Drought, Resilience, and Long-term Development

1. Drought, Resilience, and Long-term
Development
Alemayehu Seyoum Taffesse
International Food Policy Research Institute
The Future of Ethiopia’s Agriculture: Towards a Resilient System to
End Hunger and Undernutrition
Addis Ababa Hilton
December 15, 2017
Addis Ababa
1

2. 2
Outline
 Drought
 incidence and short-term impact
 response: Relief, PSNP (Address short-term impact)
 Resilience
 Definition
 PSNP
 Log-term development
 Premise – some long-term impact occur, vulnerabilities remain
(food gap, stunting), progress,
 Long term development – systems innovation;
* Source: Authors’ calculation using data on daily rainfall in millimetres (mm) extracted from the National
Aeronautics and Space Administration (NASA) website
(http://power.larc.nasa.gov/cgi.bin/cgiwrap/solar/hirestimeser.cgi?email=daily@larc.nasa.gov)

3. 3
Incidence and Impact of Drought
*Source: EM-DAT: The Emergency Events Database - Universite catholique de Louvain (UCL) - CRED, D.
Guha-Sapir - www.emdat.be, Brussels, Belgium (Created on: December 03, 2017)
Notes: Affected - People requiring immediate assistance during a period of emergency, i.e. requiring basic
Year Total affected
1965 1,500,000
1969 1,700,000
1973 3,000,000
1983 7,750,000
1987 7,000,000
1989 6,500,000
1997 986,200
1998
1999 4,900,000
2003 12,600,000
2005 2,600,000
2008 6,400,000
2009 6,200,000
2011 4,805,679
2012 1,000,000
2015 10,200,000
Incidence of Drought in Ethiopia (1965-2015)*
 Drought:
 Recurrent;
 Some widespread, others local;
 Appear to increase in frequency
(particularly in some parts of
the country);
 There are:
 other weather shocks –
flooding;
 non-weather shocks – pests,
human and animal health,
economic;

4. 4
Incidence and Impact of Drought
Source: Authors’ calculation using PSNP Surveys (2010-2016) data.
PSNP Woredas (2008-20016)
Shocks Incidence Impact
"Low-
lands"
"High-
lands"
"Low-
lands"
"High-
lands"
Drought 56.3 45.8
Loss of household income 25.7 34.9
Reduction in household consumption 25.4 32.8
Pests or diseases that affected livestock 23.2 8.4
Loss of household income 22.0 32.6
Reduction in household consumption 23.7 12.6
 Drought shocks have transitory and long-term impact;

5. 5
Incidence and Impact of Drought
 Evidence of longer-term effects based on the 1984/85 drought/famine -
Dercon (2004), Dercon and Porter (2014), Tafere (2017);
 10% lower rainfall today associated with 1 percentage point decline
in growth about 4–5 years later;
 Greater severity associated with worse outcomes – about 16
percentage points lower growth;
(Note: average CV of rainfall for the last 30 years in PSNP woredas
has been 1.35)* ;
 affected children aged 12-36 months are significantly shorter, by at
least 5 cm, as adults (20 years later);
 Second generation - Mothers’ exposure to famine in early childhood
has a negative effect on their children’s health (height-for-age z-
score), cognitive (number of years of schooling) and non-cognitive
(locus of control) human capital;
* Source: Authors’ calculation using data on daily rainfall in millimetres (mm) extracted from the National
Aeronautics and Space Administration (NASA) website
(http://power.larc.nasa.gov/cgi.bin/cgiwrap/solar/hirestimeser.cgi?email=daily@larc.nasa.gov)

6. Response to drought shocks
 Response has evolved over time, still is;
 Relief – annual appeals system, shortcomings (delay, lack
of predictability);
 PSNP – coordinated, multi-year, predictable, ‘productive’
 John’s and Kalle’s presentation on the nature and
impact of PSNP
 Part of comprehensive development planning;
 Broad social protection agenda;

7. Resilience
 Multiple conceptualisations of resilience and vulnerability;
 Resilience
 a recovery trajectory following a specific shock.
 an ex-post approach (study reported below):
 Impact of PSNP
 beneficiaries recover after no more than 2 years, rather
than taking up to four years) – resilience;
 Considerable food gap remains;

8. PSNP, vulnerability, and resilience
Findings – PSNP transfers reduce vulnerability and improve
resilience (2006-2014)

9. 9
Vulnerabilities remain
Source: Authors’ calculation using DHS (2005-2016) data. AEZ classification as earlier presentations.
Stunting Incidence among Children Under 5
(%)
2005 2016
AEZ zone Non-PSNP PSNP Non-PSNP PSNP
Drought prone 46 47 37 33
Pastoralist 50 60 35 34
Humid moisture reliable – Lowland 49 61 34 38
Moisture reliable – Cereal 50 55 38 25
Moisture reliable – Enset 55 42 38 40
Two examples:
 Up to 75% more vulnerable people in drylands in 2030 (2010=100, medium
fertility scenario)
 Child Stunting;

