Firms with fewer than 100 employees account for about 98% of all businesses, but only about 38% of employment. So, if big and small firms have net job losses at the same rate, a small-to-large firm destruction ratio of 0.56 would be expected.
What actually happened?
[F]or the period June 1990 to March 1992, firms with fewer than 100 employees shed on net 160,000 jobs versus 110,000 from firms with more than 100 employees—a small-to-large firm destruction ratio of 1.46.
For the period March 2001 to June 2003, small firms shed 79,000 jobs while larger firms destroyed 324,000 jobs—a small-to-large firm job destruction ratio of 0.24.
Using the latest available data that cover the period September 2007 to June 2009, small firms lost 467,000 jobs compared with 543,000 for larger firms—a small-to-large firm job destruction ratio of 0.86.
Thus, job losses in the 1990-1992 recession and the current financial crisis have been disproportionately in small firms, while job losses in the tech bust (2001-2003), were disproportionately in big firms.
This is quite counterintuitive.
The tech boom and subsequent bust were all about the emergence of small entraprenurial firms that got money and failed, so one would have expected job losses to be disproportionatley from small firms in the cycle, but, actually, big businesses took it on the nose then.
The financial crisis has been about the disasterous missteps of big businesses like large financial institutions and manufacturing companies, so one would have expected current job destruction to be coming mostly from big businesses. But, this too is inaccurate. Job destruction right now is coming disproportionately from big businesses.