Survey Data

Data Sources

  • Demographic and Health Survey (DHS)
  • Household expenditure and Income survery
  • Living standards measurement study (LSMS)
  • Surveys conducted by National Statistics Office (NSO) of various countries

Choosing proxy variables

  • Verifiability of the variable
  • Correlation betweent the variable and household consumption levels
  • We should not depend on any single or very few variables
  • The variables may differ based on the region. For example for a rural area, livestock is important but not for urban area

Proxy Variables

  • Human capital variables
    • Education of household head
    • Highest level of education in household
    • Female literacy
    • Number of childern in school
  • Demographic characteristics
    • Household size
    • Number of children
    • Gender / Marital status
    • Age of Household head
    • Dependency ratio
  • Household assets
    • Own home
    • Type of wall construction
    • Type of roofing material
    • Type of latrine
    • Number of rooms per capita
    • Type of cooking fuel
    • Radio, television and other forms of electronic or communication devices
    • Bicycle, car, motorcycle or other means of owned transport
    • Furniture
    • Access to electricity
    • Cooker, heater, fan, air-conditioning
  • Productive assets
    • Landholding size
    • Livestock
    • use of fertiliser
  • Livelihood options
    • Agricultural or nonfarm wage labour
    • Non-farm independent business
    • Agricultural production of cash or staple crops
    • Receipt of foreign remittances
    • Sector of work
  • Community variables
    • Presence of midwife
    • Population density
    • Asphalt road
    • Bank in Community
    • Divisional Secretariat in community

Proxies used in a PMT model

Reference:-