Performance of Japan’s R&D
Innovation is important in raising growth potential
Growth potential is gradually declining in Japan. According to the estimates of the potential GDP growth rate for Japan by the Cabinet Office shown in Figure 1, real growth rate of the potential GDP as of the third quarter of 2022 is only 0.5 percent. Since the main reason for the fall is the aging and the shrinking of the population, we have to expect that it will fall farther if no counter-measures are taken.
If we want to avoid the fall in potential growth rate, we need to increase labor input and/or raise labor productivity. The former solution requires the nation to realize: (a) a rise in labor force participation rates of females and elderlies, (b) a rise in total fertility rate, and/or (c) an increase in foreign workers. The latter solution requires: (d) an increase in business investment, (e) an increase in human capital investment, (f) an improvement in efficiency of the economic system, and/or (g) acceleration of innovation by research and development.
The focus of this month’s column is related to the last of the solutions mentioned above. By making use of the data provided by the “Japanese Science and Technology Indicators 2022” published by the National Institute of Science and Technology Policy (NISTEP) of the Ministry of Education, Culture, Sports, Science and Technology (MEXT), the past trend of Japan’s research and development (R&D), and the changes taking place in the recent years will be examined.
Inputs to R&D
Research and development (R&D) involves not only activities that would lead to changes in technology (innovation), but also activities to add new knowledge to science (discovery) and to engineering (invention). However, in order to achieve innovation, investing in R&D is essential.
In this respect, Japan has been investing in R&D in a significant magnitude. As Figure 2 shows, until mid-2000s, Japan’s R&D expenditures in yen-terms was less than that of the United States, but more than those of other countries such as Germany, France, the United Kingdom, China, and Korea (converted to yen-terms by using PPP rates). Since late-2000s, however, Japan’s R&D has stopped rising, and has been overtaken by China whose R&D expenditure has been rising significantly since early 2000s, and has also being caught-up by other countries as well.
It should be noted that the affordability of R&D expenditure depends on the size of the economy. If we are to compare R&D expenditure after controlling for the economic size of the countries, we need to look at the ratio of R&D expenditures to GDP. Figure 3 shows that the ratio was high for Japan until mid-2000s, even higher than that of the United States. In late-2000s, however, Japan has been overtaken by Korea, which has been continuously raising its ratio to GDP, and also by the United States in 2020.
We have seen that Japan’s R&D expenditure has been relatively high compared to other countries until mid-2000s. It is characteristic for a small country like Japan. However, that was not the only feature that characterized Japan’s R&D expenditure. How it was spent was also somewhat different compared to other countries.
As Figure 4 shows, the breakdown of R&D expenditure by the users of the expenditure (namely private companies, universities, public institutions, and non-profit making organizations), shows that the share of private companies was relatively higher than other developed countries such as the United States, Germany, France, and the United Kingdom. The share of public institutions, on the other hand, was the smallest among the countries except for the United Kingdom. It suggests that the objectives of the R&D in Japan has been in the application of new technology to business (innovation) rather than in the advancement of basic science and engineering (discovery and inventions).
Governments allocates their fiscal resources to other sectors in order to promote innovation. How they are allocated would have an implication to the breakdown of R&D shown in Figure 4 which is based on who spent and not on who provided the resources.
According to Figure 5, government’s budget on science and technology in Japan was allocated, not so much to private companies, but rather to public institutions. On the one hand, this is similar to the allocation of fiscal resources by the United States, China, and Korea, but these countries allocate more than Japan to private companies. On the other hand, it is in stark contrast to the patterns of Germany, France, and the United Kingdom, who allocate more than half to universities: they are the countries which have strength in basic science and engineering.
So far, we have been focusing on the financial resources spent on R&D. However, financial resources are not the only input to R&D. Inputs to R&D also includes allocation of researchers (the scientists and engineers) to R&D.
As Figure 6 shows, the number of researchers in Japan was larger than those in the European countries and Korea, but less than that in the United States and, more recently, that in China.
The availability of researcher would depend on the country’s population. Therefore, in order to see how the country is devoting its human resources to researchers, the number of researchers needs to be controlled for the size of the population.
Figure 7 shows the number of researchers per ten thousand people. The number for Japan has been relatively high until mid-2000s. However, since late-2000s, it has remained flat, and has been overtaken by Korea whose number has been increasing very quickly since late-1990s.
So far, we have examined the allocation of financial resources and human resources as inputs to R&D. In both cases, the level of inputs used to be relatively high compared to other countries. However, we have seen that the level has remained flat in the recent years, and, as one of the consequences, it has been overtaken by some of the countries that are increasing their inputs rapidly. In that respect, one can say that the relative position of Japan in R&D has declined in the recent years.
Outputs from R&D
Spending more financial resources and allocating more human resources in R&D do not promise any achievement in innovation. In fact, as Figure 8 (bars on the left) shows, the share of companies that engaged in R&D is relatively limited in Japan. It reflects the fact that it is mainly the large companies that engage in R&D and that many small and medium-sized companies did not spend resources in that kind of an activity.
