Transcript Work plan - IDA Global Development

Independent Evaluation of
Budget Support to Tanzania
2005/06 – 2011/12
Presentation of conclusions and their
implications
Andrew Lawson
Copenhagen, November 2013
Objective of the Evaluation
To Assess: How effective have GBS & SBS been in
contributing to sustainable results on growth and
poverty reduction (2006-12)?
1.How successful have GBS and SBS been in providing the
means to the Government of Tanzania to implement its
national and sectoral strategies?
2.How successful have GBS and SBS been in facilitating
improvements in the design and implementation of these
strategies?
3.As a consequence, how successful have GBS and SBS been
in attaining successful outcomes and impacts on growth and
poverty reduction?
1.How successful have GBS and SBS
been in providing the means to the
Government of Tanzania to
implement its national and sectoral
strategies?
Budget Support has been a major source of public
funding – US $ 5 billion over 7 years, on
average$690 million or $16 per head per year
Table 1: Disbursements of General & Sector Budget Support (Tsh billions) and their significance
2005/
06
2006/
07
2007/
08
2008/
09
2009/
10
2010/
11
2011/
12
592
790
963
941
1,215
928
947
592
0
790
0
963
0
727
64
906
57
831
109
777
169
Total expenditure
3,873
4,475
5,217
6,907
8,312
9,439
10,763
Nominal GDP
16,857
19,010
22,865
26,497
30,253
34,763
41,120
BS as % total expenditure
15.3%
17.7%
18.5%
13.6%
14.6%
9.8%
8.8%
BS as % total ODA
46%
40%
45%
41%
50%
35%
33%
BS as % nominal GDP
3.5%
4.2%
4.2%
3.6%
4.0%
2.7%
2.3%
Tsh. Billions
(Current prices)
Total Budget Support
General Budget Support
Sector Budget Support
Source: Aid Management Platform, MoF and IMF Article IV Reports.
Budget Support has comprised 37% of ODA
…and 13.7% of public spending
Figure 1: Budget Support disbursements (Tsh billions), and as a percentage of total ODA
Source: Aid Management Platform, MoF.
The annual predictability of BS Funds has been
better than other modalities
(actual annual disbursements as % of planned disbursement)
Source: Own computation from data provided by MoF
160%
140%
120%
100%
80%
60%
40%
20%
0%
Total ODA
Total Budget Support
Total Basket
Total Projects
…although within-year Predictability of BS
Funds remains problematic…
Differences between scheduled & actual GBS/ SBS disbursements, per quarter (Tsh. m)
Source: Own computation from data provided by MoF
Budget Support has permitted increased spending on recurrent
expenses, development projects, and new social sector staff
……..
Composition of Central Government expenditure in comparison
with sources of funding (% GDP)
30%
Devpt Foreign Financed
25%
Devpt. Domestically
Financed
Goods, Services and
Transfers
20%
Salaries
15%
Interest payments
10%
Domestic revenue
Domestic revenue + BS
5%
Domestic revenue + BS +
Other external funding
0%
2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12
8
Budget Support facilitated a rise in public spending
on priority sectors from 8.8% to 13.9% of GDP…..
with the share of priority sectors rising from 40 % to
53% of spending
Priority sectors spending (as % of GDP)
Education
Health
Water
Agriculture
Roads
Energy
Total
2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12
3.62%
4.83%
4.70%
4.60%
4.60%
5.30% 4.80%
2.24%
2.29%
2.10%
3.50%
2.40%
2.50% 2.10%
0.60%
1.10%
0.80%
0.70%
0.70%
0.60% 0.50%
0.64%
0.51%
1.10%
1.10%
1.50%
1.40% 1.30%
1.01%
2.10%
2.20%
2.40%
3.60%
3.80% 3.90%
0.66%
0.48%
0.30%
0.30%
0.40%
0.80% 1.30%
8.77% 11.31% 11.20% 12.60% 13.20% 14.40% 13.90%
Evidence suggests that the bulk of General Budget
Support funding was absorbed within the Education
and Roads sectors……
Source: RBA, MoF and own computation
Contributing the means to implement
Government policies:
• Budget Support has allowed public spending to be 13 %
higher than it otherwise would have been
• BS funding has been significantly more predictable than
other modalities, but could be better
• This funding allowed expansion of social sector staffing,
of non-salary recurrent spending, and of Development
spending
• Additional spending accrued to the 6 Priority Sectors
but especially to the Education and Road sectors
3. How successful have GBS and SBS
been in attaining successful
outcomes and impacts on growth
and poverty reduction?