10. 10
Long-term development
 Premise
 Progress - rapid agricultural growth, diversification into non-agriculture;
 Accelerating agricultural transformation still necessary, doing so gets
harder given the achievements so far;
 There are compounding factors – environmental degradation, climate
change, population growth;
 Degradation:
o Loss of top soil – over 14 million hectares of top soil have less than 50
cm depth, making it susceptible to drought.
o Soil nutrient depletion – equivalent of 30 kg/ha of Nitrogen (N), and 15
to 20 kg/ha of phosphorous (P) are lost annually through erosion on
cultivated lands.
o the average annual on-site cost of soil erosion has been estimated
between 2 and 6.7 % of AGDP’ (see Yusef et al. (2005) )

11. 11
Degradation

12. 12
Long-term development
 Focus – drought-prone or drylands (AI of 0.05-0.65)
 System perspective – challenges and opportunities should
be viewed from a single economic space perspective –
interlinked parts;
 More integrated economy – greater integration is the most
feasible avenue with the potential to generate
enduring/sustainable solutions to the development
challenge;

13. 13
Long-term development
 Radical system-wide innovations necessary;
 A 20-25 year programme to transform and integrate the drylands;
 Food and nutrition security as an outcome of an integrated
economy
 An idea:
 Degraded highlands – encourage a shift into tree crops and semi-
modern dairy farming (with fodder production as a new and
growing activity);
 Lowlands – encourage expansion of cereal production, mobile
livestock rearing;

14. 14
Long-term development
 Some implications:
 Invest on human capital – mobile schools and clinics;
 Invest on water management in all areas – irrigation, better water
use arrangements;
 Invest on disease (human and animal) control and in less moisture
reliable low lands;
 Enhance the role of the private sector in agricultural input
provision;
 Promote well-defined and focused cooperatives, particularly in
dairy production and development;
 Potential - there appear to be considerable potential yet to be deployed
productively in parts of the country;
 Land – (next slide)
 Water – renewable internal freshwater resources

15. 15
Cropped Land by AEZ (2004-2016)
Source: Authors’ calculation using CSA’s AgSS (2004-2016) data. AEZ classification as earlier presentations.
AEZ
share in
total
surface
area
(%)
AEZ
cropped area
in total
cropped area
(%)
Cropped area
share in total
AEZ surface
area (%)
Number of
woreda
Drought prone, Highland 11.4 23.2 23.5 146
Drought prone, lowland; Pastoralist 50.9 3.5 0.8 139
Humid moisture reliable, lowland 11.9 4.4 4.3 52
Moisture reliable, highland - Cereal 20.1 57.5 32.9 264
Moisture reliable, highland - Enset 5.8 11.4 22.7 126

16. 16
Thank You

17. 17
Impact of PSNP on the Local Economy
 PSNP has two components: Cash Transfer (CT) and Public Works
(PW)
 Community assets - Soil and water conservation (SWC), Irrigation,
Roads, Schools, Clinics…
 Most evaluation focuses on recipient households
 But impacts may spread far beyond recipients:
o Cash Recipients spend their money within their economy;
o Public works affect agro-ecological and economic
environment;
 Implication
 Need to evaluate the FULL impact;
 Full impacts locally, and nationwide

18. PSNP, vulnerability, and resilience
Findings – Impact (2006-2014)
 PSNP transfers improve vulnerability and resilience
 reduce food gap by half a month – ‘poverty’;
 reduce the expected food gap, given a drought has
occurred, from 4.14 months to 1.8 months (57%
decline) – vulnerability;
 reduce food gap by a further 1.75 months after a
drought (beneficiaries recover after no more than 2
years, rather than taking up to four years) – resilience;
 Robustness Check – Climate data

19. Impact of the PSNP on Productivity and Growth
19
 Yield growth:
 PSNP-related SWC infrastructures enhanced crop yields by
2.8 percent on average (econometric analysis);
 Local Economy Impact:
 PSNP generated income multipliers ranging from 1 to 2.4
ETB per ETB transferred depending on the kebele (eight
LEWIE models);

20.  Significant productivity and growth challenges continue;
Some findings
20
0.14% 0.14%
1.23% 1.23%
0.99%
0.00%
0.50%
1.00%
1.50%
2.00%
2.50%
Cost Benefit
GOE cost Donor cost value created
Approximate costs and benefits of PSNP nationwide (% of GDP)
 National Economy Impact: CGE estimates

21. Some Observations
 There were/are challenges in the implementation of PSNP:
 Arrears – early years, by and large successfully tackled;
 Targeting in the ‘lowlands’ component – weak;
 Graduation – need to know/understand more;
 Considerable food security and vulnerability remains;
 PSNP as a platform for response to shocks (relief delivery);
 PSNP and regular social protection and capacity regeneration;
 Development and resilience –
 livelihood diversification – sedentary agricultural, non-farm
activities…;
 Investments in irrigation, education and health interventions;
 Complementarity – addressing emergency, enhancing
resilience, and promoting development