Prerequisites

• Download the raw data from https://microdata.nis.gov.kh/index.php/catalog/38 and https://microdata.nis.gov.kh/index.php/catalog/44
• Extract the files to the "raw_data" folder

Table of contents

Globals

• This section sets global variables for use later in the script, including:
  • Exchange rate and currency inflation conversion to USD
  • Winsorization Thresholds
    • We correct outliers by Winsorization - setting values above the 99th and/or below the 1st percentiles to the value of the 99th or 1st percentile. The thresholds can be adjusted here or removed by setting them to 0 and 100
  • Rescaling Survey Weights
    • Defines new population-based survey weight based on population change from date of data collection

Household IDs

• Description: This section produces a dataset with Holding_ID (hhid) as a unique identifier, along with a single location identifier (province)
• Output: hhids.dta
• Coding Status: Complete
• Known Issues: None

Weights

• Description: This section produces a dataset with Holding_ID (hhid) as a unique identifier and a weight variable defined in the source dataset
• Output: weights.dta
• Coding Status: Complete

Individual IDs

• Description: This section produces a dataset with Holding_ID (hhid) and individual roster IDs as a unique identifier and contains variables that indicate an individual's gender, age, and status as female head of household
• Output: person_ids.dta

Household Size

• Description: This section produces a dataset with Holding_ID (hhid) as a unique identifier and variables containing number of household members and whether the household head is female
• Output: hhsize.dta

COVID Impact

• Description: This section produces a dataset with Holding_ID (hhid) as a unique identifier and a variable describing whether or not the household experienced a shock from the COVID-19 pandemic.
• Output: COVID_impact.dta
• Coding Status: Complete

Association Membership

• Description: This section produces a dataset with Holding_ID (hhid) as a unique identifier and variables describing the household's involvement with formal and informal agricultural associations or communities.
• Output: association_membership.dta
• Coding Status: Complete

Agricultural Households

• Description: This section produces a dataset with Holding_ID (hhid) as a unique identifier and variables describing the household's land use decisions on the homelot and parcel(s), as applicable.
• Output: agricultural_activities.dta

Household Decision Makers

• Description: This section produces a dataset with Holding_ID (hhid) as a unique identifier and variables that describe the education and gender of person or people who make household management decisions - 1: male, 2: female, or 3: mixed male and female
• Output: hh_decision_makers.dta

Livestock

• Description: This section produces multiple datasets. The primary dataset produced by this section has Holding_ID (hhid) as a unique identifier and 46 variables for each form of livestock [cattle, buffalo, pigs] describing the household's agricultural activities related to raising that livestock.
• Main Output: livestock.dta
• Known Issues:
  • The given herd growth/loss factors [S4_Q10 in 2020 and s6_q14 in 2021] may not be comprehensive. This creates an issue when constructing certain intermediate variables, e.g. num_livestock_start, that form the denominator of constructed ratios or rate indicators, e.g. livestock_growth_pct.
  • Significant outliers in the given livestock sale price [riel/head, S4_Q10e s6_q14g] were handled by calculating median absolute deviation statistics and replacing observations with provincial or national medians.
  • Missing values for livestock sale price are replaced by constructed provincial or national medians.
  • The survey does not provide a market or implicit unit price for slaughtered livestock meat. We assume equivalence to the price of live livestock sales.

Poultry

• Description: This section produces multiple datasets. The primary dataset produced by this section has Holding_ID (hhid) as a unique identifier and 46 variables for each form of poultry [chicken, duck, goose (2020 only)] describing the household's agricultural activities related to raising that poultry.
• Main Output: poultry.dta
• Known Issues:
  • The given flock growth/loss factors [S4_Q15 in 2020 and s6_q54 in 2021] may not be comprehensive. This creates an issue when constructing certain intermediate variables, e.g. num_poultry_start, that form the denominator of constructed ratios or rate indicators, e.g. poultry_growth_pct.
  • Missing values for egg_price are replaced by constructed provincial or national medians.
  • Significant outliers in the given poultry sale price [riel/kg, POULTRYSALE poultrysale] were handled by calculating median absolute deviation statistics and replacing observations with provincial or national medians.
  • Missing values for poultry sale price are replaced by constructed provincial or national medians.

Consolidating Poultry and Livestock, Adding Household-Specific Variables

• Description: This section consolidates intermediate files and adds variables related to holding size (area), rice production, cashew production, animal disease, irrigation, adverse weather events, extension training, and household family labor.
• Main Output: household_variables.dta