Despite the uproar over the past weeks regarding the errors in the most recent edition of U. Department of State?s Patterns of Global Terrorism: 2003, released on April 29, 2004, the unfortunate reality is that the document has been internally inconsistent for years, if not decades. The annual publication can be disaggregated into three main sections: regional and topical narrative analyses; a chronology for the calendar year; and a series of statistical tables and graphs. It is not uncommon to read about attacks discussed in the narrative that do not appear in the chronology. The chronology, then, is usually short in describing the date and country of a terrorist event, and often fails to identify the perpetrators, modus operandi, or resolution of the event. Additionally, rarely have the number of events or casualties described in the chronology been accurately reflected in the charts and graphs that conclude the volume. The publication has long been more fruitfully viewed as three smaller reports stitched together with occasional political posturing. For example, a review of the statistical trends reported in Patterns of Global Terrorism over the last decade reveals that, based on frequency:

? Latin America has the most terrorist attacks,