International Coercion, Emulation and Policy Diffusion: MarketOriented Infrastructure Reforms, 19771

1
International Coercion, Emulation and Policy
Diffusion Market-Oriented Infrastructure
Reforms, 1977-1999

• Witold J. Henisz
• The Wharton School
• Bennet A. Zelner
• Georgetown University
• and
• Mauro F. Guillén
• The Wharton School

2
Outline

• Background
• Motivation
• Determinants of Reform Adoption
• International
• Domestic
• Empirical Results
• Robustness
• Conclusion

3
Background

• Broader research agenda When are market-oriented
  reforms (in infrastructure services)
  efficiency-enhancing and sustainable?
  
• Methodology
• 300 field interviews in 14 countries with
  politicians, regulators, investors, activists,
  academics
  
• Quantitative analysis
• Secondary data from ITU, IEA, World Bank, private
  consultancies
  
• Project-level data gathered independently
• Expectation that domestic institutional factors
  (supporting ability of a country to credibly
  commit to reforms) would mainly determine success
  
• Investment? output efficiency gains ? public
  support? sustainability
  
• Interviewees highlighted negative effect of
  externally-imposed reforms
  
• Perception of such reforms as inappropriate or
  illegitimate
  
• This paper do international forces influence the
  domestic adoption of market-oriented reforms in
  infrastructure services?

4
Motivation Reform Adoption Widespread
5
Motivation Followed Reform Adoption
6
Conditionality Detailed Examples

• 11.1m to Bangladesh to improve the performance
  of Bangladesh's telecommunications sector through
  strengthening elements of the policy,
  institutional, and regulatory framework in order
  to promote the competitive provision of
  telecommunications infrastructure and services.
  There are two main project components. The main
  objective of the Ministry of Post and
  Telecommunications (MOPT) component is to
  separate the policy and operational roles of the
  Government. This will be achieved through 1)
  restructuring the Bangladesh Telephone and
  Telegraph Board (BTTB) and 2) strengthening the
  policy function in MOPT. Technical assistance
  will be provided to implement BTTB restructuring
  and convert it from the present Government board
  into a company, registered under the Company's
  Act.
• 120m to Congo to strengthen regulatory
  authorities in the telecommunication, transport
  and energy sectors through technical assistance
  and training. It will facilitate divestiture from
  the public enterprises by supporting formulation
  of strategies reflecting the consensus of all
  stakeholders
• 22m to Afghanistan to set up an independent
  Regulatory Commission (RC) and a
  Telecommunications Act will be formulated to
  reflect the underlying policies and objectives.
  During the 5 year period covered by the existing
  policy, the ATA's clear intention is the
  corporatization of the Ministry's fixed
  telecommunications operations. Plans for eventual
  privatization will be augmented by an
  institutional strengthening exercise to ensure
  administration procedures and accountability meet
  all international quality and performance
  expectations.

7
Countries whose World Bank Letters of Intent
Include Market-Oriented Infrastructure Reform

• Albania (12/01/01)
• Benin (26/12/00)
• Bolivia (20/12/99)
• Brazil (3/11/00)
• Bulgaria (18/08/00)
• Burkina Faso (17/4/00)
• Cameroon (6/12/00)
• Cape Verde (26/4/99)
• Central African Rep (15/12/00)
• Chad (6/7/00)
• Colombia (22/8/00)
• Rep. of Congo (3/11/00)
• Dominican Rep (22/10/98)
• Ecuador (10/8/00)
• Estonia (24/11/00)
• Ethiopia (29/01/01)

• Georgia (12/7/99)
• Ghana (25/6/00)
• Guinea (6/12/00)
• Guinea-Bissau (13/11/00)
• Honduras (13/4/00)
• Jordan (4/7/00)
• Kazakhstan (22/11/99)
• Lesotho (12/2/01)
• Mali (11/8/00)
• Mauritania (25/5/00)
• Nicaragua (13/12/00)
• Niger (21/11/00)
• Peru (2000)
• Senegal (4/6/99)
• Uganda (21/8/00)
• Zambia (30/06/00)

