EPAR Technical Report #335
Tue, 11/21/2017
Authors: 
C. Leigh Anderson
Travis Reynolds
Pierre Biscaye
Didier Alia
David Coomes
Jack Knauer
Josh Merfeld
Ayala Wineman
Abstract: 
EPAR has developed Stata do.files for the construction of a set of agricultural development indicators using data from the Living Standards Measurement Study - Integrated Surveys on Agriculture (LSMS-ISA). We are sharing our code and documenting our construction decisions both to facilitate analyses of these rich datasets and to make estimates of relevant indicators available to a broader audience of potential users. 
 

Code is currently available on our GitHub repository for three survey instruments:

  • Ethiopia Socioeconomic Survey (ESS) Wave 3, 2015/16 (LSMS-ISA)
  • Nigeria General Household Survey-Panel (GHSP) Wave 3, 2015/16 (LSMS-ISA)
  • Tanzania National Panel Survey (TNPS) Wave 4, 2014/15 (LSMS-ISA)

Code for additional waves of survey data from each of these three countries will be added to this repository as it is available.

The repository includes a separate folder for each country. Each of these folders includes master Stata .do files with all of the code used to generate the final set of indicators from the raw survey data for a given survey wave. The raw survey data files are available for download free of charge from the World Bank LSMS-ISA website. The .do files process the data and create three final data sets at the household, individual, and plot levels with labelled variables, which can be used to estimate sumary statistics for the indicators and for a variety of intermediate variables.

We have compiled a set of summary statistics for the final indicators restricted to rural households only in an excel spreadsheet, which includes estimates for the Ethiopia ESS Wave 3 (2015-16), Nigeria GHSP Wave 3 (2015-16), and Tanzania NPS Wave 4 (2014-15), as well as for two surveys that are not yet publicly available: the Ethiopia Agricultural Commercialization Cluster (ACC) Survey (2016) and the India Rice Monitoring Survey (RMS) (2016). The first three instruments include nationally-representative samples. The Ethiopia ACC includes representative samples for the four main regions of Ethiopia, and the India RMS includes representative samples for four states: Bihar, Odisha, Uttar Pradesh, and West Bengal. Estimates for these four states are reported separately.

The spreadsheet includes estimates of the following indicators created in our code files:

  • Income:
    • Total income (annual, in $ and PPP$)
    • Per capita income (PPP$/person/year)
    • Total farm income ($, annual)
    • Total crop income ($, annual)
    • Total livestock income ($, annual)
    • Total non-farm income ($, annual)
    • Share of non-farm income in total income (proportion)
    • Average daily wage rate in agriculture ($/day)
  • Agricultural production
    • Average household farm size (ha)
    • Crop yield (kg/ha, main growing season) on all plots and on purestand plots separately
    • Crop production cost per hectare ($/ha, annual), explicit costs only and including both implicit and explicit costs
    • Inorganic fertilizer application rate (kg/ha)
    • Share of crop production value sold (proportion)
    • Crop labor productivity ($/labor day, annual)
    • Crop land productivity ($/ha, annual)
    • Gender crop land productivity gap (%)
    • Milk productivity (liters/milk-producing animal, annual)
    • Egg productivity (eggs/egg-producing animal, annual)
  • Gender
    • Prevalence of women of reproductive age consuming a diet of minimum diversity (MDDW) (proportion)
    • Average number of food groups consumed by women of reproductive age in last day (MDDW Score, 0-10)
    • % of women who make decisions about use of household income (WEIA) (proportion)
    • % of women who are sole or joint owners of productive assets (including land and livestock) (WEIA) (proportion)
  • Adoption
    • Households adopting improved seed varieties (proportion)
    • Households using inorganic fertilizer (proportion)
    • Households using livestock vaccines (proportion)
    • Households reached by extension (public or private) (proportion)
    • Households using financial services (proportion)

Estimates are disaggregated by gender as appopriate depending on the indicator (by gender of the individual, the plot manager, or head of household).

In addition to the documentation of indicator construction decisions in the Stata .do files, we also have prepared a document outlining the general construction decisions for each indicator across survey instruments, focusing on the five instruments for which we produced estimates in our spreadsheet. We attempted to follow the same construction approach across instruments, and note any situations where differences in the instruments made this impossible. A final document outlines general principles and considerations for contructing these indicators.  

Type of Research: 
Data Analysis
Research Topic Category: 
Sustainable Agriculture & Rural Livelihoods
Agricultural Productivity, Yield, & Constraints
Agricultural Inputs & Farm Management
Market & Value Chain Analysis
Household Well-Being & Equity
Poverty
Food Security & Nutrition
Gender
Technology
Technology Adoption
Population(s): 
Rural Populations
Smallholder Farmers
Women
Geographic focus: 
West Africa Region and Selected Countries
South Asia Region and Selected Countries
East Africa Region and Selected Countries
Dataset(s): 
LSMS & LSMS-ISA