In an ideal world what you say is true. Unfortunately, employers don't necessarily know what's the "wise" thing to do. Take the housing boom: from a contractor's point of view it was good business to build many times more houses than the historical aggregate demand. In reality those houses were completely unnecessary: short term gain, long term damage.

Same story for the dot-com (and probably the current) bubble. Employers are notoriously bad at accurately forecasting how many jobs they can actually sustain. And in a downturn it's easy to squeeze more productivity out of fewer employees because it's easier to retain overworked employees.