under Marcos?
• Name two missing preconditions
• What does this reveal about institutional portability?

Conclusion

Conclusion

• Clientelism is a rational equilibrium, not a moral failing
• Brokers are the infrastructure; cash is secondary
• Weak states and clientelism reinforce each other
• The clearest historical exit was authoritarian
• Democratic paths exist but are slow and fragile
• Uncomfortable implication: the tools that break the trap fastest are the ones democracies cannot use

References

• Calvo, E. & Murillo, M.V. (2004). Who delivers? Partisan clients in the Argentine electoral market. American Journal of Political Science, 48(4), 742-757.
• Ferraz, C. & Finan, F. (2008). Exposing corrupt politicians: The effects of Brazil’s publicly released audits on electoral outcomes. Quarterly Journal of Economics, 123(2), 703-745.
• Kitschelt, H. & Wilkinson, S.I. (2007). Patrons, Clients, and Policies: Patterns of Democratic Accountability and Political Competition. Cambridge: Cambridge University Press.
• Nichter, S. (2008). Vote buying or turnout buying? Machine politics and the secret ballot. American Political Science Review, 102(1), 19-31.
• Scott, J.C. (1972). Patron-client politics and political change in Southeast Asia. American Political Science Review, 66(1), 91-113.
• Stokes, S.C. (2005). Perverse accountability: A formal model of machine politics with evidence from Argentina. American Political Science Review, 99(3), 315-325.
• Weitz-Shapiro, R. (2012). What wins votes: Why some politicians opt out of clientelism. American Journal of Political Science, 56(3), 568-583.