Shimer (2005) says
These [separation shocks] introduce an almost perfectly positive correlation between unemployment and vacancies, an event that essentially never been observed in the United States at business cycle frequency.
What is the data corresponding to that? Since his publication was from 2005, he must be referring to data on vacancies before the big recession. However, the standard data on job openings, JOLTS, is only available from 2005 onwards.