Presentation_Malaria Eradication in the Americas a Retrospective Analysis of Childhood Exposure_Hoyt Bleakley

7/25/2019 Presentation_Malaria Eradication in the Americas a Retrospective Analysis of Childhood Exposure_Hoyt Bleakley

http://slidepdf.com/reader/full/presentationmalaria-eradication-in-the-americas-a-retrospective-analysis-of 1/42

Malaria Eradication in theAmericas: A RetrospectiveAnalysis of Childhood Exp

Hoyt Bleakley

American Economic Journal: Applied Economics 2 (April 2010): 1 – 45



1 Research question and co

2 Methodology

3 Results

4 Conclusion




2 Methodology

3 Results

4 Conclusion




“Does malaria hold back econprogress?”

Is malaria one of the cause ofunderdevelopment?




“Identify the effect that childhoexposure to malaria eradicatio

on subsequent labor productivan adult.”




Health context:

Anemia

Lower returns to schooling + shorter tim

educational investmentsWarmer and wetter climate




Historical context:

Effort against malaria in the Southern U20s

Worlwide malaria eradication in the 50




Source: Author’s paper




2 Methodology

3 Results

4 Conclusion



2 Methodology

Data (sources):

Census sample (Integrated Public Use

Sample)



2 Methodology

Data (sources):

US: Native-born white males, between55, census year 1880-2000, cohorts withof birth ranging from 1825 to 1965.

Brazil and Mexico: native-born males ato 55, censuses from 1960 to 2000, birthfrom 1905 to 1975.

Colombia: native males, censuses of 11993, birth cohorts from 1918 to 1968.



2 Methodology

Data (construction):

Area of birth rather than current reside

Intention-to-treat design.



2 Methodology

Data (construction):Cohort-level data: average by area o

year of birth and census year

3D panel



2 Methodology

Data (use):

Outcomes of interest: Literacy, EducatIncome, Occupational/Sectorial Indic

Income.Malaria variables (malaria mortality as

fraction of total in 1980, different proxie

Controls.



2 Methodology

Research design (problem):

Reverse causality: Badinstitutions→Underdevelopment→↑M

decrease productivity→bad institutioTargeted interventions in public healt

Exogenous because innovations comoutside the study area.



2 Methodology

Research design (identification

Differences in differences



2 Methodology


Cross-area differences



2 Methodology




2 Methodology


Time differences



2 Methodology




2 Methodology

Research design (identification):Time differences:« Born after » vs « already adults » (exposure={0

Y: some socioeconomic outcome for state or municipio j

Mj,pre: pre-program malaria incidence

β: parameter of interest=to which extent a high-incidence of malaprogram (so an intensive program) would lead to an increase in osignification depends on the country, but roughly speaking, it meapercentage change in average income between an area wherecause death – and thus where there is no eradication campaing-where mortality is only due to malaria – and thus intensive eradica



2 Methodology


Time differences:

« Born after » vs « already adults » (exposure

Panel A: Different controls.

Panel B: 2SLS, because of possible measurement-erroproxies of malaria.

Panel C: Comparison between movers and non-mov

Panel D: By region.



2 Methodology


Time differences:

Functional form of childhood exposure (co

Yjkt: average outcome for a cohort of state of birth j, census yof birth k.

Expk : potential exposure to the malaria eradication-campaigmax(min(21,k-(y-21)),0))/21, with y starting year of the eradicacampaign.



2 Methodology


Time differences:

Graphical representation with flexible equa



2 Methodology


To sum it up: Compare the cross-cohoevolution of outcomes across areas widistinct infection rates to assess thecontribution of the eradication campathe observed changes.




2 Methodology

3 Results

4 Conclusion



Pre-post comparisons – United S







Effect on income of 10-15% when comparing malarious states to the highly malarious states



Pre-post comparisons – Brazil an



Around 0.25 for OLS and 0.45 for 2SLS for earne

Total income for Brazil : around 0.37 for OLS an

2SLS

Pre-post comparisons – Brazil an



Pre-post comparisons – Colomb



Pre-post comparisons – Colomb

Increase of 5% point in literacy

Increase of 0.1-0.6 in years of schooling

Increase of 0.08-0.16 in log income



Pre-post comparisons

Effect on income of (around) :

10-15% for the United States

30% in Brazil

20% in Mexico 22% in Colombia (when adjusted, otherwise 0

tables)



Panel estimates : relationship beincome and pre-campaign mal




Reduced form differences ; 95/5 percentile co:

1) effect on adult income of a given malaria b

2) magnitude of decline due to the eradicatio

campaign







2 Methodology

3 Results

4 Conclusion



Discussion of the results

Other mechanisms that could have an impact on income

Schooling

Other vector-borne diseases

Mortality selection

Spillovers

Not a real threat to the results !



Discussion of the policy implicat

Impact on income inequalities

Potentially very large benefits in region wherestill endemic today



Other related papers

Cutler, David, Winnie Fung, Michael Kremer, Monica SinTom Vogl. 2009. “Mosquitoes : the long-term effects of eradication in India”

malaria exposure reduces household expenditures in adulthosmaller magnitude than what would be expected

Lucas, Adrienne M. 2009. “Malaria eradication and edattainment : evidence from Paraguay and Sri Lanka”

more years of schooling for women, while mixed results obtaarticle



Further research

Would be interesting to find a way to include women in

Same study in Sub-Saharian Africa

Among the (new) different ways to fight malaria, which

be the most efficient ?

Presentation_Malaria Eradication in the Americas a Retrospective Analysis of Childhood Exposure_Hoyt Bleakley

Documents