Productivity: The main driver of economic growth for Japan

Date February 17, 2016
Speaker Randall S. JONES (Senior Economist and Head of Japan/Korea Desk, Economics Department, OECD)
Moderator KASHIWABARA Kyoko (Director, International Economic Affairs Division, Trade Policy Bureau, METI)
Time 12:15-13:30 (Registration desk and seminar room open at 12:00)



Randall S. JONES's Photo

Randall S. JONES

Productivity has become a unifying theme for the Organisation for Economic Co-operation and Development (OECD) work in recent years. Productivity determines wage growth and living standards, and as populations in many developed countries have found, the only real source of growth is productivity growth. Today, I would like to look at recent trends in productivity, the causes of declining productivity, and some ideas on how to revive productivity growth.

First of all, when the per capita gross domestic product (GDP) of the top 17 OECD countries is compared, Japan has per capita income of about 25%. This is in the bottom half of the OECD, and there is a big range going down to Mexico and Turkey. The most important reason for the differences in per capita income is labor productivity. There are differences in labor inputs where countries such as Korea and a few others work long hours, but most of the variation is found in productivity per hour of work. Japan, surprisingly, has a very large gap of more than 30% relative to the top half of OECD countries. This is despite an excellent education system, high research and development (R&D) spending, state-of-the-art technology, and hard workers. It seems to me that this gap should not exist, and that Japan could narrow it over time with effective policies.

Global productivity decline

Falling productivity growth did not originate with the crisis in 2008. Productivity growth was stronger everywhere in the 1990s. It fell slightly after the collapse of the bubble economy at the turn of the century, and since the crisis, it has declined even further. Japan is now in line with other countries—United States, European countries—where productivity growth has been very weak. This is a global phenomenon.

One reason that may explain weak labor productivity is investment weakness. Labor productivity is related to how much machinery is allotted to each worker. Investment growth is barely back to where it was in Q1 2008. Business investment is a problem, and that is another mystery that we need to try to unravel. In Japan, we look at things such as corporate governance, and ask why companies are so risk averse. Why are they sitting on huge piles of cash rather than giving them to workers or shareholders or investing them? We find this in other countries as well. There is a certain lack of confidence and uncertainty about the future that is making companies very reluctant to invest. Korea, for example, imposes a tax on companies that do not invest or pay wages or dividends to the extent that the government deems appropriate; the government essentially taxes away their excess cash. This seems a bit like "central planning." Companies like to have the autonomy to manage their cash as they see fit, but this shows the frustration of governments where companies sit on money and wage growth is disappointing, as in Japan, and where dividend payments are underwhelming and investment could be stronger as well.

With regard to the long-term trends in Japan, in 1990, about the time of the bubble collapse, we had 2% trend growth and labor productivity which then fell to 0.7%. It has increased a bit since Abenomics started. Labor input has turned negative which will only worsen because of population decline. The government has a target of 2% real growth and the only way to reach that would be for labor productivity to move from around 1% toward slightly above 2%. The government's productivity growth target is very ambitious.

Looking at long-term trends in Japan compared to other countries, labor productivity in Japan has been stuck at about 72%-75% of the top half of OECD countries. Working time has come down and labor productivity has edged slightly up.

Another measure is multi-factor productivity—efficiency after accounting for labor and capital inputs—a broad measure of technological progress. This also has fallen in Japan, and has been almost zero since 2007.

Reasons behind current trends

What is causing these recent trends? Looking at firm-level data for frontier firms (at the top of their respective fields with high multi-factor productivity) and non-frontier firms, the former has experienced continued labor productivity growth while the latter has lagged behind. In 2007, they turned down and have not yet recovered. We therefore see a gap between the top performers and the weaker performers. This is more marked in the service sector. The winners are still doing well, but the technology and best practices are not filtering down to the other companies.

Comparing global frontier companies to others in terms of productivity, employment, and capital stock, there is an increasing divide between winners and losers, analogous to income and equality, where you have some getting richer and others going backward.

