Transcript Presented by Rohan Sweeney 30th May, 2012

Centre for Health Economics
Have SWAPs influenced aid flows and
aid effectiveness?
Rohan Sweeney, Duncan Mortimer and David W. Johnston
14th February 2014
www.buseco.monash.edu.au/centres/che/
Centre for Health Economics
“SWAp has truly become a popular and
widespread means of coordinating and structuring
development aid”
(Sundewall and Sahlin-Andersson 2006).
Centre for Health Economics
The SWAp is a “process rather than a fixed blueprint”
(Walford 2003)
Agreement
Sector-wide health strategy
Government-led
Budgeted
Share processes
Government systems
Centre for Health Economics
SWAp has desired implications for DAH funding flows.
•
SWAp promotes increased general sector support.
•
SWAp can enable redirection of DAH towards domestic priority areas.
Important because….
•
Project-based DAH continues to dominate – only 7.7% sector support
between 2002-06 (Piva and Dodd, 2009).
•
MDGs and disease focused donors (eg. GFATM, GAVI, Clinton) have
encouraged disease specific project-based DAH.
•
Evaluations have found increases DAH support, however case study
methods don’t enable contemporaneous control, so we can’t predict
what would happen in absence of SWAp.
Centre for Health Economics
Impact of SWAp on funding flows - research questions
a) has DAH allocated to sector support increased as a result of
SWAp implementation?
b) have SWAps changed how DAH has been allocated across other
key health funding areas?
Centre for Health Economics
Methods
•
Searched for countries with implemented health SWAps.
•
Constructed a unique dataset of DAH recipient countries, which
includes total levels of DAH and levels allocated between key health
areas (IHME 2011):
–
•
HIV, TB, “maternal and child health (MNCH), malaria, sector support and NCDs.
Using a linear probability model, comparable treatment and control
countries were identified. Likelihood of SWAp implementation was
predicted given:
– GDP/capita, DAH levels, geographic region, no. of donors, life expectancy,
and population levels.
•
Fixed effect panel regression techniques employed.
Centre for Health Economics
The sample
Table 1
Variable
Total DAH
(millions $US)
DAH/capita
($US)
population
(millions)
GDP/capita
($US)
life expectancy
(years)
a
SWAp implementing
Non-implementing countries
countries (n=21)
(n=18)
P value
Mean
Std. Dev.
Mean
Std. Dev.
11.04
9.86
8.88
11.04
0.52
0.88
0.72
1.01
0.96
0.63
18.08
24.68
17.64
33.04
0.87
333.98
170.40
513.36
317.36
0.02a
49.84
6.83
51.05
6.92
0.59
t-test for difference in means with unequal variance.
Table 2:
Impact on levels of Sector Support
Log of Levels of DAH as Sector Support
(1) Centre
“Sector” DAH
(2)
for Health Economics
“Sector” DAH
-
SWApt-1
3.32*** (0.90)
Malaria t-1 (prob death/1,000)
-0.01 (0.01)
-0.01 (0.01)
HIVt-1 (% of 15-49yr olds)
0.14 (0.39)
0.13 (0.40)
TBt-1 (incidence/100,000)
0.00 (0.01)
0.00 (0.01)
IMRt-1 (per 1,000 live births)
0.03 (0.05)
0.04 (0.05)
Log(GDP/capita)t-1
1.27 (1.13)
1.42 (1.11)
Gov – effectivenesst-1
3.08 (2.30)
2.49 (2.27)
Control of corruptiont-1
-2.12 (1.79)
-1.61 (1.74)
Pre – SWAp (1-2 years prior)
1.50 (1.44)
Early-SWAp (years 1-2)
3.06** (1.46)
Later-SWAp (years 3+)
4.84*** (1.26)
N
773
773
SWAp countries
21
21
Control countries
18
18
Table 3:
Log of Levels of DAH to health areas
Impact on allocations to
other health areas
SWApt-1
Malaria t-1 (prob death/1,000)
^ Beta distributions
(1)
(2)
Malaria
Centre for Health Economics
(3)
(4)
HIV
TB
MNCH
0.24 (1.25)
-1.60** (0.65)
0.23 (1.36)
-1.56** (0.72)
0.03*** (0.01)
-0.00 (0.01)
0.03** (0.01)
0.00 (0.01)
Standard errors in parentheses
* p < 0.10, ** p < 0.05, *** p < 0.01
HIVt-1 (% of 15-49yr olds)
0.38 (0.35)
-0.00 (0.20)
0.30 (0.35)
-0.27** (0.13)
TB t-1 (incidence/100,000)
-0.00 (0.00)
0.00 (0.00)
-0.01 (0.01)
0.00 (0.00)
IMRt-1 (per 1,000 live births)
0.05 (0.04)
-0.02 (0.04)
0.05 (0.05)
-0.02 (0.04)
Log(GDP/capita)t-1
-1.12 (0.98)
1.31* (0.71)
-0.81 (0.62)
0.46 (0.80)
Gov – effectivenesst-1
2.73 (1.80)
3.90*** (1.08)
0.08 (1.65)
1.73 (1.13)
Control of corruptiont-1
1.36 (1.52)
-0.77 (1.00)
-0.31 (1.34)
1.18 (1.01)
773
773
773
773
N
Centre for Health Economics
Key messages
•
SWAps have facilitated increased levels of DAH directed as sector
support.
•
SWAps appear to have facilitated reallocations of DAH fund flows across
key health areas, specifically away from HIV and MNCH.
•
SWAp implementation has facilitated changes in funding flows
consistent with SWAp aims to increase country ownership of DAH
programmes.
Centre for Health Economics
References
Brown A, Foster M, Norton A and Naschold F. The status of sector wide approaches. Centre for Aid and
Public Expenditure. Working Paper 142. 2001. Overseas Development Institute.
Institute for Health Metrics and Evaluation , I. (2011). Development Assistance for Health Country and
Regional Recipient Level Database 1990-2009. Seattle, Institute for Health Metrics and Evaluation
OECD Development Co-operation Directorate. (2010). "Paris Declaration and Accra Agenda for Action."
Piva, P. and R. Dodd (2009). "Where did all the aid go? An in depth analysis of increased aid flows over
the past 10 years." Bulletin of the World Health Organization 87: 930-939.
Ravishankar N, Gubbins P, Cooley RJ, Leach-Kemon K, Michaud CM, Jamison DT, Murray CJL.
Financing of global health: tracking development assistance for health from 1990 to 2007. The Lancet
2009;373; 2113-2124
Sundewall J, Sahlin-Andersson K. Translations of health sector SWAps--a comparative study of health
sector development cooperation in Uganda, Zambia and Bangladesh. Health Policy. 2006 May;76(3):27787.
Walt G, Pavignani E, Gilson L, Buse K. Health sector development: from aid coordination to resource
management. Health Policy Plan. 1999(a);14:207–18.
Walford, V. (2003). Defining and evaluating SWAps: a paper for the Inter-Agency Group on SWAps and
Development
Centre for Health Economics
Appendix: model specifications
Has sector DAH increased under SWAp?
(1) sector_DAHit = αi + δSWApit-1 + β1malariait-1 + β2HIVit-1 + β3TBit-1 + β4IMRit-1
+ β5log(GDP/capita)it-1 + β6gov_effect it-1 + β7corruptionit-1 + μt + εit
Note: the impact on both absolute levels of Sector_DAH and also the proportion of
total DAH allocated to the sector is estimated.
Has SWAp changed allocations between other key health areas?
(2) tb_DAHit = αi + δSWApit-1 + β1malariait-1 + β2HIVit-1 + β3TBit-1 + β4IMRit-1
+ β5log(GDP/capita)it-1 + β6gov_effect it-1 + β7corruptionit-1 + μt + εit
Comparable specifications estimated for DAH directed to HIV, malaria and MNCH.
NCDs omitted due to lack of meaningful burden of disease control for full period.
Appendix: Table 4
Log of Levels of DAH to other health areas
Sector Support
Centre for Health Economics
(2)
(1)
(3)
(4)
malaria
HIV
TB
MNCH
Log(Sector DAH) t-1
0.022 (0.068)
-0.068 (0.041)
0.014 (0.051)
0.021 (0.032)
SWApt-1
0.460 (1.742)
-0.408 (0.608)
1.836 (1.293)
-1.257* (0.641)
SWAp*log(Sector DAH)t-1
-0.043 (0.128)
-0.130** (0.057)
-0.235* (0.117)
-0.055 (0.038)
0.026*** (0.009)
-0.001 (0.008)
0.025** (0.012)
0.003 (0.009)
HIVt-1 (% of 15-49yr olds)
0.377 (0.356)
0.009 (0.189)
0.313 (0.336)
-0.271** (0.131)
TBt-1 (incidence/100,000)
-0.004 (0.003)
0.004 (0.003)
-0.005 (0.005)
0.003 (0.003)
IMRt-1 (per 1,000 live births)
0.046 (0.037)
-0.026 (0.035)
0.040 (0.044)
-0.019 (0.037)
Log(GDP/capita) t-1
-1.133 (0.966)
1.395** (0.648)
-0.780 (0.624)
0.450 (0.806)
Gov – effectivenesst-1
2.704 (1.770)
4.196*** (1.053)
0.230 (1.534)
1.713 (1.141)
Control of corruptiont-1
1.351 (1.500)
-1.216 (1.011)
-0.702 (1.355)
1.144 (1.013)
773
773
773
773
displacement effect
^ Beta distributions
Standard errors in parentheses
* p < 0.10, ** p < 0.05, *** p < 0.01
Malariat-1 (prob death/1,000)
N
Appendix: Table 5
Impact of SWAp on Infant Mortality Rate
SWApt-1
Logged Infant Mortality Rate
(IMR)
Centre for Health
Economics
-0.086*** (0.029)
Log (DAH/capita) t-1
-0.002 (0.013)
Proportion of DAH to MNCH
0.095** (0.046)
Log(GDP/capita)t-1
0.070 (0.044)
Gov – effectivenesst-1
-0.074* (0.039)
TB (incidence/100,000)
0.000 (0.000)
Malaria (prob death/1,000)
0.001* (0.000)
HIV (% of 15-49yr olds)
Log (health expenditure/capita) t-1
-0.021** (0.010)
-0.032
N
570
SWAp countries
21
Control countries
18