What does research tell us about the economic impacts of research and development?
The Hong Kong government takes the view that economic growth is founded on research and development (R&D), and therefore that R&D must be encouraged by policy means. How well-founded is this view? What does research tell us about the relationship between R&D and growth and its policy implications?
A Belgian economist makes the following observation on the relationship between the level of R&D activities and a country’s industrial structure: “A country specialised in the finance industry (e.g. Luxembourg) would not need a high level of R&D expenditure in order to ensure growth … . Similarly, a country specialised in the tourism, fashion, services or food industries would logically have a lower R&D intensity than a country specialised in the pharmaceuticals, engineering or biotech industries.” The message of this observation is that to not a few industries, because of the nature of their businesses, R&D is of less importance or even of no importance.
Many research studies find that R&D investment makes a much smaller impact on company productivity in low R&D intensity sectors than in high R&D intensity sectors. Some even find no significant R&D impact on company output in medium and low R&D intensity sectors.
A very recent study using econometric methods analyzing data over 1,000 firms in OECD countries published in 2016 not only confirmed the findings of previous studies but further discovered that in industries with very low levels of R&D activities, the impact of R&D investment on productivity growth may actually be slightly negative.
For the R&D-productivity growth relationship at the company level, the same 2016 study also corroborated previous research results that “a certain critical mass of R&D capacity is required before significant productivity growth can be achieved from investment in R&D.” Before the critical mass is reached, the impact of R&D investment on company productivity is not significant and may even be slightly negative. The reason is that without a certain accumulated knowledge base, companies are not able to absorb, develop or use new knowledge and hence not able to benefit from R&D investment. Thus, for occasional R&D performing companies or R&D non-performing companies wanting to foster their productivity through doing more R&D, the startup investment cost can be high.
A couple of policy lessons can be drawn from the above research findings for an economy with a low R&D intensity industrial structure wanting to pursue the R&D path to productivity growth. Given an economy’s industrial structure, the productivity gain from more R&D investment will be low unless the economy at the same time expands its high R&D intensity sector if it already has one or introduces one if it does not. Nonetheless, since expanding the existing high R&D intensity sector or introducing a new one inevitably involves establishing new companies, the startup R&D investment for accumulating the required critical mass of R&D capacity for the entire sector before significant productivity growth can be achieved will be huge. It also takes little imagination to see that the total investment required for the sector to expand to a size that can significantly contribute to the GDP and employment growth of the economy will be enormous.
These considerations raise the issue of cost-effectiveness of the R&D-led growth strategy in an economy concentrated on low R&D intensity industries. The strategy is costly and the time needed for the economy to be able to gain from it will be in the order of decades. On the other hand, because of the overall size of the low R&D intensity industries in the economy, the gain from the growth of these industries is very likely to outweigh the gain from that of the high R&D intensity sector (unless the latter’s expansion in size is considerable). Furthermore, were the low R&D intensity industries to become stagnant, the overall welfare of the economy would be severely diminished. Thus, for a low R&D intensity economy, pursuing other growth paths other than R&D would look more cost-effective and appropriate. The policy question then becomes: what alternatives are available.
To this question, the following comment on R&D and innovation made by an economist at Columbia University may provide some pointers: “It is important to understand that not all of the activities and investments made by firms in innovating are conducted in R&D laboratories, or get counted as R&D.” The messages are:
- innovation is not equal to R&D
- R&D is only one of many sources of innovation
- firms can be innovative without doing R&D
- non-R&D performing firms are not necessarily less innovative
The innovation literature delineates four main categories of non-R&D innovation endeavors:
- technology adoption
- minor modifications or incremental changes to products and processes
- imitation including reverse engineering
- combining existing knowledge in new ways
The economic significance of these four areas of non-R&D innovation is not less than that of R&D innovation.
The Innobarometer 2007 survey indicates that over 50% of the companies considered innovative in the 27 European Union countries are non-R&D performers. Meanwhile, other research studies indicate that 40% of innovative companies in the German manufacturing industry do not do R&D. In the US, data from the Business R&D and Innovation Survey 2011 conducted by the US Census Bureau suggest that 72% of product innovating companies are non-R&D performing firms.
In terms of resources devoted to non-R&D innovation, in the European Union, the total expenditure on non-R&D innovation as a percentage of GDP is almost on par with that of R&D innovation. In the period from 2004 to 2008, the average total spending on non-R&D innovation was 1.35% of GDP, compared to 1.48% of GDP on R&D innovation.
In terms of companies’ economic performance, not a few recent studies find that non-R&D performers are no different from R&D performers in their sales, profitability, productivity or mortality.
Regarding the respective impact of non-R&D innovation and R&D innovation on total productivity growth, a very recent econometric analysis published in 2017 provides robust evidence that both types of innovation are statistically significant and economically important. Although it is found that the impact of non-R&D innovation is only one half of that of R&D innovation, given the expensive startup costs and total investment of R&D innovation, the importance of non-R&D innovation to total productivity growth must not be underestimated.
