Atif Mian and Amir Sufi write,

The aggregate demand channel for unemployment predicts that employment losses in the non-tradable sector are higher in high leverage U.S. counties that were most severely impacted by the balance sheet shock, while losses in the tradable sector are distributed uniformly across all counties. We find exactly this pattern from 2007 to 2009.

I am not sure that I would use the term aggregate demand to describe a phenomenon of regional variation in employment losses. In addition, we have the Autor-Acemoglu finding to contend with, which is that most of the job loss during the recession was in mid-skill jobs, with low-skill jobs holding their own. And we have to reconcile all that with the variation in job loss by education level.

Overall, I think it is a very complex picture. It would make me rather humble in trying to design policies in Washington to help with employment. It does not appear to me that generic fiscal and monetary expansion is the most promising approach, given that job losses vary so much by region, by occupation, by industry, and by education level.