The data appears to come from here[1], which describes it as, > Once we have constructed the individual hourly earnings data, we match the hourly earnings of i

The data appears to come from here[1], which describes it as, > Once we have constructed the individual hourly earnings data, we match the hourly earnings of individuals observed in both the current month and 12 months earlier. The matching algorithm results in an average of 2,400 individual wage growth observations per month. We then compute the median of the distribution of individual 12-month wage changes for each month. That seems to indicate that it does not take inflation into account. [1]: https://www.frbatlanta.org/chcs/wage-growth-tracker.aspx?panel=1 https://www.frbatlanta.org/chcs/wage-growth-tracker.aspx?pan...