This paper presents a novel monthly dataset of job vacancies in Bogota between 1960 and 2010. The dataset was constructed by counting the number of help-wanted announcements published in the most important newspaper with national circulation-namely, El Tiempo. We describe the methodology used to construct the database, discuss some possible problems associated with it, and propose ways to solve them. The paper also presents an estimation of the job vacancy rate. We depict a Beveridge curve and use the vacancy rate to forecast unemployment. This first look at the series yields sensible results, while leaving a myriad of open questions for future research.
