Arsenic Short by R8VDq9


									             Richard T. Carson
     University of California, San Diego

            Phoebe Koundouri
Athens University of Economics and Business

               Céline Nauges
French Institute for Research in Agriculture
     & Toulouse School of Economics
 Arsenic in Bangladesh Groundwater Wells
 Rural Bangladesh areas originally relied on surface water
    Walked long distances for water
    Substantial bacterial and other contamination
 Millions of shallow tube wells built starting in 1970’s
    Fraction served by these wells accelerated over time
    Now close to complete coverage (95%+)
 Widespread arsenic contamination discovered in large scale
  survey of wells done by British Geological Survey (2001)
   Our sample average arsenic concentration 62 μg/liter
        Our sample range [0.3 to 421]
        WHO standard 10 μg liter; Bangladesh standard 50 μg liter
   57 million people exposed to WHO standard or greater
     Chronic Health Impacts of Arsenic
 Short/Medium Term
    Lethargy
    Headaches/confusion
    Skin problems: hypopigmentation & keratoses
 Long term
    Skin cancer
    Internal cancers: bladder, liver, lungs
 Arsenic builds up in body and is very slow to clear
 Time from exposure to initial skin problems: 10+ years
 Time from exposure to more serious problems: 20+ years
                   Existing Work
 Large amount of epidemiology focused on:
    Body burden
    Skin problems
    Various types of cancer
 Public health/economics work focused on:
    Information/incentive for identifying/switching water
 Public health/economic work focused on:
    Medical cost of illness/WTP to avoid adverse effects
     Specific Issues This Paper Examines
 Concerning Impact of Arsenic on Labor Supply

 Relationship between arsenic exposure and household
  labor hours supplied in rural Bangladesh
 Substitution patterns involving household labor
  supply associated with arsenic exposure
 Interaction between labor hours supplied, arsenic and
  three main types of assets: land, physical capital, and
  human capital
 Key to econometric identification strategy
    Arsenic contamination function of geological conditions
    Use data from time period before widespread knowledge of
     arsenic contamination in specific well
    Spatially merge data from the British Geological Survey
     (BGS, 2001) of groundwater wells (done in mid 1998-late
     1999) with Bangladesh government’s Household Income
     and Expenditure Survey done in 2000.
    Merge takes place at thana level—5th order subdistrict small
     enough that BGS levels highly correlated with actual
     exposure but large enough that households ~independent
        220 thanas each with 20 sampled households
              Household Labor Supply
 Labor hours recorded for any type of remunerated work
    Each household member
    Paid in money or in-kind/household farm or firm
       Hours “worked” at home not recorded
 Approach taken
    Add together labor hours supplied by each house member
    Use household demographic characteristics as regressors
       Number of member in each sex/age category
       Other household demographics variables
       Choice of Modeling Framework
 Nature of Dependent Variable
   Non-negative by definition with upper-end of hours worked for
    households with large number of members quite large but finite
 Suggests survival model with hours worked as “time” variable
 Spike at/near zero and “ties” for popular hours worked
  suggest semi-parametric Cox Proportional Hazard Model
    Hazard function h(t) = f(t)/S(t), where S(t)= 1 – F(t)
    Covariates assumed to proportionately shift h(t)
    h(HHWi | Xi, ASi) = h0(HHW)exp(αASi + βXi), where
    HHWi is household hours work, Xi demographic
     composition of household, ASi is arsenic level
    Β larger than 1 indicates downward shift in HHW
         Cox Proportional Hazard Model
 Allows arbitrary (non-parametric) baseline hazard
    Hazard function h(t) = f(t)/S(t), where S(t)= 1 – F(t)
 Covariates assumed to proportionately shift h(t)
 Basic model:
    h(HHWi | Xi, ASi) = h0(HHW)exp(αASi + βXi), where
    HHWi is household hours work, Xi demographic
     composition of household, ASi is arsenic level
    Coefficient of 1 for [α, β] indicates no shift in baseline h(t)
        Smaller than 1 indicates upward shift in HHW (more hours)
        Larger than 1 indicates downward shift in HHW (less hours)
        If indicator variable, then coefficient - 1 is percentage shift
           Base Model (Arsenic Excluded)
           Pattern of Demographic Results
 Increase in HHW for females 6-25, particularly
  pronounced for 16-25 age group
 No significant deviation from 1 for older females
   Significant reversal from younger females though
 Increase in HHW for males of all ages starting with [6-10]
    HHW roughly constant from 16-55, enormously significant
 Islamic households work substantially fewer hours
    Conditional on this effect, older females work more
 Quadratic with acres, linear HHW down, quadratic up
 HHW goes up with assets, down with Max house educ.
                     Adding Arsenic
 Linear term only                 1.0108 (t=6.45)
 Quadratic specification
   Linear                         1.0226 (t=4.59)
   Quadratic                      0.9996 (t=-2.50)

 HHW is decreasing in arsenic but at a slowly decreasing rate
 Turning point is at ~300 μg (3% of data beyond that point)
  and at ~580 μg in specification with interactions
    Adding Interactions with Arsenic
 Smaller households work proportionately less

 Females, particularly older associated with lower HHW

 Prime age males associated with higher HHW

 Older head, more assets, HHW goes up

 More acres, HHW goes down
         Predicted Effect on HHW
 Reducing arsenic level to zero
    Increase HHW by 7.9%
 Reducing arsenic to WHO standard (10 μg)
    Increase HHW by 6.5%
 Reduce arsenic to Bangladesh standard (50 μg)
    Increase HHW by 3.6%
 Impact on median household substantially smaller
 because arsenic exposure highly skewed
 For poor in many places, labor hours main asset
 Bangladesh in 2000 ideal for examining the impact of
  large scale low level chronic health problems induced by
  exogenous and unknown arsenic exposure
 Estimated effect large, 7.9% reduction HHW
 Compensation mechanisms
   Given overall arsenic reduction, more work by prime age
    males, less by females
   Physical assets decrease arsenic related loss in HHW
   Land assets increase arsenic related loss in HHW

To top