Outputs of Spending: the improved road network –
length increased by 14%, while maintaining condition
Length & condition of National Road Network (paved/ unpaved regional &
trunk roads), 2008 – 2011 (Source: TANROADS)
Improved Outputs & outcomes in the
Education Sector…..
 Enrolment in primary schools doubled, with a GER of
102.3% in 2011 and 98.4% in 2012, while the Net
Enrolment Rate (NER) was 94% in 2011 and 92% in
2012.
 Transition rate from primary to secondary education
more than doubled from 20.2% in 2006 to 53.6% in
2012. The 2ary GER rose from 9.4% in 2005 to 36.9% in
2012.
 The number of university students rose from 45,500 in
2005/6 to 139,600 in 2010/11.
 Gender parity in enrolment has been achieved at
primary level.
…but also many continuing challenges in
the Education Sector
 Quality of education remains an issue:
o From 2007 to 2010 20% drop in the pass rate for the primary school leavers
exam,
o although an increase in the pass rates was achieved from 2010 to 2011.
 Secondary Pass rates remain very low, especially for
Community Schools
o In 2011, 10% of pupils passed the Certificate of Secondary Education
Examination, a fall from 11.6% in 2010.
o Pass rates vary according to type of school. In 2011 was 41.2% for
seminaries, 35.1% for Government Schools, 19.5 % for Non-Government
Schools and 6% for Community Schools.
 Teacher retention is a key difficulty in rural areas.
But Primary School Pupil: Teacher Ratios are
improving and district disparities are narrowing..
Pupils per
teacher 2008
Pupils per
% change
teacher 2011 (2008-2011 )
All districts
Mean
Median
Standard deviation
Average of the lowest decile
Average of the highest decile
MKINGA
MOROGORO URBAN
54.9
55.2
15.0
33.8
80.4
47.5
47.9
8.00
30.0
62.0
-13.5%
-13.2%
-46.8%
-11.2%
-22.9%
63.8
31.7
48.5
27.3
-24.0%
-13.9%
Pupils per teacher 2008
Pupils per teacher 2011
63.75
54.30
47.67
80.37
61.98
55.21
54.88
48.45
47.47
31.68
47.91
33.83
27.27
30.04
15.04
8.00
All districts
MKINGA
MOROGORO
URBAN
Mean
Median
Standard
deviation
Average of the
lowest decile
Average of the
highest decile
Over 2000-2010, Tanzania averaged 7% real annual
GDP growth – outperforming its neighbours
Figure 1: GDP in 2000 constant US$ (millions), for Tanzania and neighbours, 1990 – 2010
Source: World Bank Development Indicators
Yet, the rate of Income Poverty declined only
marginally, if at all…
• 2007 Household Budget Survey (HBS) reported that
the % of people living in poverty decreased from
35.6% in 2001 to 33.3% in 2007.
• This reported reduction in the poverty ratio would not
have compensated for the population growth rate of
about 2.0% per year. (HDR, 2011).
• Reported decline in the poverty rate is within the
boundaries of statistical error: the reduction in income
poverty may have been twice the reported amount but
also there may have been no income poverty reduction
over the period.
On the other hand, there was improvement
in indicators of Non-Income Poverty …
 From 2000 to 2011, Tanzania’s Human Development Index
increased from 0.364 to 0.466 (HDR, 2011), shifting from
significantly below the SSA average to slightly above it.
 The 2000/1 and 2007 Household Budget Surveys (HBS)
show improvements in some indications of welfare
(education and ownership of assets).
 Since 2003, HIV prevalence in adults (15-49 years) has
declined in both males and females.
 There has been progress in the reduction of under five and
infant mortality and Tanzania is now on track to meet
MDG4.
Overview of impact of Budget
Support in Tanzania
2. How successful have GBS and SBS
been in facilitating improvements in the
design and implementation of
Government policies and strategies?
