New digital technologies and big data have created opportunities to address persistently sluggish productivity growth and mitigate the effect of an ageing and shrinking population on the economy amongst other challenges. The Japanese government has embarked on an ambitious plan of digital transformation and businesses are often at the forefront of new technologies. The pandemic accelerated the digital transformation as households, businesses and government switched to new technologies and ways of organising themselves in order to minimise health risks. But the pandemic also exposed shortcomings. Many were underprepared having not made needed investments to exploit these new technologies. While government is pursuing initiatives across a number of areas, weaknesses in linking databases hindered its ability to target support for those most affected by the pandemic. Investment by business is concentrated in larger firms and the dynamism offered by new firms entering the market is limited. Students and workers, while highly trained, can lack the skills needed to exploit fully new digital tools.
OECD Economic Surveys: Japan 2021
2. Making the most of digitalisation following COVID-19
Abstract
The impact of the pandemic on the ongoing digital transformation
In 2016, the Japanese government set out a vision for the future society named Society 5.0, meaning "a human-centred society that balances economic advancement with the resolution of social problems by a system that highly integrates cyberspace and physical space". Harnessing digitalisation’s potential is a core part of the Society 5.0 initiative. Society 5.0 is seen as a means to achieve sustainable economic development by addressing social issues, low productivity, an ageing society, and energy and climate problems with digital technologies and innovations. This approach emphasises organisational changes to leverage what is already happening at the micro level, such as smart city or smart medicine, and roll them out across the economy and country.
One of the impacts of the pandemic has been to accelerate Japan’s ongoing digital transformation and organisational changes, for example in the way that public services are provided or how firms outsource tasks to third-party providers. However, the difficulties that firms and households faced during the pandemic also highlighted weaknesses. For example, the shift to remote working was hindered by work practices and poor telecommunication infrastructure in some homes. Indeed, over one-third of respondents in one survey felt their productivity had declined due to a poor telecommunication environment at home (Morikawa, 2021[1]). Household interactions with government required physical presence in some areas of the country, whereas in others all such services were already available remotely (Box 2.1).
In many respects, Japanese firms and households are well placed to benefit from greater digitalisation. The communications infrastructure is well developed and the skills of the population are elevated by international comparison. Companies are at the forefront of certain technology areas and investment in research and development is high. Japanese students and adults perform well in international rankings of skills and the population is increasingly educated (OECD, 2019[2]), (OECD, 2019[3]).
However, the availability of skills and technologies is not sufficient to guarantee success in a competitive environment. The full benefits require complementary investments and redirection of some efforts. Using new technologies effectively often requires changing the way firms and organisations work and the acquisition of new skills to use these new technologies. This places demands on decision makers, both in the public and private sector, to adapt. The population also can seize new opportunities, making the economy more nimble in reacting to changes.
Against this background, this chapter first assesses the existing conditions for the digital transformation. The role of government is central to making progress, both in its own use of digital technologies and through facilitating the digital transformation by firms and the population. The business sector has much to gain against a backdrop of weak productivity growth and an ageing workforce. The next section discusses how the business sector is adapting to the digital transformation and the key bottlenecks to further development. Digitalisation could give a needed boost of dynamism, but much depends on adequate investment in ICT technology and the organisational capacity and skills. The final section addresses how the flow and stock of skills need to adapt to greater digitalisation.
Box 2.1. Digital information and technology during the pandemic
The pandemic and the measures taken to cope with it suddenly altered economic agents’ behaviour. Some of the main traditional measures of economic activity turned out to no longer be informative or timely enough. Non-traditional sources of data, such as internet search requests, mobility tracking from mobile phones and banking card activity became important means of monitoring the state of the economy, the impact of containment measures and individuals’ proximity to COVID-19 carriers.
Digital technologies were mobilised to mitigate transmission risks during the pandemic. For example, a baseball match in Yokohama in November 2020 tested whether crowds could congregate safely as preparation for the Olympics and Paralympics in 2021. New technology used monitoring devices within the stadium and a smartphone application to inform spectators about congestion in different parts of the stadium.
The pandemic accelerated reforms already happening in some enterprises and government agencies. To reduce exposure to the coronavirus, teleworking was extended. This also led managers to rethink how to organise their work. In some cases, face-to-face interactions could be replaced by virtual interfaces or robots. Like in other countries, the pandemic boosted the number of firms using digital payment platforms and increased the public’s acceptance of digital settlement. In comparison with 2019, online purchases rose by 11% on average in 2020, and by almost 20% in major cities. This is an important change in Japan, where so far cash dominated consumer purchases (cashless payments’ share of total transactions was only 20% in 2016, compared with over 50% in the euro area and over 70% in the United States (Esselink and Hernández, 2017[4]) (Kumur and O’Brien, 2019[5]).
After the initial lockdowns implemented across the OECD, subsequent measures have had smaller economic impacts, as businesses and households adapted to the restrictions, often by making greater use of digital technologies. Overall, digitalisation has contributed to economic resilience by allowing greater flexibility in working arrangements and reducing the need for face-to-face interactions at the point of sale. This helps not only in the face of a pandemic but also to cope with natural disasters.
Where does Japan stand with respect to digitalisation?
Communication infrastructure and internet usage
Japan performs well on a number of communication infrastructure indicators. For example, the overall number of people with access to broadband is high. This is especially true for the number of mobile broadband subscriptions per 100 inhabitants, which is the highest in the OECD (Figure 2.1). Major telecom companies have already begun to roll out 5G technologies. In contrast, fixed broadband subscriptions are around the OECD average at 32 per 100 residents (OECD, 2020[6]), even though the share of future-proof fibre in fixed broadband subscriptions ranks amongst the highest in the OECD (Figure 2.2). The share of firms with broadband connections was around the OECD average in 2019, and the share for companies with over 250 employees was the lowest in the OECD.
The use of communication services, however, lags behind the high number of broadband subscriptions. Before the pandemic, the share of the population using the internet for online commerce was around the OECD average (Figure 2.3). In 2019, mobile data usage was below the OECD average, notwithstanding the high subscription rate. Education’s use of ICT technologies also appears relatively weak. The share of individuals attending an online course was below the OECD average, at 9.4% against 13.6 in 2019. Finally, the share of individuals using e-government services has been amongst the lowest in the OECD. In 2019, just 7.9% of the population have used the internet to fill in forms via the public authorities’ websites (the unweighted OECD average was over 40% of 16 to 74-year olds). However, the coronavirus pandemic resulted in a number of changes. Communication operators noted marked increases in mobile data usage, partly also driven by new pricing plans for broadband that reduced costs dramatically. Education also switched to distance learning wherever feasible and government has been pushing for greater digitalisation of its services.
Figure 2.1. Access to mobile broadband is widespread
Mobile broadband subscriptions per 100 inhabitants, by technology, June 2020
Source: OECD, Broadband database.
Figure 2.2. Most connections are high-speed broadband
Percentage of fibre connections in total fixed broadband
Note: 2010 rather than 2009 for Canada, Turkey and United Kingdom.
Source: OECD Broadband database.
Japanese businesses are at the digitalisation frontier in many ways. The use of robots in manufacturing is third only to that of Korea and Singapore, with the number of robots per 10000 workers estimated to be 364 in 2019. Furthermore, Japan accounts for 47% of global robot production (IFR, 2021[7]). In this respect, Japanese manufacturers have maintained their lead over the United States and Europe (Dekle, 2020[8]). The automotive industry makes extensive use of welding robots while precision manufacturing, such as electronics, makes more use of assembling robots. These two sectors account for around two-thirds of all robots in use in Japan. Japan is also an important player in digital technologies development, accounting for a sizeable share of the patents in emerging digital technologies (Figure 2.4). Firms are amongst the leaders in the OECD in using cloud computing, which is more prevalent amongst larger firms and with the gap to smaller firms growing over time (Figure 2.5). On the other hand, radio-frequency identification, which collects data remotely throughout the production process, is comparatively less developed.
Figure 2.3. Online purchasing is around the OECD average
Share of individuals participating in e-commerce, 2020 or latest available year
Note: Data refer to 2019 for Columbia, France, Japan, Israel, Switzerland, and United States, 2017 for Chile, and 2016 for Australia.
Source: OECD ICT Access and Usage by Individuals database.
Figure 2.4. Japan is an important developer of emerging digital technologies
Share of top five economies’ patents in top 25 technologies fast accelerating from 2010 onwards, 2013-16
Note: IP5 denotes the five largest intellectual property offices worldwide (European Patent Office, Japan Patent Office, Korean Intellectual Property Office, National Intellectual Property Administration of People’s Republic of China and United States Patent and Trademark Office). Fast-accelerating ICT fields are defined as technology fields where substantial increases in the number of patents are observed.
Source: OECD (2019b), calculations based on STI Micro-data Lab: Intellectual Property Database.
Figure 2.5. Some digital technologies are widely used
As a percentage of enterprises with ten or more employees, 2020 or latest year
Note: RFID stands for Radio frequency identification; CRM for Customer relationship management; ERP for Enterprise resource planning. Data for cloud computing refer to 2020, except for Canada, Japan (2019) Australia, Korea, United Kingdom (2018) Switzerland and United States (2017). Data for big data refer to 2020, except for Canada (2019) and Korea (2018). Data for RFID refer to 2017 except for Canada, Japan and Korea (2019). Data for CRM refer to 2019, except for Switzerland and Iceland (2017). Data for ERP refer to 2019, except for Switzerland and Iceland (2017). The statistical unit for Korea is the establishment and not the enterprise and covers public and private sectors. The scope of coverage for Japan extends to the firms that have 100 or more employees, while for Australia, Korea and New Zealand all enterprises are included.
Source: OECD ICT Access and Usage by Businesses database.
The backbone of past high growth has been manufacturing, particularly machine tools, automotive products and steel. As economies are changing and the service sector becomes increasingly important, the drivers of productivity growth are also changing. Greater use of digital technologies creates new opportunities, both in manufacturing and elsewhere. For example, the rollout of 5G telecommunication networks facilitates the transmission of large data volumes at far higher speed, with low latency and hyper-connectivity. Applications using these attributes stretch from the smart cities local governments are already exploring, e-health, autonomous vehicles, the internet of things, smart grids and factories. This will introduce new goods and services - and by integrating machine learning and artificial intelligence techniques - make production more reactive to changes in demand and supply as well as potentially lifting productivity. In this regard, the digital transformation will support productivity and make the economy more resilient to shocks. Central to the digital transformation is investment in new technologies and the diffusion of knowledge.
Promoting the government’s role in the digital transformation
Governmental digital transformation: capacities and challenges
Japan is relatively advanced in e-government, especially with respect to infrastructure. The Japanese government has made efforts to make its data more open and to promote their reuse by the public and private sectors. According to the OECD Open, Useful and Re-usable data (OURdata) index the country ranks highly in terms of the design and implementation of open data policies (Figure 2.6).
Figure 2.6. Government policy settings for data provision appear relatively good
Open Useful and Re-usable data index (2019)
Note: Data for 2019 are not available for Hungary, Iceland, Turkey and the United States. Data availability assess formal requirements to promote open government data. Data accessibility assesses policies to promote unrestricted access to understandable data, ensuring data quality and implementation of these requirements once data are published. Government support to re-use measures how proactive governments are in promoting the re-use of data within and outside government.
Source: OECD OURdata Index 2019.
However, the digital services provided by the government are not fully utilized, though with variations across levels of government. The use of the internet to interact with the public sector is not very common in comparison to other OECD countries (Figure 2.7). A Cabinet Secretariat and Ministry of Internal Affairs and Communications survey conducted in 2020 on the use of online administrative procedures found that the major central government services for the private sector were relatively well digitised with an online application rate of 50.6% (Table 2.1). However, the procedures of local governments are less well digitised and heterogeneous across areas. The survey also found that the main private sector users of government online services were firms, while individuals were much less likely to use online applications. This also would reflect on the relevance of adopting people-driven approaches towards understanding the needs of citizens and embedding these into the design and delivery of digitally-enabled public services (OECD, 2020[9]).
Table 2.1. Administrative procedures and applications
Note: Online application ratio calculated by the number of applications.* 165 of 307 procedures could not be done online by municipalities or local branches. Some online applications were not reported in the research because of difficulty to count by individual local entities.
Source: (Cabinet Secretariat and Ministry of Internal Affairs and Communications, 2021[10]) and OECD calculation.
Figure 2.7. Internet use for public services is limited in Japan
Individuals using the Internet for sending filled forms via public authorities' websites, 2019 or latest
Note: As a percentage of individuals in age 16-74 or 75.The data for Canada and Costa Rica refer to 2018. Different methodologies are used in EU and non-EU countries. For example, there are differences in the format of the questions. In addition, the sample for Japan covers individuals aged 15-74, while the sample for majority of the countries refers to those aged between 16 and 74.
Source: OECD, ICT Access and Usage by Households and Individuals database.
The pandemic highlighted some e-government shortcomings. The lack of linked information in different government databases affected the government’s ability to support individuals and households. For example, online procedures occasionally took longer than paper ones to organise the cash handouts of JPY100,000 to every resident in some municipalities. This arose because local governments during the state of emergency had insufficient administrative capacity and lacked the necessary database for the online processing needed with the national personal ID (My Number) to connect personal information during the pandemic. As a result, local government officials had to check the consistency manually. It took around three months to deliver the benefit to over 90% of households. Moreover, local governments could not easily determine household income, which led to delays in getting support to low-income households affected by the pandemic. This contrasts with the experience in Estonia for example, where 99% of administrative procedures are digitised, which enabled the authorities to provide a one-off benefit within about a fortnight. Ultimately, efficient public sector digital arrangements make it possible to provide services to the public much more quickly and to target them as needed, without any overwhelming pressures on the public sector during periods of stress, such as in a pandemic.
