Eight Questions (2005)

• Corruption in models:
  • As taxes
  • As lobbying
  • As rent-seeking
• Measuring is extremely difficult, for obvious reasons                

Eight Questions (2005)

• Corruption in models:
• Measuring is extremely difficult, for obvious reasons
• Complementary explanations of corruption:
  • Institutional:
    • Economic factors/human capital => institutions => corruption
    • Colonization => legal institutions/regulation/religion => corruption
  • Markets and competition
    • Open trade policies
    • Free press
    • Voting

Eight Questions (2005)

• Corruption in models:
• Measuring is extremely difficult, for obvious reasons                  
• Complementary explanations of corruption:
• Evidence? Mixed, noisy and not causal
• Can higher public service wages reduce corruption?
• Can competition reduce corruption?
• Why isn't there a clear-cut answer?
• Does it actually hurt growth?

Eight Questions (2005)

• Does it actually hurt growth?                                                            

Simple Economics of Extortion (2009)

• AKA, corrupt officials are profit maximizers, too
• Level of corruption depends on the market, demand, and coordination
• 304 trucking trips in Indonesia, with 6000 payments (13% of total cost of a trip is spent on bribes)
• Market structure change: 30,000 officials withdrawn from Aceh (but not North Sumatra)
  • Average bribe in North Sumatra increases in response to the reduction in checkpoints in Aceh
• Bargaining power: Downstream checkpoints have higher bribes than those at the beginning of the route
• Price discrimination: bribe charges change on observables

Simple Economics of Extortion (2009)

• Model(s) predict:
  • The price per checkpoint is decreasing in the number of checkpoints

Simple Economics of Extortion (2009)

• Model(s) predict:
  • The price per checkpoint is decreasing in the number of checkpoints
  • More bargaining power means higher bribe

Simple Economics of Extortion (2009)

Monitoring Corruption (2007)

• Can formal or informal monitoring reduce corruption?
• Increasing government audits reduced missing expenditures by 8 percentage points
• Increasing grassroots participation had little impact
• RCT in Indonesian villages, where a new road was being constructed
  • Treatment 1 is 100% chance of audit (vs. baseline 4%)
  • Treatment A is accountability meetings invite
  • Treatment B is accountability meetings invite with a chance to share an anonymous comment
  • You could receive Treatment 1 and A or B
• Constructed estimates of real cost of construction to compare to reported cost (super cool!)

Monitoring Corruption (2007)

• Increasing government audits reduced missing expenditures by 8 percentage points

Monitoring Corruption (2007)

• Increasing government audits reduced missing expenditures by 8 percentage points
• Increasing grassroots participation had little impact
• Evidence of substitution into nepotism (a "legitimate" expense)