Innovation and firm size
In economics, “technology” is used to describe how much output can be produced from a given amount of inputs. Depending on how it is measured, cross-country variation in technology is arguably the biggest driver of differences in labour productivity between countries and therefore the biggest driver of differences in living standards.
Considering the impact on growth, technological advancements are a significant factor in explaining why some countries achieve better growth than others. Findings from analysis of 23 OECD countries show that investments in applied or experimental research lead to increased productivity growth rates through more efficient use of technology. Investment in basic research takes on average two- to three-year periods to take effect through an expansion of technological opportunities (Sun, Wang & Li 2016).
Innovation benefits consumers by satisfying unmet needs or disrupting the existing status quo by replacing old ways of doing things with new ones. Innovation also creates significant private rates of return for innovative companies. In the UK, the long-run effect of an additional unit of innovation on profit margins has been estimated to be between 6% and 17% (Geroski, Machin & Van Reenen 1993).
Arguably the most important policy question is how government support for innovation should be targeted, which in turn depends on where innovation comes from.
If innovation comes primarily from small firms then it is wasteful to focus support on incumbents, large research consortia or big corporations. On the other hand, if large firms are more innovative than small players in the market, then aggressive antitrust legislation and regulating R&D intensive industries might impede innovation and growth.
Previous research has established two interesting empirical patterns: (a) bigger firms tend to be more R&D intensive with R&D activity increasing with firm size, and (b) SMEs are found to be more productive in carrying out R&D activities – SMEs are observed to hold more patents per dollar spend on R&D compared to large firms.
But if R&D expenditures increase with firm size while at the same time R&D productivity decreases, this potentially counterintuitive finding suggests that large firms are inefficient with their R&D spend. Three potential explanations have been identified:
1. R&D activity could display decreasing returns to scale at the firm level. Returns to scale is the term that economists use to describe what happens to output after a change in inputs given an underlying production function. Under decreasing returns to scale an increase in all research inputs leads to a proportionally smaller increase in research output (i.e. innovation). Indeed, Plehn-Dujowich (2009) showed the existence of decreasing returns to R&D for two different measures of R&D output (patents and citations).
2. The explanation could lie in the limitation of measuring R&D intensity, which could be subject to size-related biases. There is anecdotal evidence that small firms systematically under-report their R&D expenditures due to informal ways of carrying out R&D and informal roles within the company.
3. Large firms may innovate more in terms of processes i.e. finding new ways of producing things that drive down cost of production, whereas small firms tend to engage more with product innovation i.e. developing new products or improve existing ones. Given that product innovation tends to be more heavily patented – as it is more easily copied - this could also be an explanation for why small firms appear to produce more innovations per dollar spend on R&D.
Unfortunately this is an area were the conclusion for policy-making is not straightforward, but tentatively we might conclude that support targeted narrowly at formal R&D activity should be focused more on smaller firms, whereas larger firms should be encouraged to pursue a broader range of innovative activity outside of formal R&D. It is a difficult balance to strike.