Arnold Kling has been posting sample chapters from a hypothetical book on macro. There's too much for me to comment on now, but I feel the need to bookmark his chapters for the futre:
suppose that unemployment did not exist, but that instead there was a very low-value job (say, dishwashing) that is always available, at whatever wage will clear the market. In that case, we would always have full employment. However, I would argue that we could still observe recessions in such an economy. During good times, the only dishwashers would be people who have no other skills. During a recession, many skilled people would be working as dishwashers, and they would drive down the wages of dishwashers. Income would be much lower in a recession, even though there is full employment.
Similarly, I would suggest that a recession creates a flood of refugees in the job market. In ordinary times, the job transitions that people undertake are ones for which they have done some preparation and planning. The market is able to absorb job losers and new labor force entrants fairly routinely, just as a region can absorb fairly routinely the normal ebb and flow of population. However, just as a community cannot quickly absorb a sudden flood of refugees, labor markets cannot absorb a large influx of unemployed workers.