Abstract
The empirical analysis of datasets covering a large number of countries and time periods has become an integral part of conflict and peace economics. As such, numerous studies examine relationships between and among macroeconomic, political, and conflict variables and this often involves the merging of disparate datasets to combine relevant variables for which the country unit of analysis, however, is not necessarily the same. This article highlights difficulties in the data merging process and, by way of example, presents detailed country coding unit comparison for two economic (UN Comtrade and World Development Indicators), two democracy (Polity IV and V-Dem), and two conflict datasets (UCDP/PRIO Armed Conflict Dataset and COW Militarized Interstate Disputes Dataset). We find that merging datasets can result in the elimination of very large numbers of observations due to unmergeable records and that dropped observations often include the very countries or territorial entities most of interest in conflict and peace economics.
Highlights
In conflict and peace economics, the construction of large panel datasets nowadays forms the basis for the majority of empirical cross-country studies
We emphasize the importance of discussing the merging process in empirical studies in conflict and peace economics
In this article we show that the extent of these “missing values” is vast and of particular relevance to empirical research in conflict and peace economics
Summary
Note: *** Statistically significant at the 1% level. PolityIV A second dataset, capturing political authority patterns worldwide and over long periods of time, is the PolityIV project’s dataset on “Political Regime Characteristics and Transitions, 1800–2016” (for short, the PolityIV dataset). In the dataset countries are identified by their name, an alphabetic country code, or a numeric code. These identifiers supposedly follow the COW country coding scheme. Table 1 displays the results from merging the PolityIV data with the COW country list, finding that 13 percent of the countries are unmergeable when merging by country name, 6 percent when merging by numeric code, and 10 percent when merging by alphabetic code. The unmergeable groups largely consist of countries of particular interest in conflict and peace economics such as the Koreas, Congos, Germanies, and Serbias. In both datasets, the unmergeable group had a significantly lower average level of democracy. We first discuss the countries listed in the UN Comtrade data, those in the WDI, and compare the country coding schemes of both datasets. There is no information on whether territories changed, and on whether or how much this change was incorporated in the coding This becomes a severe drawback to the data when complementary variables for the analysis of trade flows, such as country size, GDP, measures of distance and—most importantly—borders are taken into account.. The case of Sudan (see Table A6) illustrates the problem: WDI codes “South Sudan” and “Sudan” For the latter, the measure of trade openness is available for the whole time series (1960–2016). UN Comtrade codes “Sudan” (2012–2015) and “Former Sudan” (1963–2011, with gaps)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
