A casual look at regional unemployment rates reveals that there are vast differences which cannot be explained by different institutional settings. Our paper attempts to trace these differences in the regions' labour market performance back to the regions' specialisation in products that are more or less advanced in their product cycle. The model we develop shows how individual profit and utility maximisation endogenously leads to decreasing employment in the presence of process innovation. Things deteriorate even further if the region under observation is less innovative than others. Our model suggests that the only way to escape from this vicious circle is to specialize in products that are at the beginning of their economic life.