Interstate War Initiation and Termination (I-WIT) data set

Résumé 0

Project Summary:The Interstate War Initiation and Termination (I-WIT) data set was created to enable study of macro-historical change in war initiation and termination. I-WIT is based on the Correlates of War (COW) version 4 list of interstate wars, and contains most of the interstate wars in the COW list; those excluded were wars the researchers believe do not meet the COW criteria for interstate wars. For each war, research assistants (RAs) coded a host of variables relating to war initiation and termination, including whether each side issued a declaration of war, the political and military outcomes of the war (which are coded separately), and the nature of any agreement that concluded the war. One argument made in several publications based on these data (also part of a larger book project) is that the proliferation of codified international humanitarian law has created disincentives for states to admit that they are in a state of war. Declaring war or concluding a peace treaty would constitute an admission of being in a state of war. As international humanitarian law has proliferated and changed in character over the past 100 years or so, it has set the costs of compliance – and also the costs of finding a state to be out of compliance – very high. Thus, states avoid declaring war and concluding peace treaties to try to perpetrate a type of legal fiction – that they are not at war – to limit their liability for any violations of the laws of war. Data Abstract: The data cover the period from 1816 to 2007 and span the entire world. Dozens of graduate and undergraduate RAs working between 2004 and 2010 compiled existing data from secondary sources and, when available online, primary sources to code variables listed and described in the coding instrument. RAs were given a coding instrument with a description and rules for coding each variable. Typically, they consulted both secondary and primary sources, although off-site archival sources were not consulted. They filled in a spreadsheet for each war with variable values, and produced a narrative report (henceforward, “narrative”) of 5-10 pages that gave background information on the war and also justified their coding. Each war was assigned to at least two RAs to check for inter-coder reliability. If there was disagreement between the first two RAs, a third RA was brought in to code discrepant variables for that war. Where possible, a 2/3 rule was followed in resolving discrepancies. Remaining discrepancies are addressed in the “discrepancy narrative,” which lists the discrepancies and documents final coding decisions. Files Description: Some sources were scanned (e.g., declarations of war or peace treaties) but for the most part, RAs took notes on their assigned cases and produced their coding and narratives based on these notes. The coding instrument and the discrepancy narrative are included in the data documentation files, and all data files produced – including original codings that were discrepant with later codings – are included in the interest of allowing other researchers to make their own judgments as to the final coding decisions. A companion data set – C-WIT (Civil War Initiation and Termination) – is still under construction and thus not shared at this time.