Establishing causality: one view

Economists agonise, rightly, over causality: how can we tell whether one thing causes another thing rather than the other way round or whether both things might just both caused by some underlying third thing.

Meanwhile, here's how the debate on causality plays out in the political arena.

Under Labour minimum wage went up significantly and unemployment went down significantly. Time for a Living Wage. It creates jobs.
— Trevor Mallard (@TrevorMallard) February 25, 2015

Mallard is far from an idiot, but this where public debate on causality is at.

Granger wept.

PS: via Lindsay Tedds - lagged variables don't cut it either.

VINDICATED: Lagged explanatory variables should never be used for identification purposes. http://t.co/bBZaMy8jTc @comptonjrc @Lachlan07

— Lindsay Tedds (@LindsayTedds) February 24, 2015