It's based on his work using functional magnetic resonance imaging brain scans that first demonstrated that, when people are asked to imagine themselves in retirement, the parts of their brains that usually "light up" when they think about themselves don't light up at all.