Rachel Glennerster IGC Lead Academic for Sierra Leone and JPAL Joint work with Tavneet Suri IGC Agriculture and MIT Sloan Overview The importance of good data in a crisis Initial sectors of concern and policy focus not supported by the emerging evidence ID: 693401
Download Presentation The PPT/PDF document "Economic l essons from crisis" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Economic lessons from crisis
Rachel Glennerster (IGC Lead Academic for Sierra Leone and JPAL)Joint work with Tavneet Suri (IGC Agriculture and MIT Sloan)Slide2
Overview
The importance of good data in a crisisInitial sectors of concern and policy focus, not supported by the emerging evidence
Economic impacts were not concentrated where the
disease was
Impact of food aid during the crisisThe role of researchers during a crisisMaintaining confidence while calling for help
2Slide3
Ebola cases end June, Sierra LeoneSlide4
Ebola cases, cordon areas, end
AugSlide5
Ebola cases, cordon areas, end
SeptSlide6
Ebola cases, cordon areas, end
OctSlide7
Ebola cases, cordon areas, end
NovSlide8
Ebola cases, cordon areas, end
DecSlide9
Ebola cases, cordon areas, end
JanSlide10
Ebola cases, cordon areas, mid
MaySlide11
Incentive is to grab headlines
WHO “up to 90%” death rate from EbolaMin of Ag, GDP
had already fallen by 30%
GDP fell by
av of 5% per year during devastating civil warSept 5, FAO reported 40% of farms in Kailahun abandoned, and lack planting materialsIn sept rice is maturing in fields: what does abandoned mean?Planting pre Ebola, would not take place again till May 2015
Reports of skyrocketing food prices
Restrictions on transport did make this a concernSlide12
Messaging ignored basic psychologySlide13
Call reps at 208
markets, 1-2 times monthSource:
Glennerster and
Suri
, 2015 www.theigc/country/sierra-leone Slide14
Food prices similar to 2011/12
Source:
Glennerster and
Suri
, 2015 www.theigc/country/sierra-leone
Domestic Rice
Imported Rice
Imported rice price lower particularly in cordon areas
Consistent with results data on rice prices from household surveySlide15
More high price outliers in 2014
Source:
Glennerster and
Suri
, 2015 www.theigc/country/sierra-leone
Slide16
Household cell phone survey,
Jan-Feb ‘15Source:
Fu,
Glennerster
, Himelein, Rosas, and Suri
, 2015
LFS sample
= 4199
66% had cell phones
68% of these reachedSlide17
Decline in employment in urban areas
Source:
Fu,
Glennerster
, Himelein, Rosas, and Suri, 2015
Slide18
Nonfarm HH
enterprises hit badly Source:
Fu,
Glennerster
, Himelein, Rosas, and Suri
, 2015
P
ercent of HH with non-farm business no longer operating rose from 4% to 12%
1/3
cite
Ebola
as
reason their business no longer operates
Average business revenues shrunk by 40%
>90% urban women worked in non-farm HH enterprises pre-ESlide19
Little evidence of impact on ag
Ebola coincided with the growing and harvest season93 percent of farming households grow riceIn Nov more
than half
had
rice in field, mainly because of rain (72%), not EbolaMore than half farming HH hired outside laborSome cite labor constraints for harvest14% labor constraints in HH vs. 6% labor constraint in communityno significant differences across quarantine areasMost farmers never sell rice, prod estimates unreliableNo clear signs of probs in cocoa but sample smallSlide20
Economic impacts uncorrelated with disease burden
No difference in range of different economic outcomes between cordon vs noncordon areasOnly difference is imported rice prices are lower in cordon areas, possibly because of food assistance
Main differences between urban and nonurban
In Liberia similar declines in employment and revenue in badly hit and less badly hit (
xxxxx, Werker, in progress)As in Sierra Leone biggest impacts in capital (Monrovia)Some sectors hit worse than others (e.g. construction)20Slide21
Aid: 9
% HH received social assistanceSlide22
Aid correlated with Ebola and transport disruption Slide23
F
ood prices lower where food aidMatched market price data with HH assistance data
Difference-in-difference specification
Two way causality
possible:Receiving food aid reduces pricesHigh prices in an area trigger food aidWithout detailed data on timing cant separate the two
Moving from 0-100% assistance reduces rice prices by 10%
Suggests food aid on average reduced food prices by 1%
Net consumers gain from lower food prices, net producers lose
Most poor, even farmers are net consumers riceSlide24
Conclusion: role of researchers
Researchers can play and important role in providing good data in a crisis, to reduce risk of policy effort being diverted to wrong areas
But needs much faster turn around time than normal studies
Much easier if building on existing research, experience, and in country team
Good causal identification can be hard as impacts may not be correlated with direct cause of crisisBut good descriptive evidence is very valuable, especially if good pre crisis data to compare withDifferent methods can be useful: e.g. SMS, cellphone, modeling
The press may only remember the high case scenario
24Slide25
Conclusion: attention vs confidence?
Early stages of Ebola crisis showed how difficult it can be to attract the attention and support of the international community to crisisDisease spread from end April-end June with little attention
World grown weary of calls for help, only threat of total collapse gets through the noise
But risks in overplaying likely impacts
Can further reduce domestic confidence with important economic impactsIf final results not as big as predicted, risks people concluding it “was not so bad after all”Raises the stakes for the next crisis
25Slide26