Research Summary 2017-3
Since the 1980s, US industries have become increasingly dominated by large firms across almost all sectors. David Autor and coauthors find that the market share of the top four firms grew 4% in manufacturing and services industries on average with increases in other sectors ranging up to 15% in retail. Some worry that this rising industry concentration is causing higher prices, less innovation, and greater wage inequality.
But that interpretation depends on what is causing the rise in concentration. One possibility is that large firms have become dominant because antitrust authorities have allowed too many mergers and acquisitions (see Grullon et al.). In this case, declining competition might very well imply high prices and a decline in economic dynamism. A new paper by James Bessen explores another possibility: that leading firms have been better at harnessing information technology (IT) for competitive advantage, allowing them to grow faster. If that is the case, large firm dominance might instead reflect greater innovation by these firms, possibly bringing greater benefits to consumers. The IT story raises some major policy questions, but not necessarily those around antitrust enforcement.
Bessen’s paper, “Information Technology and Industry Concentration” finds a strong link between industry investment in IT systems and the growing dominance of large firms in industries. The focus is not on spending on off-the-shelf computers, but on investment in proprietary systems. Increasingly, business investment in IT is dominated by custom systems. Think of Walmart’s logistics and inventory tracking systems or the systems that large banks use to manage credit card customers. Such systems require large upfront investments and developers and managers with specialized knowledge and skills. For this reason, successful IT systems provide leading firms with marketplace advantages that are difficult for rivals to imitate.
Looking at detailed industries aside from those that produce IT from 2002 through 2012, Bessen finds a strong association between industry concentration and the workforce share of software developers and related occupations involved in developing IT systems. There is also a strong association with the growth in industry concentration from 2002 to 2007. An instrumental variable estimation provides some evidence that the relationship is causal. In addition, Bessen finds that in IT-intensive industries, the top four firms have larger plant sizes, generate more revenue per employee, and have higher operating margins, both in absolute terms and relative to smaller firms in the same industry. In sum, IT systems appear to boost the relative performance of large firms, allowing them to increase their share of the market.
Moreover, the effect is large enough to account for most of the general rise in industry concentration. On the other hand, mergers and acquisitions are not correlated with industry concentration. The general rise in industry concentration does not clearly call for strengthening US antitrust enforcement, although there may well be other reasons to do so. On the other hand, if IT is enabling large firm dominance throughout the economy, this does raise some policy issues. Why is it, for instance, that advanced technology is inaccessible to smaller firms? New technologies sometimes “diffuse” quickly through the economy and sometimes not. The inability of most firms to acquire the best technology holds back productivity growth and may contribute to growing wage inequality. Difficulties training workers in new technologies might be a barrier and so might intellectual property. More research is needed to know.