README for the files distributed with the Urban Social Disorder dataset =========================================================================== Version: 2.0 (7 Jun 2017) Peace Research Institute Oslo, Norway Citation: ------------ When using the USD data please cite: Urdal, Henrik & Kristian Hoelscher, 2012. ‘Explaining urban social disorder and violence: An empirical study of event data from Asian and Sub-Saharan African cities’, International Interactions 38(4): 512–528. More detailed information is included in "Codebook.pdf" and should be cited as: Bahgat, Karim, Halvard Buhaug, and Henrik Urdal. 2017. “Urban Social Disorder Codebook, version 2.0”, Peace Research Institute Oslo. https://www.prio.org/Data/Armed-Conflict/Urban-Social-Disorder/ Fileformats: ------------ All files in this data package come in two formats: - .xls (Excel) - .dta (Stata) Overview: --------- The USD data distribution consists of two main tables, and a number of auxiliary files aggregated to different units of analysis for convenience. - "events": -------- Main table of urban social disorder events used for analysis. See "Codebook.pdf". - "reports": -------- Supportive table containing the reports that the event codings are based on. Each observation is a subsection of a Keesings report pertaining to a particular event. For users interested in quality assurance, textual analysis, or data mining. The reports listed in this table are those that were deemed to contain an event during the initial screening. A report continues to be listed here even if the event was later dropped from the events table. This means that not all reports have an event that uses it. - "citymonth": -------- Pre-aggregated to the city-month level, with observations for every month and year regardless if no events happened. About 50 events had unknown month precision and were dropped before aggregating, instead of setting them to an arbitrary month and biasing the results. - "cityyear": -------- Pre-aggregated to the city-year level, with observations for every year regardless if no events happened. - "cities": -------- An overview of the cities contained in the dataset, including associated variables and aggregated values for the entire period. This table also includes useful link IDs to other city-level datasets. Variables: ---------- The following lists and explains the variables for each of the files: - "events": -------- See "Codebook.docx". - "reports": -------- * REPORTID Report ID (text; ID) Unique identifier of the Keesing’s report, corresponding to REPORTID1-3 of the events table. IDs consists of a city-ID, followed by a dash and an alphanumeric code. In the alphanumeric code, the number refers to the report, while the character refers to logically grouped subsections of the report referring to a particular event. * TITLE Report Title/Header (text) The (unstructured) title header of the Keesings report to provide some context on the date and topic of the report. This will be the same for all REPORTIDs with the same report number. * CITY_ID ID of City (text; ID) Unique identifier for the city. * CITY Name of City (text) Name of the city for which the report was gathered. * COUNTRY Name of Country (text) Name of the modern-day country in which the city is located. This is based on the situation at the time that the data is released and does not change to reflect changing ownership over time. * ISO3 ISO3 Alphanumeric Country Code (3 character text; ID) Alpha-3 country code according to the ISO 3166-1 standard. This is based on the situation at the time that the data is released and does not change to reflect changing ownership over time. * GWNO Gleditsch-Ward Country Code (1-3 digit numeric; ID) Numeric country code according to the Gleditsch and Ward list of independent states. This is based on the situation at the time that the data is released and does not change to reflect changing ownership over time. * FULLTEXT The full text of relevant parts of the report (text) Full text of all relevant parts of the report needed to understand the background, context, and setting of the event. This fuller version of the text is NOT provided in the events table. * SUMMARY Summary of the event itself (text) Summary of only the parts of the text describing the event itself. Is either the same or a subset of the FULLTEXT text. The SUMMARY variable in the events table combines this variable for all the relevant REPORTIDs. - "citymonth": -------- Same as the events table, except for the following changes: * NEVENTS Total number of events in the city that started during the month. * DEATHEVENTS Number of events with at least one death in the city that started during the month. * NODEATHEVENTS Number of events with no known deaths in the city that started during the month. This includes events where it is unknown whether any deaths occured (-99). * COUNTRY_HIST GW country code of the city was determined on the 1st of each month. * GWNO_HIST GW country code of the city was determined on the 1st of each month. - "cityyear": -------- Same as the citymonth table, except for the following changes: * NEVENTS Total number of events in the city that started during the year. * DEATHEVENTS Number of events with at least one death in the city that started during the year. * NODEATHEVENTS Number of events with no known deaths in the city that started during the year. This includes events where it is unknown whether any deaths occured (-99). * COUNTRY_HIST Country name of the city was determined on the 1st of january of each year. * GWNO_HIST GW country code of the city was determined on the 1st of january of each year. - "cities": -------- Same as the cityyear table, except aggregated for the whole period. This table also includes the following city-level link IDs: * ID_UNWUP An ID corresponding to the "City Code" in the UN's World Urbanization Prospects datasets. Can be used as a unique key to join with city-specific population data (https://esa.un.org/unpd/wup/CD-ROM/). * ID_UNDAT An ID corresponding to the "Code of City" in the city-specific indicators from the UN Data website. Can be used as a unique key to join with various city-specific indicators provided by the UNSD Demographic Statistics, including types of housing units, roofing, lightning, electricity, sanitation, water, internet, etc. When downloading city data from the UN Data website make sure to click "Select Columns" and then check off "Code of City" in order to include this ID variable with your dataset (http://data.un.org/Search.aspx?q=city).