The new HLFS data is out. And so it's time to run the old regression again. The specifications on it are all here.
This quarter we find unemployment in NZ among those aged 20+ is 4.99%; the rate for those aged 15-19 is 27.6%. Prior to 2008, the highest youth unemployment rate ever was 23.9%: September quarter 1991. In that quarter, adult unemployment was 9.7%. When adult unemployment was almost double what it's now at, youth unemployment was lower than it is now. Explanations for current youth unemployment outcomes that hinge on the poor overall market seem unlikely to be catching what's going on.
In the graph below, the blue line traces the adult unemployment rate. The red line traces the youth unemployment rate. The green line gives the predicted youth unemployment rate based on a very basic regression taking youth unemployment outcomes as a function of the adult unemployment rate conditioned on all the quarters prior to the vertical red line: the point at which Labour and the Greens abolished the differential lower youth minimum wage. Quarters subsequent to the vertical red line are the new regime. If the current youth unemployment rate followed historic trends relative to the adult unemployment rate, we'd have a still high 15.5% youth unemployment rate; 18.4% if youth unemployment were no worse than the worst ever prior performance relative to the adult rate. But the youth unemployment rate is more than nine percentage points higher than that.
I still think the abolition of the differential lower youth minimum wage is the most plausible explanation of this rather obvious break in the data.
Here are results since 2008 in tabular form. Where the graph above just takes the best fit, the tabulation below adds the largest prior residual to the fitted value to get a worst-case expected youth unemployment rate; this gives the expected rate if youth unemployment performance were no worse than in the worst prior quarter relative to the adult unemployment rate: