I haven't looked very closely at the NBER paper's methodology but the conclusion flies in the face of the raw data on total compensation as a share of GDP/GDI f

I haven't looked very closely at the NBER paper's methodology but the conclusion flies in the face of the raw data on total compensation as a share of GDP/GDI from the BEA[1] and other sources[2]. The NBER paper bases its conclusion on a suggestion that GDP figures are not accurately estimated; this seems like somewhat of a strange position but I am not qualified to evaluate it on merits. The raw data indicate that total compensation as a percentage of GDP/GDI has been falling since at least the early 1970s. These percentages are derived by comparing nominal dollar values within the same year (e.g. 2018 total compensation divided by 2018 GDP, both in 2018 nominal dollars). Unlike estimates of long-term real wages, these percentages are not dependent on inflation assumptions. The long term trend paints a clear picture that people who earn most of their income (including benefits) from wage labor have been losing ground economically for decades. Even without metrics, this conclusion is supported by living memory history. In 1970, a single middle-class full-time income could support a comfortable lifestyle for a family of four. In 2020, a single middle-class full time income can't support housing costs in most (possibly all?) urban centers. We're going backwards. [1] https://fred.stlouisfed.org/series/A4002E1A156NBEA https://fred.stlouisfed.org/series/A4002E1A156NBEA [2] https://fred.stlouisfed.org/series/LABSHPUSA156NRUG https://fred.stlouisfed.org/series/LABSHPUSA156NRUG