Also noteworthy is that in the context of the European Union, research has found that non-R&D innovation plays a role in companies in both high R&D intensity sectors and low R&D intensity sectors and non-R&D innovative companies are found in all sectors and company sizes. Nonetheless, this type of innovation is especially important for company in low R&D intensity sectors and small and medium size companies.
Available evidence shows that companies doing little or no R&D can be innovative and industries doing little or no R&D can be economically vibrant.
For an economy concentrated on low R&D intensity industries, the most straightforward policy option for fostering growth would be to introduce measures to support non-R&D innovation. Although compared to R&D innovation’s impact on total productivity growth, non-R&D innovation’s impact is weaker, given the considerable size of low R&D intensity industries, its impact on the overall wellbeing of the economy will not be weak. At the same time, the non-R&D growth path is less costly.
A more balanced policy option nonetheless is to foster growth while at the same time allow room for diversifying the economy. This option involves introducing an innovation policy incorporating both the R&D and non-R&D aspects of innovation, rather than single-mindedly focusing on either. The aim of such a policy is to provide support for companies to develop their innovative capabilities, of which R&D is only one of many aspects. Support measures should be expressly conceived according to the particular types of companies. Measures of policy support for high R&D intensity sectors should be different from those addressing low R&D intensity sectors. From the perspective of this policy, deciding a balanced allocation of support for non-R&D innovation and R&D innovation is as important an issue as deciding the types of support. Failing to give the low R&D intensity sectors sufficient support and letting them stagnate will seriously jeopardize the health of the economy.
In Hong Kong, at least 80% of GDP is contributed by service industries, such as import/export, wholesale and retail trade, accommodation and food services, transportation and storage, postal and courier services, financing and insurance, real estate, professional and business services, and social and personal services. R&D has little or no relevance to these industries. Because of this economic structure, contrary to the Hong Kong government’s understanding, research tells us that the R&D-led growth strategy may not be an appropriate growth-fostering strategy for Hong Kong. Therefore, what Hong Kong needs is a comprehensive innovation policy providing balanced support for both R&D innovation and non-R&D innovation. The goal is not to make Hong Kong’s economy more R&D intensive but to make it more innovative. R&D is only one of many inputs for this endeavor. The underlying consideration should be finding a balance between maintaining the economic vibrancy of the existing industrial structure and exploring the growth potential of investing in high R&D intensity industries. Regarding this consideration, it is very important not to overlook that the health of the service industries mentioned above is vital to the health of Hong Kong’s economy.
George Cautherley (高德禮)
Hong Kong Democratic Foundation (香港民主促進會)
8 January 2019
This article entitled “Is R&D the magic solution for Hong Kong?” was published on “ejinsight” <http://www.ejinsight.com/>, the English service of the Hong Kong Economic Journal, on 24 January 2019, <http://www.ejinsight.com/20190124-Is-R&D-the-magic-solution-for-Hong-Kong/>.
The Chinese version of this article entitled “研發與經濟增長的關” was published in Hong Kong Economic Journal on 25 January 2019.
 Bruno van Pottelsberghe de la Potterie. 2008. Europe’s R&D: missing the wrong targets? Intereconomics 43(3): p. 222.
 d’Artis Kancsa and Boriss Siliverstovs. 2016. R&D and non-linear productivity growth. Research Policy 45(3): 634–646.
 Ibid., p. 643.
 Richard Nelson. 2000. National innovation systems. In Acs, Zoltan. Ed. Regional Innovation, Knowledge and Global Change. London and New York: Pinter Publishers, p. 13.
 Anthony Arundel, Catalina Bordoy, and Minna Kanerva. 2008. Neglected Innovators: How Do Innovative Firms that Do Not Perform R&D Innovate? Results of An Analysis of the Innobarometer 2007 Survey No. 215. INNO-Metrics Thematic Paper. Brussels: European Commission, DG Enterprise.
 Anthony Arundel, Catalina Bordoy, and Minna Kanerva. 2008. Ibid.
 Oliver Som. 2012. Innovation with R&D: The Roles of Non-R&D-Performing Firms in the Industrial Innovation System. PowerPoint Presentation presented at EU-SPRI Conference, 12-13 June 2012.
 You-Na Lee. 2015. Expanding Understanding of the Innovation Process: R&D and Non-R&D Innovation. PhD Dissertation, Georgia Institute of Technology, p. x.
 Jesus Lopez-Rodrigueza and Diego Martinez-Lopez. 2017. Looking beyond the R&D effects on innovation: The contribution of non-R&D activities to total factor productivity growth in the EU. Structural Change and Economic Dynamics 40(March): 37-45.
 Oliver Som. 2012. Op Cit.
 Jesus Lopez-Rodrigueza and Diego Martinez-Lopez. 2017. Op Cit.
 Anthony Arundel, Catalina Bordoy, and Minna Kanerva. 2008. Op Cit. and Lesley Potters. 2009. R&D in Low Tech Sectors. IPTS Working Paper on Corporate R&D and Innovation No.08/2009, Joint Research Centre – Institute for Prospective Technological Studies. Luxembourg: Office for Official Publications of the European Communities.
 Lesley Potters. 2009. Ibid.