Transcript PPT

Impacts of CAADP on
Africa’s Agricultural-Led
Development
Sam Benin, IFPRI
IFPRI Discussion Paper 01553
August 2016
Impacts of CAADP on Africa’s Agricultural-Led
Development
Download at:
https://www.ifpri.org/publica
tion/impacts-caadp-africasagricultural-led-development
Samuel Benin
or send me email
[email protected]
Development Strategy and Governance Division
Introduction and objectives
• Learned about growth and poverty reduction in
Africa in recent years, and CAADP is implicated
• Question: how has CAADP actually contributed to
these achievements? What are the impacts?
• Objective of study: assess impact of CAADP on:
 Government agriculture expenditure, agricultural
growth and productivity, income, and nutrition
Fundamentals of impact evaluation
direct effect
indirect effect
CAADP
XCAADP
Y
XCAADP&OUT
Outcome
XOUT
Total Effect = direct effect + indirect effect
(control for XCAADP, XOUT, XCAADP&OUT)
Key Assumption: Xi is known, observed, and used
CAADP country-level process
& conceptual framework
A
Launch of
CAADP
G
Joint sector review
& mutual
accountability
F
B
Analysis of growth
options, investment, &
capacity needs
C
Financing and
implementation of
plan and programs
E
Preparation of
investment plan &
programs
D
Consultations with
stakeholders and
validation of results
Preparation and
signing of compact
by all stakeholders
CAADP country-level process
& conceptual framework
A
Launch of
CAADP
G
Joint sector review
& mutual
accountability
F
B
Analysis of growth
options, investment, &
capacity needs
Assumption
CAADP involves processes and
C to manifest.
actions
that
take
time
Consultations with
Financing and
The longer
or more intensive
a
stakeholders and
implementation
of
of results
plancountry
and programs
engages, the validation
greater
the
likelihood
ofDsuccess
E Preparation
of
Preparation and
investment plan &
programs
signing of compact
by all stakeholders
Two definitions
of CAADP:
1. Whether
signed
compact
(0=no, 1=yes)
2. Level reached:
0=precompact
1=compact
2=NAIP
3=1 ext fund
4=>1 ext fund
Concepts and methods
• Identify factors that determine a country’s decision
to implement CAADP (d):
 whether it signs a CAADP compact (d1 = 1,0)
 level of implementation reached (d2 = 0,1,2,3,4)
• Controlling for above factors as well as those that
affect realization of outcomes, estimate impact of
implementing CAADP on annual change in:
 Agricultural performance: agriculture expenditure,
agricultural growth and land & labor productivity
 Broader outcomes: income (GDP per capita), nutrition
(prevalence of adult undernourishment)
Influential factors and hypothesis
Conceptual factor (expected)
Variables and measures
Relevance of CAADP/
Importance of agriculture (+)
Share of agricultural value added in total GDP,
share of agricultural area in total area
Political will (+)
Number of AU charters/treaties ratified by 2003
Peer pressure (+)
Share of bordering countries at next stage of
CAADP implementation (physical or REC)
Negotiation posture (-)
Total expenditure per capita, share of GDP in
Africa’s total GDP
Capacity of government (+)
Cumulative years agricultural minister in place
Demand and capacity of
citizens (+/-)
Voice and accountability index (-2.5 to 2.5),
autocracy-democracy index (-10 to 10)
Pretreatment outcomes
Lagged values of the outcome variables
Cross-country effects
Population density, rainfall, AEZ-economic class
Global effects
Financial crisis (0 up to 2008, 1 after 2008)
Data sources and estimation
• Data from various international and national
sources from 2001 to 2014
• Use panel-data regression methods to estimate
treatment effects of CAADP and deal with several
relevant econometric issues
• Use different model specifications to evaluate
sensitivity of results to different issues and
assumptions  generate greater confidence in
results
Distribution of countries (46)
Signed CAADP compact in:
2007–
2010–
2013–
Not
2009
2012
2014
signed
(13)
(16)
(8)
(9)
Benin
Burundi
Cape Verde
Ethiopia
Gambia
Ghana
Liberia
Mali
Niger
Nigeria
Rwanda
S. Leone
Togo
Burkina Faso
Cent Afr Rep
Congo, D.R.