8
International Factors ?Reform Adoption (Coercion)

• Strong states or IFIs pressure countries to adopt
  reforms
  
• Offer incentives to (agents to) adopt reforms
• Examples
• US investors demand reforms as a condition for
  more investment (and threaten pullout otherwise)
  
• IMF / WB demand reforms as a condition for more
  lending
  
• As determinants of successful reform grew more
  expansive, so did the scope of conditionality
  (IMF 1980s 10 terms ? 1990s 26 terms World
  Bank 1980s 32 ? 1990s 56)
• 1993, market oriented reform required for
  sector lending
  
• Other funds held hostage until reform initiated
  (46b to Indonesia in 1997 2.6b to Ukraine in
  2001)
  
• Hypotheses
• H1 Adoption rate of reform increases w/ external
  
  
• H2 Adoption rate of reform increases w/ external
  at an increasing rate over time.

9
International Factors ?Reform Adoption
(Emulation)

• Emulation
• Rational attempt to emulate success of reforms
  elsewhere
  
• Domestic pressure to conform with reforms
  elsewhere
  
• Examples
• Government hires consultancy to compare/contrast
  viable reform models chosen in similar countries
  
• Opposition party proposes government emulate
  neighboring countrys reform program
  
• Hypothesis
• H3 Adoption rate of market-oriented reform by a
  given country increases w/ the adoption of such
  reforms by peer countries

10
Domestic Factors ? Reform Adoption

• Economic conditions and performance
• Political economy of reform literature highlights
  role of status quo as counterfactual against
  which highly uncertain reforms are evaluated
  
• Costs of poor sector-level performance of
  national economic significance and previously
  shown to promote reform
  
• H4 Adoption rate of market-oriented reform is
  inversely related to economic performance at the
  sector level.
  
• Political Institutions
• Political institutions that provide checks and
  balances limit discretion of policymakers and
  enhance commitment
  
• Prior qualitative and quantitative studies show
  improved private sector outcomes in
  infrastructure when checks and balances are
  present
• Payoff or potential returns to reform increase in
  checks and balances
  
• H5 Adoption rate of market-oriented reform
  increases with the presence of credible checks
  and balances in the political environment.

11
Empirics Dependent Variables

• Unit of observation is country-year
• Nine market-oriented reform indicator variables
  in both telecommunications and electricity
  
• Deregulation
• Separation of regulatory function from ministry
• Separation of regulatory function from incumbent
• Subjective judgment that regulator
  semi-autonomous from government
  
• Subjective judgment that regulator autonomous
  from government
  
• Privatization
• Minority
• Majority
• Complete
• Liberalization
• Long distance (autoproduction)
• Local (generation for external sale)

12
Empirics Independent Variables

• International factors
• Coercion
• FDI as percentage GDP
• Multilateral lending as percentage of GDP
• Emulation
• Adoption by other countries with each countrys
  prior reforms weighted by their trade share with
  the focal country
  
• Domestic Factors
• Sector-level performance
• Waiting list for telecommunications service as
  of population served
  
• Electricity lost in transmission or distribution
  as of electricity generated
  
• Political institutions
• Political constraints index (Henisz, 2000)
• Other
• Size (log of population)
• Wealth (log of per capita GDP)

13
Empirics Modeling Procedure

• Event history analysis (Weibull)
• Pool nine reforms within each sector
• Primary specification (Tests for H1, H3, H4, H5)
• Allow intercepts and time parameters to vary by
  reform type
  
• Cluster standard errors by country and reform
  type
  
• Testing time-varying effect of conditionality
  (H2)
  
• Allow multilateral lending / GDP to have a
  variable effect over time
  
• Collapse reform strata into single intercept and
  time-varying parameter
  
• Still cluster
• Examining economic significance of results
• Collapse reform strata and report hazard ratios,
  which are otherwise a function of reform type
  
• Choose 1977 (telecommunications) and 1978
  (electricity) as base year for analysis based on
  Chilean adoption of reforms in those years

14
Time-varying Effects of Time-varying Covariates
Vs.
15
Regression Results
16
Economic Significance of Results

• A 1 SD increase in variable leads to a
  predicted increase in the adoption rate of reform
  of X in telecoms (electricity)
  
• Coercion (H1)
• FDI/GDP 4.4 (4.4) percent
• Multilateral lending/GDP 59 (n.s.) percent
• For severely indebted countries 100 of GDP
  400 (n.s.)
  
