In 2018, the U.S. Government Accountability Office published a study of non-citizens in U.S. prisons and jails. It found that their criminal records contained 33,300 arrests or transfers for homicides committed within the United States. This study doesn't account for many other murders committed by non-citizens because (1) there "are no reliable population data on all criminal aliens in every U.S. state prison and local jail," (2) the study doesn't identify murders that don't result in arrests, (3) 38% of all murders in the U.S. go unsolved, and (4) rates of unsolved murders are significantly higher in areas with large minority populations. However, the study also double-counts some homicides because the data didn't allow the researchers to "distinguish between a new arrest and a transfer from one agency to another." By reporting on rare events while ignoring others that are far more common, media outlets often create misleading impressions about their relative frequency.