One issue is that dynamism has weakened. Startup companies are the most innovative, bringing new technologies and ways of doing things to the fore. But in many countries, the rate of business creation has fallen dramatically. The average age of companies is getting older. The United States, for example, has come down by about one-third. Japan only has two data points and hasn't come down but is quite low. Why aren't as many companies being created as in the past? What has happened to risk-taking and entrepreneurial initiative? It is worrisome to see things becoming less dynamic and filled with older companies that are less interested in innovation. Japan has many large firms that are quite old. This problem might be more severe in Japan than in other countries.

The same problem exists among small companies. In Brazil, 40% of small companies are new, and only about 10% are old companies founded more than 10 years ago. In Japan there is a large number of small and medium enterprises (SMEs) that have been around a long time. Why do they stay small rather than grow up?

Increasing productivity

The next issue is how to increase productivity. Techno-pessimists say that the technological changes of the 20th century—in transportation and electricity, etc.—were tremendous and resulted in huge productivity gains, and that there aren't many more really dramatic technological innovations existing that will boost productivity. Techno-optimists think the rate of innovation has not slowed and will continue to contribute to technological improvement and growth just as it has throughout history. Some people think technology will save us, others think there is not that much technology left.

Korea has greater Internet penetration than is imagined—even older people use mobile phones to access the Internet—but Korean companies do not use it. It is underused for e-commerce and productivity. That makes me wonder whether technology will really make companies more productive or if they are just entertainment machines.

Japan is a country with a high level of R&D; the third highest in the world after Israel and Sweden. Despite this, total factor productivity growth is sluggish. This raises the question: where is the payoff from investing in R&D?

Taking a look at countries across the board in terms of their openness, basic research, etc., and seeing how those relate to growth: France is doing the most basic research while Belgium is doing the least. What is the spillover if there is a 2% increase in multi-factor productivity at the global frontier? How much does that help each country? Countries that are more involved in basic research receive a big gain and those that are not have less of a gain. The same thing works with university-industry collaboration.

In the case of Japan, business enterprises account for 75.5% of R&D spending, 98% within companies. Only 0.5% is done at universities, and 0.6% at government research institutes. The business enterprises are in a sense in their own world making technology. Universities likewise are 99% at the university, 0.1% within businesses, and 0.6% within the government. The university sector is also often alone, doing its own research. Only 0.5% comes from overseas. Korea, another country that is relatively low, is at 0.7%. For countries at the technological frontier, being part of global innovation chains is very important, so this could be a concern for Japan.

Japanese universities are not very highly ranked. Demographics are changing the environment for universities, and we would hope that as the number of new high school graduates goes down, the university system will be able to consolidate and become better.

Japan is last in the OECD in international co-patenting. Why is Japan not more involved in the global innovation networks? Is it that very powerful companies such as Sony Corporation develop their own technology and are not interested in sharing it or adopting technology from other companies? Costs can be reduced if companies can operate more globally, and risk is reduced if cooperation can be achieved.

We have to find sources of productivity growth that will bring greater gains. The three sources I will talk about are diffusion, resource allocation, and risk-taking. These overlap somewhat. Good diffusion promotes good resource allocation, etc.

Accelerating diffusion

What would happen if companies reduced product market regulation (PMR) from the level of Greece, which is quite high, to the OECD average? We see very large gains in the most backward companies. We stress regulatory reform, which is a very difficult topic. In Japan's case, PMR is about average. If Japan were to move toward the level of Denmark, Japan would also probably be able to achieve the same kinds of regulatory reform gains. Another problem is the gap between manufacturing and services. Since the 1970s, we have seen a steady rise in manufacturing productivity while services have been very stagnant. Manufacturing is more exposed to global competition, which puts pressure on companies either to innovate and become more productive or go out of business, while the service sector is less open to international trade and the PMR in services tends to be much more severe.

In Korea, we counted 15,000 regulations; four times as many in services as in manufacturing. Services are more complicated and many involve social welfare, education, health, and long-term care, etc., and deregulating areas such as healthcare is a sensitive subject. Citizens will wonder what will happen to them if such industries are deregulated. There has been a focus on making the service sector as dynamic as manufacturing in many countries.

In manufacturing, the greater the amount of trade with countries at the global frontier, the larger is the spillover effect when those countries advance. With regard to participation in global value chains, the question is how much production is derived from intermediate inputs produced overseas? Significant gains can be achieved in knowledge-based capital and business R&D. These indicators help a country become more open and gain from spillover at the global frontier.