Côte d’Ivoire
Djibouti
Guinea
G-Bissau
Kenya
Malawi
Mauritania
Mozambique
Senegal
Swaziland
Tanzania
Uganda
Zambia
Angola
Cameroon
Chad
Congo, R.
Lesotho
Madagascar
Sudan
Zimbabwe
Algeria
Botswana
Egypt
Eritrea
Mauritius
Morocco
Namibia
S. Africa
Tunisia
Level implementation reached by end of 2014
Level 0 Level 1 Level 2 Level 3 Level 4
(9)
(8)
(8)
(9)
(12)
Algeria
Botswana
Egypt
Eritrea
Mauritius
Morocco
Namibia
S. Africa
Tunisia
Angola
Chad
Congo, R.
Lesotho
Madagascar
Sudan
Swaziland
Zimbabwe
Cameroon
Cape Verde
Cent Afr Rep
Congo, D.R.
Djibouti
Guinea
G-Bissau
Mauritania
Burundi
Gambia
Liberia
Mali
Niger
S. Leone
Togo
Uganda
Zambia
Benin
Burkina Faso
Côte d’Ivoire
Ethiopia
Ghana
Kenya
Malawi
Mozambique
Nigeria
Rwanda
Senegal
Tanzania
Determinants of CAADP implementation
Signed compact (logit)
Level reached (ologit)
Model 1
Model 2
Model 1
Model 2
Importance of agriculture
0.57 ***
0.48 ***
-0.01
0.02
Political will
0.70 **
0.63 **
-6.69
-6.38
Peer pressure
0.06 ***
0.11 ***
0.00
0.02 ***
1.47
1.56
Negotiation posture
-1.49 **
-1.56 **
Government capacity
3.71 ***
5.38 ***
2.11 ***
2.27 ***
Citizens’ demand&capacity
0.58
0.80
0.63
0.16
Financial crisis
Population density
Intercept
Chi-square statistic
23.24 ***
0.03 ***
4.86 ***
0.02 ***
-75.72 *** -65.06 ***
58.85 ***
81.28 ***
0.06 *
0.11 ***
n.a.
n.a.
803.2 ***
769.0 ***
*, **, and *** = statistical significance at the 10%, 5%, and 1%, respectively
Determinants of CAADP implementation
Signed compact (logit)
Level reached (ologit)
• Regarding compact
signing,
variables
representing:
Model
1
Model
2
Model
1
Model 2
Importance
of agriculture
0.57
***
0.48 peer
*** pressure,
-0.01
 Role
of agriculture,
political
will,
0.02
government
Political will
capacity,
have positive
0.70 and
** financial
0.63 ** crisis
-6.69
-6.38
and statistically significant
Peer pressure
0.06 *** influence
0.11 ***
0.00
0.02 ***
 Negotiation
posture
influence,
Negotiation
posture
-1.49has
** negative
-1.56 **
1.47 likely due
1.56 to
alternative
sources
development
Government
capacity (non-agriculture)
3.71 ***
5.38
*** of 2.11
***
2.27 ***
 demand&capacity
Citizens’ demands0.58
and capacity
significant 0.16
Citizens’
0.80 are not 0.63
Population
densityof
• For level
0.03 ***
0.02
***
0.06
implementation
reached:
*
0.11 ***
Financial
crisis
 Only
23.24
*** of implementation
4.86 ***
peer pressure
(stage
of neighbor)
Intercept and government-75.72
*** (how
-65.06long
*** minister
n.a. of
n.a.
