Is Politics Bankable? Philip Keefer Development Research Group

1
Is Politics Bankable? Philip KeeferDevelopment
Research Group
2
Why do politics matter?

• Politicians are the ultimate arbiters of
  welfare-enhancing, growth-promoting, equitable
  policies.
• A fortiori, they are the ultimate arbiters of
  success of foreign assistance.
• The political economy question What are the
  incentives of politicians to pursue
  development-oriented policies?

3
The paradox

• Efficient public goods and broad public policies
  (e.g., appropriate regulation) are the
  cornerstone of development.
• Government incentives are therefore more
  pro-development the more they favor these over
  rent-seeking and private good provision.
• Why? governments can improve social welfare
  with 100 Rs. of public goods by much more than
  they can improve welfare with 100 Rs. of transfer
  payments/private goods.
• The paradox in many countries, policies that
  benefit fewer people predominate, when with the
  same resources politicians could choose policies
  that benefit more people.

4
Priority questions for addressing the paradox in
development policy

• The primary development question what
  interventions improve political incentives to
  pursue development-oriented policies?
• The secondary question How can we design sector
  interventions to be compatible with political
  incentives?
• The tertiary question How can we build
  constituencies for reform?
• The quaternary question are key decision
  makers supportive of reform?

5
These priorities reverse usual practice

• Our general tendency is to put the cart before
  the horse.
• Great concern about the identity of key
  politicians about building constituencies to
  secure passage of specific reforms.
• However, sustainable development/projects require
  direct attention to the incentives of
  politicians, more than their leadership qualities
  and personal dedication to reform.

6
Political market imperfections

• Ill focus on four conditions that reduce
  political incentives to reform.
• Incomplete citizen information
• Limited political credibility
• Social polarization
• Violence and extra-institutional political
  competition
• Project success more likely when projects are
  designed to take these into account. More
  importantly. . .
• these are themselves significant obstacles to
  development should themselves be the object of
  donor attention.

7
Political market imperfection info

• We cant expect political accountability for
  development outcomes when
• public doesnt know what political decisions were
  made
• public cant observe outcome of decisions.
• public cant observe the impact of decisions on
  their welfare.
• Most donor interventions do not increase citizen
  info most that do fail to provide the right kind
  of info.
• But they should informed citizens (exposed to
  media) much more likely to receive transfers
  (India, US) (probably) more likely to demand
  public goods.

8
PMIs credibility

• Politics is not about policy/public goods in poor
  countries
• high tax/high redistrib. vs. low tax/low
  redistrib competition in social service delivery
  versus no competition deregulation versus
  regulation.
• Few programmatic political parties in poor
  countries
• at best, we see populist reputations (e.g., free
  power) more often, purely clientelist
  (vote-buying/vote blocs/jobs in gov).
• Why?
• Politicians cannot credibly promise high quality
  public goods, public policy to most citizens.
• Can sometimes credibly promise populist transfers
  (free power) to most citizens.
• Can usually credibly promise clientelist benefits
  (pork barrel, jobs in government) to a few
  citizens.

9
PMIs credibility

• Donors do not generally monitor whether
  development-oriented policies are the subject of
  political debate, nor consider how to bridge the
  credibility gap.
• But they should young democracies and
  democracies lacking programmatic political
  parties
• do significantly worse on governance
• spend significantly more on government employees
  and pork barrel projects
• exhibit significantly larger market share of
  government-owned newspapers
• significantly less secondary school enrollment
• are significantly more vulnerable to civil war.

10
PMIs polarization

• Political market imperfections social
  polarization
• Citizen polarization leads them to care more
  about who benefits from policy than the welfare
  effects of policy.
• Related to credibility and information in
  credibility- and information-free environments,
  promises to co-religionists most credible,
  easiest to monitor.
• Donors pay insufficient attention to how they
  might mitigate (1) the political factors that
  exacerbate polarization and (2) effects of
  polarization on implementation of reforms.
• But they should substantial evidence (India,
  Kenya, US, etc.) that public good provision
  suffers in multi-ethnic settings.

11
PMIs violence

• Violence and extra-institutional forces
• Short-circuit institutionalized political
  competition
• Citizens care more about who can protect them
  politicians care more about who has the guns or
  who can threaten them (strikes that cut off
  energy).
• The most difficult situation that donors
  confront.
• Donors are far better at delivering aid to poor
  households than changing the political economy
  equation that undermines development (e.g., lack
  of confidence in central gov).

12
Implications for donors- info

• Information components already in some
  programs/projects.
• Citizen report cards, PETS dissemination, media
  outreach
• Often lack information crucial to political
  accountability.
• Outcomes (benchmarking of school/health
  performance) decision process (e.g., how much
  money approved, by whom).
• PMI analysis yields design elements that increase
  impact.
• Uttar Pradesh best practice interventions to
  encourage better-informed participation have no
  effect on education (Khemani, et al). New
  PMI-informed design being tested in Karnataka.
• PMI analysis can allow citizens to use impact
  evaluation to hold governments accountable.

13
Implications for donors-particip

• CDDs rely on local governance to improve
  service delivery.
• However, PMIs need not be fewer at the local
  level (Khemani, et al.).
• Most successful CDD KDP-Indonesia. Donors
  participate in ongoing governance, however.
• PRSPs build local ownership for reform agenda.
• PRSPs have no/limited effect on PMIs, but these
  determine who owns what.
• Uninformed citizens in non-credible environments
  cannot own reform.
• CDDs and PRSPs need to focus on credibility,
  information.
• E.g. Directly, CDDs do nothing for credibility
  of national policy makers may help indirectly by
  building up credible local challengers to
  national politicians.

14
Implications for donors-how to make tradeoffs

• PMI analysis pinpoints tradeoff between political
  development and economic development.
• Example Aid to households in strife-torn areas
  of Nepal (or food baskets in Iraq, infra in
  Afghanistan).
• Objectives alleviate suffering AND reduce
  conflict.
• Tradeoff well-targeted, hard to monitor versus
  badly-targeted, easy to monitor.
• Second is more transparent about central gov role
  and about whom should benefit gives citizens
  greater confidence in central government.
• High political efficiencies could offset economic
  inefficiencies.

15
Implications for donors monitoring

• Dont just ask whether finance ministers are
  pro-reform. Ask whether politicians are
  investing political capital in promises to
  provide public goods.
• Track whether citizens have info. to monitor
  these promises.
• Use supervision strategically to improve
  credibility of governments to citizens to
  substitute for accountability where PMIs are high
  and citizen leverage over government officials
  low.
• Increase supervision budgets where
  accountability is low reduce where high.

16
What about the 3rd question?

• Weve addressed first two questions
• incentives of politicians, and design of sector
  interventions.
• What about the 3rd Building constituencies for
  reform?
• Difficulty of building reform constituencies is
  precisely a function of political market
  imperfections/violence.
• Have to ask why havent consituencies reformed
  already? PMIs are likely answer.
• If they are deals with reform opponents and
  mobilization of reform supporters likely to be
  too difficult unless PMIs identified and
  addressed.

17
What about the 4th question?

• Are key decision makers supportive of reform?
• How should we support reform champions?
• Natural to look for cooperative, visionary
  counterparts.
• But we have to ask do they support reform
  because of or despite political incentives.
• If the latter (if PMIs large), sustainable
  reform must take into account the alleviation of
  political market imperfections that will confront
  subsequent leaders who are less enlightened.
• CANT rely on the promise of snowballing reforms.

18
Conclusion

• Ample evidence that political market failures are
  at the root of development failure.
• Strategies of building reform constituencies and
  supporting visionary leaders have not generally
  been informed by an analysis of these failures .
  . . but should be.
• But more importantly, far more can be done to
  shape development strategies to mitigate these
  political market failures.
• This does not mean that every project is
  dedicated to political market failures (is
  political).
• It does mean that every project can be seen as an
  opportunity to chip away at these failures.