Bedru Balana, Research Fellow, IFPRI, presented these slides at the AAAE2023 Conference, Durban, South Africa, 18-21 September 2023. The authors acknowledged the contributions of CGIAR Initiative on National Policies and Strategies, Google, the International Rescue Committee, IFPRI, and USAID.
Scaling up coastal adaptation in Maldives through the NAP process
Anticipatory cash for climate resilience
1. Anticipatory Cash for
Climate Resilience
Findings from a randomized experiment,
NE Nigeria
Presented by Bedru Balana
AAAE2023 conference
Durban, South Africa • 18-21 Sep. 2023
2. Anticipatory cash for climate resilience:
Findings from a randomized experiment, NE Nigeria
Presenter: Bedru Balana (IFPRI)
Co-authors/Contributors: Dolapo Adeyanju1, Alan de Brauw1, Kwaw Andam1,
Clare Clingain2, Olukunbi Olarewaju2, Ishaku Yohanna2, & Molly Schneider2
1 International Food Policy Research Institute (IFPRI)
2 International Rescue Committee (IRC)
AAAE2023 conference, Republic of South Africa
Durban| 18-21 Sep. 2023
5. Background
• Nigeria is highly vulnerable to climate-related
shocks and ranks high among the countries that
are susceptible to climate change: perennial
flooding
• 2012 flood - over 2.3 million people displaced, 363
deaths, and impacted the livelihoods of over 16
million people
• 2022 flood - over 1.4 million people displaced, over
603 deaths, injured over 2,400 persons, over
200,000 homes completely/partially destroyed, and
332,327 hectares of land affected (Oguntola, 2022)
Photo credit: Daily post (13 Oct. 2022)
Photo credit: Premium Times (27 Oct. 2012)
6. Background – (contd.)
• Northeast Nigeria: decades of climatic shocks, security
threats, and the consequent humanitarian crisis.
• In 2022, faced the worst floods in seven years, which
affected over 300,000 people and with casualties five times
more than those reported in previous years.
• Adamawa state among the most affected states with an
estimated 260,000 people exposed to floods and around a
1,000 km.sq. of land submerged (REACH Initiative, 2022).
• Besides climate variability, protracted and frequent violence
conflicts further disrupt agricultural activities and food
supply chains
• Limited coping strategies among poorer population
Photo credit: Early warning volunteers
(Fufore, Adamawa: Jul.2022)
7. Anticipatory vs. Post-shocks Action
• Most humanitarian supports largely focus on responding to post-
shock crises - targeting vulnerable households
(food aids, temporary shelter, or basic services, one-off or monthly
cash transfers)
• Anticipatory actions can potentially help household to better
prepare and respond to climate shocks.
• Project Goal: Testing anticipatory vs. post-shock cash transfers:
• In reducing the use of negative coping strategies and support households
building resilience against climate shocks by providing climate risk
reduction payments and early warning messaging to smallholder farmers.
8. Evidence gaps
Feasibility: limited evidence on feasibility of ex-ante climate risk payments
Effectiveness: “small” cash transfers have been found to be effective, for some
outcomes, yet may hinder overall effectiveness, larger transfers should be evaluated
Cost-Efficiency: general feeling that anticipation is more cost-efficient than post-ante,
yet cost-efficiency data is not available for most programs
Impacts: there is limited data on impact disaggregated by gender or other sub-
populations
Scalability: limited integration cross-organization or between organizations and
governments of anticipatory services
12. Hypothesis:
When climate vulnerable communities have timely access to information and the financial
and resources to act upon that information, they will avoid negative coping strategies,
build more diversified and climate resilient livelihoods.
Experiment/Intervention:
(1) Provision of information – early warning messages (flood alert, disaster
risk/emergency preparedness, disaster management and CSA)
(2) Transfer of a lump sum of anticipatory cash payments to a sample of 725 flood-prone
households when triggered by the climate data risk thresholds, and same No. HHs
(‘control group’) received equal amount of cash post flooding shock.
• Treatment group: ₦ 195, 000 (~400 USD) anticipatory cash to 725 households (27
Jul.2022)
• Control group: ₦ 195, 000 (~400 USD) post-shock cash to 725 households (7
Nov.2022)
13. Where? Fufore LGA
Intervention site
What climatic shock? Flooding
Why Fufore? Situated close to the River Benue
floodplain, Faro River & exposed to the spillover from
the Ladgo dam in Cameroon.
Why Flooding?
(1) Flooding has high impact on the livelihoods and frequency
of occurrence.
(2) Flooding is predictable climatic shock with well-established
prediction data compared to other climate shocks.
14. • We conducted blanket registration of
all eligible households in six
communities. A total of 2095 eligible
households
• 1450 study households were
randomly selected with proportional
allocation by community and gender
• Baseline data collected between 25
April and 31 May 2022
• Endline data collected 5 months
after the baseline (between 1 Dec
and 9 Dec 2022)
Study communities
Gender
total
Dasin
Hausa
Gembeusi
Ribado
Rico
Farang
farang
Dullo
Bwatiye
803
245
138
238
60
41
81
Male
647
132
138
211
13
31
122
Female
1450
377
276
449
73
72
203
Total
Sampling Frame & Sample selection
18. Key Outcomes
Outcome1:
Increase in climate
adaptive actions
Number of pre-
emptive
actions/practices
adopted.
Household labor re-
allocation in
response to flooding
(including migration)
Number of post-
shock actions taken
Outcome 2:
Increase in climate
resilience
Livelihood
diversification
Productive
Investment
(agricultural assets)
Productive
Investment
(productive
livestock)
Outcome 3:
Reduction/avoidance in
negative coping
strategies
Reduced Coping
Strategy Index (rCSI)
Livelihood Coping
strategy Index (LCSI)
Outcome 4:
Household welfare
Food consumption
score (FCS)
Wellbeing indicator
measured using an
11-point scale (0-10)
Cantril's ladder of
life satisfaction
(‘scarcity mindset’)
19. Other Outcomes
Household aggregates
Non-transfer income (non-agricultural)
from non-farm employment and small
businesses
Agricultural income (crop production
and livestock incomes)
Number of livestock (TLU)
Value of durables household assets
Food vs. Non-food consumption
Expenditure on food items
Expenditure on non-food household
consumables
20. Estimation strategy
(1) Average treatment effect of anticipatory cash (without controls)
Considering treatment (i.e., anticipatory cash payments) as a randomized control trial
(RCT), an unbiased treatment impact is estimated using a standard linear model as
expressed in equation 1
𝑌 = 𝜑 + 𝛽. 𝑝𝑟𝑒𝑠ℎ𝑜𝑐𝑘 +𝜓 . 𝑣 + 𝜀 ------------------------------(1)
(2) Average treatment effect of anticipatory cash (with controls)
We augmented equation 1 by adding observable covariates/controls.
𝑌 = 𝜑 + 𝛽. 𝑝𝑟𝑒𝑠ℎ𝑜𝑐𝑘 +𝛾. 𝑋 + 𝜓 . 𝑣 + 𝜀 ------------------------------(2)
(3) Heterogeneity analysis by wealth index
𝑌 = 𝜑 + 𝛽 . 𝑇 +𝛽 . 𝑊𝐼 + 𝛽 . 𝑊𝐼 ∗ 𝑇 + 𝛽 . 𝑊𝐼 + 𝛽 . 𝑊𝐼 ∗ 𝑇 + 𝛽 . 𝑊𝐼 + 𝛽 . 𝑊𝐼 ∗
𝑇 + 𝛽 . 𝑊𝐼 + 𝛽 . 𝑊𝐼 ∗ 𝑇 + 𝛾. 𝑋 + 𝜓 . 𝑣 + 𝜀 ..…………………….…(3)
21. (1) Impacts on Consumption (FCS) and coping strategies (rCSI & LCSI) (n=1447)
22. (2) Impacts on livelihood diversifications
non-farm
activities
Mixed crop-
livestock
Diversify
crop
production
Labor re-allocation
(inc. migration)
Margins
Coef. (dy/dx)
0.132
0.087
0.074
0.034*
0.236*
Treatment (anticipatory)
(0.114)
(0.107)
(0.115)
(0.019)
(0.142)
-0.886***
-0.466**
0.848***
-
-1.753***
Cons.
(0.081)
(0.076)
(0.081)
-
(0.104)
No
No
No
No
No
Household level controls?
No
No
No
No
No
Community fixed effects?
• Anticipatory cash has significant effect on household labor re-
allocation decisions inclu. migration of HH members (at 10% level)
• Anticipatory cash is not statistically significant on livelihood
diversification indicators (crop diversification, mixed crop-
livestock production, and engagement in non-farm activities)
23. (3) Impacts on household climate adaptative actions
Number of post-shock actions
Number of pre-emptive actions
(Margins,
(dy/dx))
(Poisson)
(OLS)
(Margins,
(dy/dx))
(Poisson)
(OLS)
0.106
0.052
0.106
0.169**
0.081**
0.175*
Treatment (Pre-shock
cash)
(0.075)
(0.036)
(0.082)
(0.075)
(0.036)
(0.101)
--
0.753***
1.889***
--
0.688***
2.136***
Cons.
--
(0.082)
(0.185)
--
(0.026)
(0.232)
No
No
No
No
No
No
Household level controls?
Yes
Yes
Yes
Yes
Yes
Yes
Community fixed effects?
• Pre-shock cash payment has significant impacts on the number of
pre-emptive actions households took in anticipation of flood
shocks.
• No significant differences between the pre- and post-shock cash
recipients in terms of post-shock actions taken.
24. (4) Impacts on household aggregates
Durable
assets
(N)
Livestock
numbers
Agriculture
income
Non-transfer &
non-agric.
income
0.135
0.437
50738*
0.534**
Treatment (Pre-flood cash recipient)
(0.232)
(0.293)
(21836)
(0.193)
5.921
1.192**
-338005***
3.546***
Cons.
(0.129)
(0.362)
(10614)
(0.094)
Yes
Yes
Yes
Yes
Initial conditions (lagged values)?
Yes
Yes
Yes
Yes
Community fixed effects?
• Anticipatory cash transfer has statistically significant effect on
seasonal non-transfer incomes (non-farm empl. & small business) and
incomes from agriculture (crops and livestock) measured at endline.
• No significant difference in the possession of durable assets and
number of livestock measured at the endline between the anticipatory
and post-shock cash recipients.
25. (5) Impacts on HH’s long-term climate resilience
Utilization of IRC cash transfer
Investment Expenditure
Consumption
Expenditure
Nonfarm
small
business
Invested on
productive
livestock
Invested on
productive
agric. assets
Nonfood
household
expenditures
Food
expenditure
-0.173
0.352**
0.496***
-0.123
-0.368
Treatment (anticipatory cash)
(0.153)
(0.146)
(0.106)
(0.112)
(0.352)
-1.763***
-1.878***
-0.490***
0.783***
3.926***
Cons.
(0.105)
(0.109)
(0.076)
(0.080)
(0.269)
Yes
Yes
Yes
Yes
Yes
Community fixed effects?
0.259
0.016
0.000
0.271
0.297
P-values (treatment var)
• No significant differences between pre- and post-shock cash
recipients on food purchases and expenditures on non-food household
items.
• Anticipatory cash transfer has statistically significant effect on
long-term investment decisions on productive agric. assets and
productive livestock.
27. Conclusions
Households who received anticipatory cash were less likely to employ negative coping
strategies in response to the flood shocks (rCSI, LCSI).
Anticipatory cash payment had significant impacts on the number of pre-emptive actions
taken by households in anticipation of flood shocks.
Anticipatory cash has positive and statistically significant impacts on productive investments
(agricultural assets and productive livestock). Productive investments could enhance
household’s future income generating capacity and reduce their vulnerability to future shocks.
Thus, anticipatory cash may support long-term resilience.
Wellbeing (life satisfaction) – we couldn’t find significant differences in wellbeing measures
between the pre- and post-cash households.
Heterogeneity analyses showed that the impacts of anticipatory cash on various outcomes
measures were not sensitive to the wealth status (wealth quintiles) of households, which may
be linked to the general poverty rates in the study community.