step,rule,n_before,n_after,n_removed,rationale
1,drop quarters after 2019Q4 from merged quarterly panel (GDPC1 x complete-quarter UNRATE averages),314,288,26,information rule: quarters after 2019Q4 must not enter any estimate
2,keep merged quarters 1960Q1-2019Q4 with non-missing du and g,288,240,48,regression sample per TASK; 1947Q1-1959Q4 quarters (GDP/UNRATE overlap begins 1948Q1) only supply lagged levels