Getting Serious about Stakeholder Analysis

1
Getting Serious about Stakeholder Analysis

• Piloting Political Science Methods in World Bank
  Operational Work
• Barbara Nunberg
• January 26, 2005

2
Stakeholder AnalysisBackground

• Potential winners and losers of policy reforms
  can actively influence reform outcomes
• To account for role of politics, Bank has
  performed intuitive analyses of stakeholder
  preferences
• Need for more systematic political stakeholder
  analysis
• Pilot studies conducted by EASPR using the
  Expected Utility Stakeholder Model

3
Stakeholder AnalysisWhat the Model Can Do

• Identify stakeholder positions on policy reform,
  weigh their potential influence and assess the
  strength of their commitment
• Data cross-checked for internal consistency and
  comparability
• Simulate round-by-round negotiations to gauge
  whether proposed reforms are feasible as designed
• Determine possible strategic options for
  optimizing reform levels using knowledge about
  political dynamics

4
How the Model WorksData Collection Process

• Interviews with country experts to define context
  and relationships related to reform issues not
  opinions or predictions
• Defined policy issue
• Steps in the reform process, in order of
  difficulty
• List of stakeholders with an interest in the
  policy outcome and those with a veto
• Each stakeholders initial bargaining position
• Relative power of each stakeholder on this issue
• Salience of issue to each stakeholder

5
How the Model WorksSimulation Bargaining Process

• Stakeholders try to influence each other to
  secure an outcome they see as favorable
• Model provides a round-by-round simulation of
  prospective political bargaining
• Predicts how key stakeholders will shift their
  positions
• Assesses the level of consensus in support of a
  particular outcome
• Estimates effect of different initial stakeholder
  positions on likelihood of reform success and
  level of policy consensus

6
Stakeholder influence stakeholder resources
stakeholder salience
Opposite extreme stakeholder position
Extreme stakeholder position
Stakeholder A position
Stakeholder C position
Stakeholder B position
Round x Stakeholder positions and influence are
analyzed to determine the winning outcome based
on each stakeholders expected utility.
Iteration
Model goes through risk propensity, stakeholder
perceptions, policy proposals, and stakeholder
policy shifts to simulate bargaining process.
Negotiations stop if stakeholders see no further
gains from continuing discussions
Outcome forecast, predicted timeframe
7
Case Study AnylandProcurement Reform

• Integral part of anti-corruption reform agenda
• Corruption in procurement pervades both public
  and private sectors
• Procurement process highly decentralized,
  allowing for individual discretion
• Prime Minister advocates modest reform, but a
  strong coalition of interests opposes it entirely

8
Anyland ProcurementStep One Defining the Issue
9
Anyland Procurement Step Two Positions and
Influence
10
Anyland Procurement Step Three Bargaining
Dynamics
11
Anyland Procurement Step Four Anticipated
Outcome
12
Anyland Procurement Step Five Potential for
Further Reform

• Re-analyzed using various donor starting
  positions above and below current positions
• Banks initial position a Restructuring of
  agencies dealing with large procurement items
• When lowered 1 step to support internal audits
  within ministries for improved enforcement a
  Successful negotiation with PM and other leaders
  for this level of reform
• Greater potential reform progress

13
Benefits of the Model (1)

• Wide applicability
• Has been used to analyze a diverse set of
  negotiated issues from business to economics to
  politics
• Can be applied to various areas of Bank work
• Accuracy rate
• Approximately 90 in real-time prediction of
  thousands of cases since 1981
• More accurate than traditional methods using
  expert opinion

14
Benefits of the Model (2)

• Structured format for data collection
• Predicts the types of coalitions that may form in
  support of various levels of policy reform
• Allows dual analysis of a macro issue along with
  its component policy parts for better reform
  targeting

15
Limitations of the Model

• Garbage in, garbage out a Quality of data
  collection is critical
• Shortage of qualified experts on some issues
• Need for in-depth country knowledge
• Operational utility depends on close alignment
  with country program tasks
• Ability to forecast has not yet been tested for
  Bank-related policy issues
• No in-house capacity to run the model

16
Next Steps

• Establish framework within country operations for
  ongoing application of these techniques
• Pilot studies on additional country issues
• Develop in-house modeling capacity
• Assess further experience, including accuracy of
  forecasts and recommendations