capacity
Chi-square
statistic
58.85 in***
81.28
803.2 ***
769.0 ***
agriculture
has been
place)
are***
important
*, **, and *** = statistical significance at the 10%, 5%, and 1%, respectively
Est. impacts of CAADP, % change (2001-14)
Annual change in:
Ag expenditure as %
of total expenditure
Ag expenditure as %
of ag value added
Agricultural value
added per hectare
Agricultural value
added per worker
Agricultural value
added
GDP per capita
Prevalence of adult
undernourishment
Signed
compact
Level of implementation achieved
Level 1
Level 2
Level 3
Level 4
-3.6
-4.0
-5.1
6.3 change
-23.0***
Interpretation:
percentage
in the-5.0
outcome-7.8
in countries
that-20.1**
are
-1.9
-2.3
implementing CAADP, compared to
-6.5 the general
12.9** trend
-7.0*in countries
8.3** that
16.5***
are8.7*
not implementing
CAADP11.6***
-4.1
-9.0**
3.6
4.9
8.6**
7.4**
10.8**
16.7***
-0.6
-1.3
-0.3
-0.2
1.5
1.3
1.1
1.1
3.0
-0.1
*, **, and *** = statistical significance at the 10%, 5%, and 1%, respectively
Est. impacts of CAADP, % change (2001-14)
Level of implementation achieved
Level 1
Level 2
Level 3
Level 4
-3.6
-4.0
-5.1
6.3
-23.0***
-1.9
-5.0
-7.8
-2.3
-20.1**
Puzzling impacts
-6.5
12.9**
-7.0*
8.3**
16.5***
-4.1
8.7*
-9.0**
3.6
11.6***
4.9
8.6**
7.4**
10.8**
16.7***
-0.6
-1.3
-0.3
-0.2
1.5
1.3
1.1
1.1
3.0
-0.1
*, **, and *** = statistical significance at the 10%, 5%, and 1%, respectively
Largest significant impacts
Ag expenditure as %
of total expenditure
Ag expenditure as %
of ag value added
Agricultural value
added per hectare
Agricultural value
added per worker
Agricultural value
added
GDP per capita
Prevalence of adult
undernourishment
Signed
compact
No significant impacts
Annual change in:
Est. impacts of CAADP, % change (2001-14)
*, **, and *** = statistical significance at the 10%, 5%, and 1%, respectively
Largest significant impacts
Signed
Level of implementation achieved
• Signing a compact
alone
has
no significant impact
compact
Level 1
Level 2
Level 3
Level 4
Negative
impact on expenditure  substitution effect,
Ag• expenditure
as %
-3.6
-5.1
6.3
-23.0***
of total
expenditure
largest
for level
4, which-4.0has more
than one
external
Ag expenditure
sourcesasof% funding
-1.9
-5.0
-7.8
-2.3
-20.1**
of ag value added
Puzzling impacts
Agricultural
valueimpacts on agricultural value added:
• Positive
-6.5
12.9**
-7.0*
8.3** 16.5***
addedlevel
per hectare
4=17%;
level
3=11%;
level
2=7%;
level
1=9%
Agricultural value
-4.1
8.7*
-9.0**
3.6
11.6***
added
per
worker
• Mixed impact on land and labor productivity: positive,
Agricultural value
but negative for
countries)
4.9level 2 (small
8.6** number
7.4** of 10.8**
16.7***
added
GDP
capita insignificant
-0.6
-1.3 on income
-0.3
-0.2nutrition
1.5 
• per
General
impact
and
Prevalence
of adult
positive
gains in
yet
1.3production/productivity
1.1
1.1
3.0 to translate
-0.1
undernourishment
into broader positive outcomes
No significant impacts
Annual change in:
Est. impacts of CAADP, % change (2001-14)
Indicator
Year of compact
GDP per capita
2007-2009
2010-2014
Undernourishment
2007-2009
2010-2014
Level of implementation achieved
Level 1
Level 2
Level 3
Level 4
-0.8
-1.4
1.7**
-1.1***
-0.6
0.2
2.0*
0.0
-0.5
0.2
4.8**
0.9
3.7**
0.0
2.4*
-0.9**
*, **, and *** = statistical significance at the 10%, 5%, and 1%, respectively
• Strongly positive impact on income for early implementers at
level 4  broader benefits of reforms take time to manifest
• Counterintuitive impact on nutrition reflects weaker
emphasis on nutrition in early NAIPs compared to later
NAFSIPs (FS = food security)
Overall implications
• Because CAADP is a framework for inclusive participation,
ownership, evidence-based policy making, and donor
alignment for an agricultural-led development
 it takes time to gain buy-in from all stakeholders to safeguard
successful implementation
 as such, finding a shortcut is unlikely
• We can expect (greater) benefits from processes that
include a systematic effort to
 identify strategies that are likely to work (as expected of the growth
options and investment and capacity requirements analyses)
 articulate those strategies in a plan that is adequately funded and
implemented accordingly
 to monitor and evaluate progress to continuously refine the
investments and programs