Budget Support as a mechanism for better
policy dialogue (& implementation)
• Tanzania faces major policy challenges in promoting
growth and poverty reduction, and meeting the needs of
an expanding population efficiently and fairly
• Budget Support should help:
– By providing a framework to set policy targets and
measure progress on Mkukuta and sector strategies
– By keeping attention on Macro, PFM & core issues
– By providing advice & giving access to TA
– By providing incentives to faster implementation
Better Structures for Policy dialogue have
been established ….
• The Partnership Framework Memorandum, signed in
2006 by GoT and 14 DPs, set a clear collaborative basis for
a Policy Dialogue to ‘enrich the country’s strategies’
• It established an integrated framework for assessing
performance against Mkukuta targets and against specific
disbursement triggers & conditions, based upon:
– A standardised annual calendar for review/ assessment
– A common Performance Assessment Framework (PAF)
• Thus, in terms of harmonisation, alignment and
stakeholder engagement the framework scores well.
And Budget Support has been influential
in certain policy results ….
• Budget Support has contributed to:
– Consistent political and administrative commitment to
the PFM reform programme
– Consistent support for the Road Fund and related
road maintenance arrangements
– The introduction of Capitation Grants for secondary
schools
– The introduction of Budget Transparency targets
– The establishment of the high-level dialogue on
Corruption
Yet, Evaluation found that Policy Dialogue
was not working well ….
• Senior Stakeholders have pointed to some major
problems, for which we have found evidence :
– A declining level of ownership over the process by
Govt., prompted by a loss of trust by DPs and an
increased level of conditionality
– Excessive transactions costs
– A lack of consistency, coherence and “S.M.A.R.T-ness”
in the definition of PAF indicators
– The absence of a strategic, problem-solving
orientation to the dialogue process
Government’s ownership and leadership
of Policy Dialogue has declined ….
• In 2005/06, the spirit of partnership was strong :
– Targets in the PAF related closely to Mkukuta and the
principles in the Partnership Framework Memorandum
– MoF officials participated actively in defining targets.
• By 2011, far from a Govt-led framework, PAF had
developed into ‘the sum total of all DPs’ individual
preferences and requirements’ (Claussen & Martinsen,
2011)
• After EPA scandal in 2007, trust was lost and PAF
changed from a joint monitoring mechanism to a DPled instrument to exert ‘policy leverage’.
Transaction Costs of Annual Assessment &
related Policy Dialogue have increased
• Assessment process has become more burdensome with
more indicators, more meetings, larger reports being
written, and assessments finalised later in the year due to
time taken to bring necessary information together.
• PAF indicators grew from 55 in 2004/05 to 89 in 2007/08
PAF Indicators by type and year
FY 05 / 06 FY 06 / 07 FY 07 / 08 FY 08 / 09
FY 09 / 10
FY 10 / 11
FY 11/12
Outcome indicator
20
26
34
34
37
34
36
Temporary process
20
20
32
25
25
12
12
Underlying process
24
64
22
68
23
89
14
73
20
82
14
60
8
56
Total
Poor coherence, consistency and quality of PAF
indicators undermine their value..
• Over 12% of PAF indicators have proven unmeasureable
• Of 64 TPAs ‘partially completed’, ‘not completed’ or
‘delayed’, only 16 (25%) were followed up in the next year
• In 4 out of 6 years, over 35% of indicators were new.
Total of new indicators each year
FY 06 / 07 FY 07 / 08 FY 08 / 09 FY 09 / 10
FY 10 / 11
FY 11/12
Outcome indicator
3
7
12
9
13
23
Temporary process
13
29
24
0
10
12
0
16
24%
2
38
43%
3
39
53%
6
15
18%
0
23
38%
1
36
64%
Underlying process
Total New Indicators
As % of total indicators
Tentative conclusions & some implications
• Budget Support facilitated expanded spending on priority
sectors and brought benefits in service delivery.
• There has been steady progress in PFM reform, in
transparency & in strengthening of anti-corruption work.
• Non-income poverty has fallen.
• These are important results which should be celebrated.
But…
• The Policy dialogue process is not working effectively.
• As a result, important policy problems remain unresolved.
• Put simply, Budget Support scores 6.5 out of 10, where
there is a potential for 8 or 9, out of 10.