Variables Included

xSub datasets include variables from two categories:

  1. conflict and violence
  2. local demographics, geography, ethnicity, weather

ID variables

The following variables identify spatial and temporal units, and can be used for merging and aggregation:

  • ID_0: GADM country ID (numeric)
  • ISO: ISO3 country ID (string)
  • NAME_0: country name (string)
  • ID_1: GADM province ID (numeric)
  • NAME_1: province name (string)
  • ID_2: GADM district ID (numeric)
  • NAME_2: district name (string)
  • CLEA_CST: electoral constituency ID / format is “A.B.C”, where A is the CLEA country code, B is the election year, C is the CLEA constituency number (string)
  • CLEA_CST_N: name of electoral constituency (string)
  • PRIO_GID: PRIO-GRID cell ID (numeric)
  • PRIO_XCOORD: longitude of cell centroid (numeric)
  • PRIO_YCOORD: latitude of cell centroid (numeric)
  • PRIO_COL: PRIO-GRID column number (numeric)
  • PRIO_ROW: PRIO-GRID row number (numeric)
  • YEAR: year (numeric)
  • YRMO: year-month (YYYYMM) (numeric)
  • WID: week ID (numeric)
  • DATE: date (YYYYMMDD) (numeric)
  • GEOPRECISION: geographic precision code (only included in event-level data):
    • settlement: event geo-coded to nearest settlement (e.g. village, town, city)
    • adm2: event geo-coded to center of nearest second-order administrative unit (e.g. district, county, rayon)
    • adm1: event geo-coded to center of nearest first-order administrative unit (e.g. province, state, oblast)
    • adm0: event geo-coded to country level
  • TIMEPRECISION: temporal precision code (only included in event-level data):
    • day
    • week
    • month
    • year

Event counts: conflict and political violence

ACTORS

  • SIDE_A: incumbent government, pro-govt militia, third party acting on incumbent’s behalf
  • SIDE_B: rebels, anti-government militia, third party acting on rebels’ behalf, and other armed groups directly challenging the government
  • SIDE_C: civilians
  • SIDE_D: local militia, tribe, other non-state actors not directly challenging the government

The actor dictionary specifies which factions belong to SIDE_A SIDE_D for each conflict.

  • INITIATOR variables:
    • INITIATOR_SIDEA, INITIATOR_SIDEB, INITIATOR_SIDEC, INITIATOR_SIDED
  • TARGET variables:
    • TARGET_SIDEA, TARGET_SIDEB, TARGET_SIDEC, TARGET_SIDED
  • ACTION variables:
    • ACTION_ANY, ACTION_IND, ACTION_DIR, ACTION_PRT
  • INITIATOR_ACTION variables:
    • SIDEA_ANY, SIDEA_IND, SIDEA_DIR, SIDEA_PRT
      SIDEB_ANY, SIDEB_IND, SIDEB_DIR, SIDEB_PRT
      SIDEC_ANY, SIDEC_IND, SIDEC_DIR, SIDEC_PRT

    NOTE: ACLED and UCDP-GED do not explicitly distinguish between the initiators and targets of conflict events. In those cases (undirected dyads), INITIATOR_* and TARGET_* are to be interpreted as “participant 1” and “participant 2,” respectively. and the INITIATOR_ACTION variables are to be interpreted as “tactics used in event involving participant 1.” In all other cases (directed dyads), the “initiator” and “target” labels apply as written.

DYAD

  • DYAD__: interaction between actors and

    NOTE: ACLED and UCDP-GED do not explicitly distinguish between the initiators and targets of conflict events. In those cases (undirected dyads), the variables DYAD_* are to be interpreted as “event involving actors and .” In all other cases (directed dyads), they are to be interpreted as “action by actor against actor .”

ACTION

General categories:

  • ANY: any use of force
  • IND: indirect violence (e.g. indirect fire, shelling, air strikes, chemical weapons)
  • DIR: direct violence (e.g. direct fire, arrest, assassination)
  • PRT: protest

Specific categories (only included in event-level data):

  • ACTION_AIRSTRIKE
  • ACTION_AMBUSH
  • ACTION_ARMOR
  • ACTION_ARREST
  • ACTION_ARTILLERY
  • ACTION_BOMB
  • ACTION_CHEMICAL
  • ACTION_CIV_ABUSE
  • ACTION_DISPLACE
  • ACTION_FIREFIGHT
  • ACTION_KIDNAP
  • ACTION_KILLING
  • ACTION_KILLING_A
  • ACTION_OCCUPY
  • ACTION_PROPERTY
  • ACTION_PROTEST
  • ACTION_PROTEST_V
  • ACTION_RAID
  • ACTION_RIOT
  • ACTION_RIOTCONTROL
  • ACTION_ROBBERY
  • ACTION_ROCKET
  • ACTION_SIEGE
  • ACTION_STORM
  • ACTION_SUICIDE
  • ACTION_TERROR
  • ACTION_UNKNOWN

Local demographics, geography, ethnicity, weather

  • POP_1990: population per square kilometer, 1990 (GPW v3) (numeric)
  • POP_1995: population per square kilometer, 1995 (GPW v3) (numeric)
  • POP_2000: population per square kilometer, 2000 (GPW v3) (numeric)
  • ELEV_MEAN: average elevation, meters (ETOPO05) (numeric)
  • ELEV_SD: standard deviation of elevation, meters (ETOPO05) (numeric)
  • ELEV_MAX: maximum elevation, meters (ETOPO05) (numeric)
  • OPEN_TERRAIN: proportion of land covered by open terrain (shrublands, savannah, grasslands, barren or sparsely vegetated) (GLCC) (numeric)
  • FOREST: proportion of land covered by evergreen, deciduous or mixed forest (GLCC) (numeric)
  • WETLAND: proportion of land covered by wetlands or water bodies (GLCC) (numeric)
  • FARMLAND: proportion of land covered by croplands (GLCC) (numeric)
  • GREG_NGROUPS: number of local ethnic groups (GREG) (numeric)
  • GREG_GROUPS: names of local ethnic groups (GREG) (string)
  • WLMS_NLANG: number of local languages (WLMS) (numeric)
  • WLMS_LANGS: names of local languages (WLMS) (string)
  • NBUILTUP: number of built-up areas (GGIS) (numeric)
  • BUILTUP: names of built-up areas (GGIS) (string)
  • NPETRO: number of petroleum fields (PRIO) (numeric)
  • PETRO: names of petroleum fields (PRIO) (string)
  • DIST2PROVCAP: distance to nearest provincial capital, km (GGIS) (numeric)
  • DIST2CAP: distance to national capital, km (GGIS) (numeric)
  • ROAD_XING: number of primary, secondary roads crossing border (DCW) (numeric)
  • ROAD_LENGTH: length of primary, secondary roads, km (DCW) (numeric)
  • AREA_KM2: area of spatial unit, km2 (numeric)
  • ROAD_DENSITY: road density, km/km2 (DCW) (numeric)
  • TEMP: average temperature in Celsius (NOAA) (numeric)
  • RAIN: average precipitation in millimeters (NOAA) (numeric)