In addition, Figure 8 (bars on the right) shows that the number of companies that achieved product innovation (NB: excludes process innovation), as a share in the number of companies engaging in R&D, is also limited compared to the other countries. It suggests that R&D in Japan has not been so rewarding.
How are the outputs of R&D in Japan? To see that, we will measure the outputs of R&D by looking at the number of scientific papers, and the number of patents.
When measuring the output of R&D by the number of scientific papers, we must be careful about the contribution made to the papers, since the papers could be a product of a joint effort of researchers. A way to take that into account is to count the number of scientific papers on the basis of fractional counting: giving each of the N authors of an N person coauthored paper a weight of 1/N. We must also be careful about the quality of the paper: there are influential papers and ones that are not. In order to pick up the influential ones, we can focus on those that are frequently cited by other papers.
Figure 9 shows the number of top one percent well-cited papers based on fractional counting. It shows that the number of the contribution by Japanese authors had been increasing gradually, but peaked out in 2009 and has been falling since then. Compared to other countries, contribution of Japanese authors was next to US, UK and Germany in early 1980s, but has been overtaken by China and France in 2006, and by Korea in 2020.
Output of R&D can also be measure by the number of patents. Some caution also needs to be taken when using the number of patens. That is because there are again patens that are influential and those that are not. Those that are influential tend to apply for patents in more than one country: we call them “patent families”. By focusing on patent families, we can also avoid counting the same patents more than once when totaling the number of patents across countries.
Figure 10 shows the total number of patent families and the sum of the total number of patent families and the total number of patents applied to only one country. It shows that the latter has shown a significant increase in the recent years, but the former has been showing a much slower increase.
As Figure 11 shows, among the patent families, which are relatively more influential set of patents, share of Japan has been high and has ranked top among the countries considered here for most of the years since early 1980s. However, the share has peaked in mid-2000s and has been falling since then. In contrast, the share of China and Korea is rising rapidly and is closing-in to Germany.
In order to look into the reasons behind the slowing down of the number of patent families in Japan, breakdown of patent families by technology area should be informative. As Figure 12 shows, the industrial area that Japan’s application concentrates are electrical engineering and general machinery. The strength in these technology areas was certainly an advantage in achieving economic growth in the past. The question, however, is whether the strength is still advantageous in the current situation.
According to Figure 13, the fasting growing areas of patent families at present are information and communications technology, electrical engineering, and bio and medical equipment.
As a result of the differences in the rate of change that has taken place, the main areas of family patent at present, according to Figure 14, are information and communications technology, electrical appliance, and mechanical engineering.
Comparing the main areas of Japan’s patent families with the main areas where family patents are growing globally shows that Japan’s weakness is in the lack of technological advancement in information and communications technology and in bio-related technology. This seems to be at the background of the gradual decline in the output from Japan’s R&D.
Diffusion of the new technology
Finally, we should look at the diffusion of the new technology that has been developed.
New technology would not contribute to the economy unless it is implemented to the production process and actually start producing goods and services. If the new technology is accessible to all countries, and all countries actually implement the technology, the labor productivity should not be so different across countries. However, as Figure 15 shows, there is a considerable difference in labor productivity among countries. In particular, Japan’s labor productivity is only 62 percent of that of the United States in 2019.
The reason for the low labor productivity in Japan seems to lie in the services industries: labor productivity in the services industries in Japan is considerably lower than that in other countries.
To see the point, Figure 16 compares the increase in labor productivity in the manufacturing industries with that in the services industries. It shows that, in the recent years, the increase in labor productivity in both the manufacturing and services industries is the lowest in Japan compared to other countries. Also, the difference between the rate of increase in the manufacturing industries and that in the services industries is the largest in Japan among the five countries. In fact there is no difference in the case of the United States. It suggests that not enough implementation of new technology is made in Japan, especially in the services industries, compared to other countries.
Structural problems to be addressed
In this month’s column, we first saw that (A) the inputs to R&D in Japan has been relatively high in the past, but has gradually declined. We also saw that (B) the output from Japan’s R&D tended to be high in the past but has also been falling in the recent years, especially in the area where innovation is actively taking place globally. Towards the end, we found that (C) the new technologies are not implemented enough to the production process, especially in the services industries.
Why did these changes take place? Detailed discussion of the reasons for the changes has to be made in another opportunity. Here, I will confine myself to summarize the main points of what I think are the reasons. I think the changes are taking place in R&D because of the structural problems that the economy faces:
(A) should be a result of the aging and the shrinking of the population which makes it hard to secure enough resources, monetary and human, for R&D.
(B) would be a consequence of the Japanese economic system which make it difficult to have more startups which are at the center of innovation in information and communication technology and in bio-related technologies.
Finally, (c) could be explained by the weak incentive for the services industries to implement new technologies: lack of competitive pressure to improve productivity.
In order to restore R&D in Japan, these structural problems need to be addressed.