The Employment to Population Ratio for Males Aged 25-54

Following up on Brad DeLong's post commenting on General Glut's view that employment to populatiopn statistics show weakness in the labor market, here is the employment to population ratio for males age 25-54 along with an estimated cubic trend. NBER dated expansions are shown in green, contractions in blue (these are annual data so the association is imperfect). The model used to estimate the trend is

EmpToPop = β₀ + β₁t + β₂t² + β₃t³ + ε_t

The estimated trend is the curved black line in the graph. I also eyeballed a piecewise linear version as shown in red:

In general, this ratio rises during expansions and falls during contractions. Notably in reference to the current episode, the ratio generally rises above trend during a recovery. The most recent observations are below the trend level estimated by the (simple) cubic trend model, and this would be true even if the trend sloped slightly downward at the end. Consistent with Brad and the General's claim, this indicates we aren't there yet...