Matt Yglesias I like, but he does wrap up empirical claims as slightly folksy wisdom. For example, he claims that job growth comes from small firms growing, and from few other sources. That is an important insight because it stops you thinking that big means good. But it does lead you towards a different set of mistakes.
Large firms can get larger, small firms can get larger. They can all add jobs or fail miserably both. In fact, job creation and firm growth is more or less a random process. At least random in the sense that you cannot predict where growth, productivity or jobs will come from next. This including technologically advanced sectors, as Alex Coad argues…
…high-growth firms are the central drivers of job creation in the economy but are neither clustered in high technology sectors nor are necessarily young and small. The evidence on the determinants of firm growth confirms that firm growth is difficult to predict. The finding that firm growth is well approximated by a random process does not only reflect the heterogeneity at the firm level but is also associated with the low persistence of growth rates over time.
Matt still reaches the right conclusions; that disruptive change is to be encouraged (or at least not retarded), but he misunderstands something about disruptive change. Disruptive change, in new products or services, comes from all manner of firms, both large and small.
Large profitable firms can internally fund otherwise risky research and development projects. Twitter started small and got big, Apple started large and got larger. Both brought disruptive technological change, but I bet Apple has created more jobs than Twitter. Don’t make prediction, you will always be wrong.