Japan has relatively high barriers to trade and investment. Hopefully, the Trans-Pacific Partnership (TPP) will cause Japan to eliminate this to a great extent. This is another indicator of openness and ability to benefit from spillover.

Japan has the lowest foreign direct investment (FDI) in the OECD. How can FDI into Japan be promoted? Mergers and acquisitions (M&As) could have a significant impact.

Services are key to globalization. In Japan, the share of domestic services in value added is rising. Japan tends to have high import barriers on key service sectors which, in a sense, facilitate trade and the fusion of technology across frontiers.

With regard to resource allocation, it is important for productivity not only to narrow the gap with the frontier but also to grow. Italy has only a moderate productivity gap, but the problem is that its companies are relatively small. They are not able to grow as much as companies that are close to the global frontier. The United States, on the other hand, has a larger productivity gap with global firms, but the size of its companies is large. Companies do not want to grow; they want to stay small. Is the government too generous to smaller companies by offering tax benefits, etc.?

In Japan, the number of bankruptcies has been trending downward. We had a spike in 2008—as expected, given the crisis—but the number has fallen every year since. In an ideal world, inefficient small companies would exit and their resources would be acquired by more efficient companies. The number of bankruptcies in Japan is toward the low end of the OECD. Are banks and other financial institutions keeping non-viable companies alive because they don't want to have non-performing loans on their books, so-called "evergreening"?

What policies would help with reallocation? Lowering the barriers to entrepreneurship would help, as they increase the difficulty of benefiting from spillover from the global frontier. In addition, the cost of bankruptcy where it is more difficult to exit means you don't benefit as much because inefficient companies survive but remain unable to gain from the global frontier.

For Japan, in terms of PMR, the entrepreneurship barriers are lower than average for the OECD. Is the lack of enthusiasm for entrepreneurship cultural, or is it regulations or financial factors holding things back? In polls, Japanese people express a fairly low level of respect for entrepreneurship. Parents say they want their children to go to the University of Tokyo and then to the Ministry of Economy, Trade and Industry (METI) rather than going into entrepreneurship. The United States is much more positive toward entrepreneurship.

The next issue is mismatch. If companies get workers that lack the necessary skills, it holds them back. Having overskilled workers also holds the economy back because workers can go to other firms and be more productive. Japan is in the middle with regard to the level of mismatch.

Resource allocation

I would like to summarize the research on factors that promote efficient resource allocation. The first is imposing low burdens on new firms and lowering the barriers to entrepreneurship.

The second is less stringent employment protection legislation. That's difficult in Japan where there is a dualistic labor market in which one-third of workers are non-regular, and non-regular workers tend to be fixed-term or dispatched workers who can be easily let go, while the regular workers tend to have a fairly high level of protection. That does not show up in the legal system, but looking at it practically, courts tend to put workers back into companies from which they had been fired. Women who come back after leaving the labor force and young people and older people once they hit age 60 are the ones that get pushed into non-regular employment. From an OECD perspective, that is completely unfair, as the skills among regular and non-regular workers are not very different. Regular workers also get training on the job while non-regular workers do not, so a gap emerges over time due to the dualism.

The third is bankruptcy legislation that does not penalize business failure. In many countries, startups obtain funding based on collateral. If you go bankrupt and lose your house, that discourages innovation.

The fourth is the availability of early-stage financing. That is the risk-taking aspect. Venture capital has taken off in Israel and the United States. Other countries are struggling to develop a venture capital sector. In Korea, most of the money comes from the government, causing a lack of dynamism. Fixing this issue could give rise to a larger number of innovative startups.

Do schools teach you what you need to be an entrepreneur? Japanese people have said that they do not learn very much about entrepreneurship. Some countries begin teaching entrepreneurship in early childhood.


There are four priorities. The first is innovation: keep investing in R&D and pushing the frontier. The second is the diffusion machine: ensuring that the technology invented benefits not only the global leaders but also all companies in a country. The third is improving resource allocation to enable innovative and successful companies to grow and allow inefficient companies to either shrink or disappear. The fourth factor is encouraging more risk taking.

