Documenting and Improving the Hourly Wage Measure in the Danish IDA Database

This paper overhauls the hourly wage measure that is most often used in Danish research, the TIMELON variable in the IDA database. Based on a replication that we have constructed, we provide a documentation of the wage variable, the first of its kind, then continue with a performance analysis. We find four puzzles. 1) The wages of part-timers fall steeply from 1992 to 1993, 2) the wages of full-timers fall from 2003 to 2004, 3) the level of the part-timer wages is around 12.5% higher than it should be, and 4) the wages of new hires fall steeply from the first year of employment to the second year. We analyse these puzzles in depth and solve almost all of them. Finally, we propose a new hourly wage measure that incorporates all the solutions and we show that it performs much better.