• Time varying effect of multilateral lending (H2)
• Predicted adoption rates 24-50 lower in year 5,
  24-27 in year 10, and 0-12 in year 20
  
• Generates weakly significant (p 0.082) support
  for effect of multilateral lending on adoption of
  electricity reforms
  
• Policy emulation (H3) (trade-weighted) adoption
  by peers 178 (120)
  
• Economic conditions (H4) sector-level
  performance indicator 66 (70)
  
• Political institutions (H5) political
  constraints 79 (n.s.)

17
Robustness I

• Selection (i.e., countries that need to go to the
  WB/IMF need to reform)
  
• Stage 1 Predict which country goes to IMF or
  World Bank
  
• Domestic need (Reserves as a ratio of imports,
  Budget balance, Debt service, Ratio of FDI/GDP
  and Ratio of Portfolio investment/GDP)
  
• Constraints at IMF/WB (number of other countries
  currently drawing funds)
  
• Stage 2 Our model
• Stronger support for effect of multilateral
  lending in both sectors
  
• No support for effect of political constraints
• Robustness across different types of reform
  (deregulation subjective and objective,
  privatization and liberalization)
  
• Loss of statistical power
• In telecommunications
• Robust at aggregated measure of deregulation and
  for privatization (except H4)
  
• FDI/GDP and sector performance driving objective
  measures of deregulation
  
• Multilateral lending, trade weighted reform and
  political constraints driving subjective measures
  of deregulation
  
• In electricity
• Very robust across aggregated and disaggregated
  measures of deregulation
  
• Privatization influenced only by FDI/GDP
• Competition influenced only by trade-weighted
  reform in other countries

18
Robustness II

• Do IMF, WB, RDB and other multilaterals have
  equal effects on all types of reform?
  
• of IMF funding has greater impact than WB
• IMF funding enhances likelihood of de jure
  deregulation whereas WB enhances de facto
  
• IMF and WB both show greatest impact on
  privatization
  
• Regional Development Banks often an impediment to
  reform
  
• No significant effects in electricity
• No evidence of omitted variables bias
• Democratization some evidence that democracies
  reform more in telecoms
  
• Regime or government durability some evidence
  that older governments and regimes reform more in
  telecoms
  
• Openness (trade and portfolio investment)
• Debt service (as share of GDP or exports)
• Size of government or composition of spending
  some evidence that smaller governments more
  likely to implement reform in telecoms
  
• Sector-level composition of output
• Countries with more rapid population (but not
  income) growth more likely to reform in both
  sectors
  
• Time-varying effects for other independent
  variables
  
• Sector performance shortfall increasing over time
  in telecom
  
• Trade weighted reform in other countries and
  political constraints increasing over time in
  electricity
  
• Varying initial year led to no qualitative
  changes in results

19
Conclusion

• Privatization programs across developed and
  developing countries can be understood only with
  an appreciation of their international context
  (Ikenberry, 1990)
• Likelihood of adoption of market-oriented reform
  in infrastructure services jointly influenced by
  
• International forces of coercion and emulation
• Domestic economic and political forces
• Initial adopters influenced relatively more by
  domestic factors as compared to late adopters
  where international forces played a greater role
  
• Normative implications of this trend unclear but
  warrant additional research
  
• Policy and welfare effects of conditionality on
  post-reform
  
• output and efficiency gains
• evolution of market structure
• durability of reform
• Political hazards faced by investors may be a
  function not just of domestic political
  institutions and preferences but also of origins
  of policy reform