AI summary by TwinMind

Lecture 20: Public Finance (Part 3)

MIT OpenCourseWare

1h 21m 2s

Cash vs In-Kind Transfer Programs

  • Conceptual difference: cash programs = demand shock, in-kind programs = both demand and supply shock
  • Price effects differ between programs
    • Cash transfers increase demand only
    • In-kind transfers may change aggregate supply depending on government sourcing
  • Example: public housing (government builds) vs housing vouchers
  • Remote vs non-remote regions matter for sourcing patterns
    • Remote locations: government sources externally (net supply shock)
    • Less remote regions: more elastic markets, government may buy locally
  • Empirical findings: prices lower in in-kind areas relative to cash areas in remote locations
  • Real consumption higher in in-kind areas due to price effects

Conditional vs Unconditional Cash Transfers (CCT)

  • CCTs condition aid on fulfilling criteria (school attendance, health checkups)
  • Randomized trials show increased incentivized behaviors
  • Key research question: what is impact of conditionality per se?
  • Two effects of conditioning:
    • Changes price/incentives for desired behavior
    • Cuts off transfers for those who can't meet conditions (lose income effects)
  • Empirical findings:
    • CCT increases school enrollment vs unconditional transfers
    • Better test scores in CCT group
    • Early marriage rates: lower in unconditional treatment than CCT
    • Among non-enrolled students: large program effect on early marriage in unconditional group, not in CCT group
  • Explanation: kids not on margin for school attendance lose income effects in CCT
  • Cost considerations: CCT monitoring costs vs reduced transfer payments (30% of sample cut off)

Transfer Program Design Literature

  • Expanding research area on implementation details
  • Key questions: delivery mechanisms, transparency, corruption reduction
  • Technology impacts: smart cards, biometric authentication, digital payments
  • Government vs private distribution systems
  • Enrollment processes and lump sum vs regular payments

Social Insurance in Development

  • Low/middle income countries starting to consider health insurance, disability insurance, weather/crop insurance
  • Important frontier for understanding challenges in low information, low governance environments

Taxation in Developing Countries - Overview Facts

  • Tax revenue as share of GDP much lower in poor countries
  • Different tax structure composition:
    • Income taxes smaller share in poor countries
    • Corporate income tax much larger share of total income tax in poor countries (50% vs 1/6 in rich countries)
    • Tariffs non-trivial revenue source
    • Seigniorage (inflation tax) more important
    • Informal sector much larger share of GDP

Theories for Different Tax Structures

Gordon-Lee Financial Sector Theory

  • Financial sector generates information for taxation
  • Banks observe money flows through accounts
  • Government can tap into banking records
  • Larger organizations need formal financial sector more
  • Richer countries have more financial development
  • Larger firms can't operate in cash (unlike small pizza shop vs General Electric)
  • Explains focus on corporate taxation in developing countries
  • Inflation as optimal tax on cash economy

Third-Party Reporting Theory (Kleven-Kreiner-Saez)

  • Government gets information from multiple parties, compares for consistency
  • Example: employer and employee both report wages to IRS
  • Under-reporting very low when third-party reported, high when not (even in Denmark)
  • Easier to implement in large firms vs small firms

Firm Size and Tax Evasion Model

  • Random shock version: each employee denounces with probability ε
  • Probability of getting away = (1-ε)^n where n = number of employees
  • Large firms can't sustain evasion due to high probability of denunciation
  • Whistleblower version: US pays 15-30% of recovered taxes to whistleblowers
    • Individual employee can claim entire 30% regardless of firm size
    • Makes evasion unprofitable in large firms
  • Both models predict: as economy develops → larger firms → more tax compliance

Jensen Paper on Employment and Tax Structure

  • Individual income taxation only works in formal employment relationships
  • When employment share too low at income level, governments exempt from income tax entirely
  • As countries develop: employment share increases, moves down income percentiles
  • Countries endogenously change tax structure
  • US historical evidence 1870-present:
    • 1870: employee share only at top income deciles, no income tax
    • Income tax introduced ~1918
    • Over time: employee share moves down distribution, tax exemption threshold moves left
  • Cross-country evidence shows similar patterns (India like historical US, Tunisia more developed pattern)
  • Employee share vs self-employed increases with country wealth
  • Personal income tax base (share above exemption) increases with development
  • Fraction of employees in tax base remarkably constant across development levels

Improving Third-Party Reporting

  • Mexico pension reform example: old system (flat pensions) vs new system (pensions tied to reported wages)
  • Creates worker incentive to ensure honest wage reporting
  • Found better wage matching after reform
  • Policy ideas: government text messages to workers about reported wages

Value-Added Tax (VAT) Structure and Theory

  • Firms taxed on output, receive credit for input taxes paid
  • Imports taxed, exports receive rebates (equal treatment principle)
  • Export rebate process often challenging in developing countries
  • Equivalent to consumption tax if everyone participates
  • Advantage: creates opposing incentives in supply chain
    • Seller wants to under-report transaction
    • Buyer wants to over-report (gets larger tax credit)
  • Third-party reporting throughout chain except final retail level
  • Requires tax ID matching, labor-intensive government analysis

VAT Evasion Types and Experimental Evidence

  • Two evasion types:
    • Collusive evasion: omit entire transaction
    • Unilateral evasion: disagree on transaction amounts
  • Spillover experiment: audit some firms, measure effects on upstream/downstream partners
  • Deterrence experiment: send audit threat letters, measure differential responses by chain position

Experimental Results

  • Spillover effects: strong upstream effects when auditing firms
    • Audited firms pay more taxes
    • Their suppliers also pay more taxes
    • Retail firms show no response (more evasion flexibility)
  • Deterrence effects: largest for downstream firms (more pre-existing evasion)
  • Evidence supports VAT chain theory

VAT Implementation and Extensions

  • Data collection creates network of supplier relationships (research opportunity)
  • Final retail level remains challenging
  • Brazil lottery ticket solution: VAT receipts become lottery tickets
    • Small incentive (2% of purchase price)
    • 24% increase in reported revenues from retail firms
  • Demonstrates potential for consumer-side incentives

Research Frontiers

  • MVPF framework for comparing transfer programs to infrastructure spending
  • Need welfare framework accounting for redistribution effects
  • Long-run evaluation of programs (10-20 year impacts)
  • Tax administration and enforcement optimization
  • Digital payments and automation impacts on tax collection

TwinMind

Get summaries like this for your own meetings

TwinMind takes notes in the background during meetings, lectures, and calls, then writes the summary for you.

Free plan to get started

No bot joins your calls

Phone, desktop & Chrome apps

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome