Presentation by John Hoddinott at the event, “2013 AAEA & CAES Joint Annual Meeting” which took place on August 4-6, 2013 in Washington, DC. It offers AAEA members, CAES members, and other applied economists a chance to interact and learn over the course of the three day meeting.
2. Literature & Research Questions
• Food versus Cash transfers
• Timeliness advantages, and in some cases cost (Gentilini 2007;
Lentz et al. forthcoming).
• Key advantage of cash is that it provides recipients with choice
• Several contributions
• Randomized design
• Important setting for understanding impacts of cash transfers
• Geography (land‐locked), poverty, food insecurity
• Functioning food markets
• Government and agency movement toward cash transfers
• Food security analysis
3. Context
• Niger, and the intervention region
• Fifth poorest in per capita GNI; 186/187 on the HDI; 93% suffering
from deprivation
• Food insecure (availability, access, and use); severe food crises in
parts of country in 2005‐2006, 2010, and again in 2012
• Zinder; surplus production region, yet often hardest hit by food
crises
• The Project (2011)
• Large‐scale cash/food pilot implemented by the World Food
Programme
• Cash/food for work (April‐June), followed by unconditional
transfers (July‐Sept)
• 126 villages in 12 departments of Mirriah, Zinder
• Food basket, and cash transfer (of equivalent value)
5. Research Design
• Randomization at worksite level (52 worksites)
• Two survey rounds:
• Following public works, all households in all evaluated villages (5,668
households)
• Follow‐up with unconditional transfer recipients (2,268 households)
Mirriah District
(52 worksites)
Agricultural Zone
(29 sites)
RANDOMIZE
Cash for work
(15 sites, 1747 HHs)
Cash transfers
(15 sites, 686 HHs)
Food for work
(14 sites, 1658 HHs)
Food transfers
(14 sites, 635 HHs)
Agro‐Pastoral Zone
(23 sites)
RANDOMIZE
Cash for work
(12 sites, 1202 HHs)
Cash transfers
(12 sites, 493 HHs)
Food for work
(11 sites, 1061 HHs)
Food Transfers
(11 sites, 395 HHs)
7. Outcome Variable Definitions
• Foods and food groups consumed
• How many occasions in the past 7 days
• Household Dietary Diversity Index (HDDI)
• Sum of different foods consumed (1 to 25)
• Household Dietary Diversity Score (HDDS)
• Sum of different food groups consumed (1 to 11)
• Food Consumption Score (FCS)
• Weighted sum of food groups consumed, based on dietary quality
• Categories: Poor (≤ 21), Borderline (> 21, ≤ 35), Acceptable (> 35)
• Coping Strategies Index (CSI)
• Index based on reliance on a diverse set of coping strategies
• Food and Non‐Food Expenditures
8. Method
• A single‐difference estimator of the form:
,
, 1
• Lack of baseline => cannot estimate a dif‐in‐dif; however, dif‐
in‐dif only preferable when autocorrelation of outcomes is
high (which it is not in this case)
9. Results: Food Diversity
Food Security Outcomes (Table 4) July October
Dietary Diversity Index (DDI) 0.356* 0.544**
(0.207) (0.229)
Food Consumption Score (FCS) 3.923*** 4.647***
(1.424) (1.139)
“Acceptable” FCS (WFP cut‐off) 0.109** 0.121***
(0.043) (0.041)
Number of Households 2256 2187
Standard errors (clustered at worksite level) in parentheses
*, **, and *** indicate significance at the 10%, 5%, and 1% level, respectively
Controls included (but not shown): age, sex, education, and ethnicity of household head; household
size; asset score; whether in pastoral zone; village access to market, health clinic, and mobile phone
coverage; distance to the main road; livestock prices; change in millet price during period; millet price
at end of period; commune‐level fixed effects.
10. Results: Coping Strategies
CSI and Selected Coping Strategies (Table 7) July October
Coping Strategies Index ‐3.708* ‐3.168***
(1.916) (0.411)
Relied on less‐preferred foods ‐0.039* 0.024
(0.022) (0.020)
Borrowed food from relatives or neighbors ‐0.082*** ‐0.022
(0.024) (0.021)
Had to cancel debt repayments ‐0.038** 0.057***
(0.017) (0.009)
Reduced number of meals per day ‐0.025 ‐0.036**
(0.024) (0.015)
Number of households 2256 2187
Notes: see Table 4 (previous slide)
11. Results: Food Consumption and Expenditures
Food Groups (Table 5) Were items consumed Number of days consumed
July October July October
Cereals ‐‐ ‐‐ 0.093* 0.109***
(0.051) (0.035)
Pulses 0.064** 0.021 0.638** 0.820***
(0.032) (0.013) (0.314) (0.168)
Oils 0.106*** 0.042** 0.959*** 1.010***
(0.033) (0.017) (0.258) (0.186)
Food Purchases (Table 6) Made purchase Expenditure (CFA)
July October July October
Bulk Grain purchases ‐0.273*** ‐0.400*** ‐14289*** ‐25015***
(0.020) (0.034) (1570) (432)
Consumption/purchase estimated using a Probit model; number of days using a Poisson; expenditures
using a Tobit. Results reported are marginal effects.
Additional notes in Table 4 (above).
12. Results: Non‐Food Expenditures
Non‐Food Purchases (Table 6) Made purchase Expenditure (CFA monthly)
July October July October
Total non‐food expenditure ‐‐ ‐‐ 1874.7*** ‐592.0
(502) (1010.7)
Construction, repair, housing ‐0.034* 0.002 ‐2870.8* 495.2
(0.021) (0.016) (1686.3) (403.9)
Wages, animal care, seeds ‐0.105*** ‐0.090*** ‐1779** ‐5819**
(0.035) (0.029) (816.2) (2604)
Purchase estimated using a Probit model; expenditures using a Tobit. Results reported are marginal
effects.
Additional notes in Table 4 (above).