Policies that would accomplish this include pro-competition market reforms focusing on the service sector. The second policy priority should be promoting entrepreneurship. The third would be bankruptcy and exit, the fourth is labor mobility, and the fifth is international openness.


Q1. Is the business sector really a force for innovation or productivity? Businesses cannot take long-term risks. They have to gain short-term returns, thus they cannot invest in the long term. Does the government have to take the lead in innovation?

Randall S. JONES
I agree with you for the most part. Major scientific breakthroughs like the Internet came through government research, as well as aerospace and antibiotics. That type of research takes place in universities or government research institutes, and it constitutes a public good.

On the other hand, business R&D is found to be most linked to multi-factor productivity. It has an impact on products that are actually sold. I think we need a high level of R&D for both businesses and governments.

One thing is striking. Japan is fairly close to the frontier. Over time, productivity growth in the 1970s was 5%. As Japan gets faster, productivity slows down. In countries that are dynamic, slowing productivity growth is a sign of success.

Q2. Japan believes SMEs to be employment preservation vehicles. I don't think this is a Japan-only idea. Do you think we have to change this conceptualization to an idea of SMEs as more active and focused on improving their productivity? What do you think about the balance between these two ideas?

Randall S. JONES
That's very important, and it is a social issue to an extent. My preference would be that the government, through its safety net, worries about people who can't get jobs at SMEs or big companies and then the SMEs become productive. The gap in SME productivity versus big companies is so large that, in some countries, SMEs are becoming unimportant. Growth and productivity don't trickle down from big companies to SMEs.

Q3. The difficulties in Japan's labor system have been talked about for a very long time. Do you think it might be time to talk about how to change that? I also see a chicken-and-egg problem: Japanese universities are not very attractive destinations for international students precisely because of the lack of competitiveness. Do you have any recommendations on how to change that?

Randall S. JONES
On the labor market, the only thing that has been successful is the social safety net idea. Denmark and Sweden have built expensive and generous social welfare systems which bring labor flexibility and higher taxes. The other is grandfathering. Non-regular workers can be paid more but without a lifetime contract, improving their stature while allowing them to retain their flexibility.

For the university system, it is a question of competition. Regional development is linked. With the shrinking population, high-ranking universities should be closed. They should be subject to market forces. Government should ease up as well.

Q4. What is your opinion on the effect of patent systems on reviving productivity growth in Japan? It seems to enable the acquisition of funds as venture capital for startups. How can that be accommodated to revive productivity growth in Japan?

Randall S. JONES
It is hard to make strict rules on how far to go in protecting new inventions. Most economists at the OECD would think that the current systems are too protective and last too long. In all of these wars between Apple Inc. and Samsung Electronics Co. Ltd., judges are trying to decide if it is brand new or a derivative. The system is too protective of inventors' rights. The trend now is global innovation networks where companies float around taking innovation and contributing, and they are less focused on protecting what they have. I think the patent system needs to go in that direction in Japan as well. It means less return to the inventor, but the people who invent often do it more as a hobby or passion than in the hope of striking it rich.

Q5. You mentioned consolidation of Japanese universities. I would like to hear your views on the competition between university academic departments. Last year, the Minister of Education implied that he would cut resources in human and social science departments at public universities, and there was a huge backlash. My impression is to prioritize policy and resources in science/technology fields instead.

Randall S. JONES
This is a sensitive issue. In an ideal world, for me, the government wouldn't have a role in deciding what departments to keep. There should be transparency. Every university should show by department what happens to the graduates. Do they get high-paying jobs? Do they work in their field? Students coming in then could expect certain outcomes. Universities would then see which departments students are entering, those departments would grow, and others would naturally shrink.

There should be more connection between business and universities so that the latter knows what the former wants. I would get the government out of the universities in terms of trying to guide the curriculum. Korea has a huge over-education problem. It had 84% of high school graduates going on to university, and the students ended up as taxi drivers or unemployed. So, Korea developed these 867 national competency skills, such as what you need to know to do a specific job. Junior colleges and universities develop courses that teach these skills, which have been ratified by companies in the country. The curriculum is based on these skills. Students who matriculate from these schools will be welcomed by companies looking for these skills. And this system has been working.

*This summary was compiled by RIETI Editorial staff.