If only there were measurable things to try and work out why this might be.
Like levels of absenteeism due to illness or mental health, and then mark these against, I don't know, cuts to mental health services or real term cuts in health spending. And then how that snowballs as those left are...