The problem of translating policy into effective delivery experiences from a case study in Tanzania

1
The problem of translating policy into effective
delivery experiences from a case study in
Tanzania

• Bertil Tungodden
• Chr. Michelsen Institute and Norwegian
• School of Economics and Business
• Administration, Bergen, Norway
• Presentation,
• Unleashing Capacities to Achieve the MDGs
• UNDP and Chr. Michelsen, Solstrand, May 1-3, 2005

2
Plan for the presentation

• Background.
• Main results.
• The problem of implementation.
• Central level.
• What explains the variation among schools.
• Summary of findings/policy implications.

3
Background

• Field study in Tanzania, 2003-2004.
• Part of the REPOA team undertaking the public
  expenditure tracking on PEDP commissioned by the
  Public Expenditure Working Group.
• Report published in the fall of 2004.
• Further analysis of data and related literature
  in the fall of 2004.

4
PEDP (2002-2006)

• Primary Education Development Plan (PEDP)
  launched in 2002 collaboration between the GOT,
  various donors and the World Bank.
• Sub-sector program including all parts of primary
  education.
• Enrolment expansion (huge success).
• Quality improvement (?).
• Capacity building (?)

5
Focus of the study

• Capitation grant.
• Cash part.
• Book part.
• Development grant (building and construction).
• Classrooms.
• Teacher houses.
• Desks and latrines.
• Capacity grant (not discuss today).

6
What is leakage?

• Money that are not traceable in district or
  school records.
• Not the same as not reaching the school level.
• Leakage may reflect poor recording or delays in
  disbursements (for example, MOEC, December 2003).

7
Overview per student Capitation
8
Overview Development grant
9
Some findings on absolute changes (average of
sampled schools)
10
Problems of implementation central level

• Lack of knowledge.
• Lack of focus.
• Lack of capacity.
• Lack of continuity.
• Lack of collaboration.
• Lack of transparency.

11
Some common perceptions before the tracking study

• Not much disbursed in capitation grant from the
  central level in 2002 and 2003.
• Joint review (less than 2 USD).
• Not much leakage.
• Pilot Study Report to the World Bank.
• Most of the money disbursed by PO-RALG.
• PER Education.
• The equivalent of 10USD supposed to reach the
  schools both in 2002 and 2003.
• Donors.

12
Flow of PEDP funds
13
Variation among schools and districts

• There are huge variations in the reported amount
  of capitation grant received at the school level,
  both within and between districts.
• What can explain this?

14
Possible explanatory factors

• Resources available at the school level
• Toilet holes per student in 2001
• Classrooms per student in 2001
• Desks per student 2001
• Size of the school
• Location (rural/urban)
• Distance to district HQ
• Information available at the school level
• Perceptions of entitlement
• Availability of newspapers
• Head teachers experience
• The quality of the district administration

15
Summary statistics (information)
16
Results (on total capitation)

• Location is important and statistically
  significant.
• Rural schools received on average 40-1300 Tsh
  less in capitation grant per student in
  2002-2003.
• Perception of entitlement slightly lower for
  rural schools in 2002 (mean 2364 Tsh).
• Other indicators of resources and information not
  statistically significant or unimportant.

17
The district administration is what really
matters!

• District dummies are highly statistically
  significant and extremely important.
• Indicates that local empowerment is probably not
  sufficient (but probably necessary) to make such
  a system work.

18
Similar studies in other countries

• Reinikka and Svensson (Uganda, 2004).
• Substantial leakage in the early phase.
• Regressive transfers.
• Local empowerment essential.
• Das et. al. (Zambia, 2004).
• Substantial leakage in parts of the program.
• Rules, not discretion.

19
Summary of main findings

• Lack of reliable information at all levels and
  among all stakeholders.
• Central level disbursement of capitation
  according to the plan.
• Improvements in the identification of flows.
• Huge differences in expectations at the school
  level.
• Simple plan, complex system.
• Three ministries involved.
• Three funding partners.
• Four levels (government, region, district,
  school).
• Substantial leakage in capitation grant.
• Main leakage in the book part of the capitation
  grant.
• Book availability has probably declined in the
  period.
• Explanations of leakage in capitation grant.
• Rural schools receive less capitation grant.
• The quality of the district administration is the
  main explanatory factor.
• The development grant performs better.
• Less leakage.
• Has been disbursed according to needs.
• Has contributed to improvements at the school
  level.

20
Policy implications

• Establish reliable and simple measures of
  performance.
• Some leakage is to be expected.
• Monitor carefully policy changes.
• Establish reliable data for the whole period.
• Study performance at school level.
• Simplify the system!
• Both funding and disbursement.
• Provide more easily accessible information to all
  stakeholders.
• Donors should be selective when requesting
  information.
• Empower the rural schools.
• Monitor carefully underperforming district
  administrations.
• Enforcement mechanisms.

21
Important reminders

• Decentralization difficult in all poor countries.
• First phase always a challenge.
• Some positive trends.
• Important to loook at the average performance of
  a system, not only the worst cases (PCB).
• Main problem identified. Policy change initiated.
  Need for a follow up study!
• What is leakage? Need for a follow up study!
• What happens at the school level? Need for a
  follow up study.