Digital transformation and innovation in public service delivery
The digital transformation of government can achieve several objectives. First, it can improve the efficiency of administrative processes. For example, individuals do not need to provide their personal information repeatedly. When databases are interconnected and accessible to different government departments and across levels of government the administrative burden associated with moving can be reduced. In addition, inefficient processes can be eliminated. For example, the use of hanko (stamps) on physical documents often serves no practical purpose and increases transaction costs. Almost all of the uses of hanko could be replaced. Indeed, the city of Fukuoka systematically reassessed the use of hanko in their procedures and eliminated around 3 800 of them. These administrative procedures could then be conducted online or even in convenience stores. During the pandemic this reduced the need for face-to-face interactions. The central government has also made a strong push to eliminate hanko and more generally for regulatory reforms to encourage more effective use of digital technologies.
Digital transformation can also raise productivity in the public sector, especially against a backdrop of an ageing and declining population, and contribute to improving conditions in the private sector. In the construction sector, the Ministry of Land, Infrastructure, Transport and Tourism (MLIT) aims to achieve by 2022 an infrastructure and transport database and platform, with the resulting “Digital twins” providing a means to raise productivity, enhance the provision of operating information and promote research and development by helping co-ordinate industry, academia and government. In addition, the MLIT is promoting “i-Construction”, which uses ICT technologies throughout the construction process: survey, measurement, design, construction, inspection and maintenance. Digitalisation also facilitates inspection of different types of infrastructure assets, such as by using drones (e.g. for checking structural integrity of otherwise hard to access locations) or robots (e.g. in-water or in zones undergoing natural disasters), with the identification of emerging issues helped by machine learning and artificial intelligence. These new methods allow cheaper more regular checking, potentially identifying when repairs and maintenance are needed and so reducing the need for investment in replacements. In addition, automated remote-controlled construction machines using 5G technologies (high precision real-time control based on global positioning satellites) reduce labour intensity and improve safety in construction sites. In 2016, the Japanese government set a target to improve productivity at construction sites by around 20% by 2025. In the construction sites under the direct control of MLIT, i-Construction reduced the total time of civil engineering ICT earthworks by around 30% in 2018 (MLIT, 2020[11]).
Smooth and efficient transportation can also improve productivity. In the port sector, Japan is building a data platform that enables port operators to conduct vehicle logistic procedures digitally. At present, the trials are relatively limited in coverage (notably the port of Yokohama use of the “Container Fast Pass” or “CONPAS”). These types of systems are already well developed in some other countries. For example, the Port of Freemantle in Australia uses a vehicle booking system, OneStop, that manages the available slots for containers through the day using a compulsory app (Productivity Commission, 2020[12]). The Freemantle port also uses a fully automated IT system to manage vessel congestion. Australian ports have rolled out driverless vehicles for moving containers around their ports. Against this background, digitalising port logistics, administration and use of infrastructure and integrating data more effectively will help improve port productivity and strengthen their international competitiveness.
For vehicle transport, mobility as a service, such as on-demand ride sharing applications, has been developed in some local areas in Japan. With the support of the government, such as the Smart Mobility Challenge Project, local governments are developing “Mobility-as-a-Service” initiatives that often harness digitalisation. For example, in Shiojiri an on-demand bus service (without predetermined routes or schedules) uses artificial intelligence to estimate the most efficient routes for buses. The city is also planning to trial driverless vehicles for some urban routes to address shortages of drivers. As the city’s population is relatively elderly, passengers are often dependent on bus services for mobility. Digitalisation has also helped rural areas experiencing population declines exploit big data to predict demand and rationalise their provision of public transport (OECD, 2016[13]). At the same time, automated driving is currently being tested with a view to widespread practical use on highways by 2025. Following the revision of the Road Transport Vehicle Act in 2019, a Japanese car with level 3 autonomy (automated driving in special situations with support by a driver if necessary) received authorisation in November 2020 (MLIT, 2020[14]) and came onto the market in March 2021. The autonomous car market will develop through technology development and commercialisation, but also regulatory reform.
Regional revitalisation strategies and the Abe government’s structural reform agenda aimed to improve productivity in rural areas. The Ministry of Agriculture, Forestry and Fisheries (MAFF) plans to digitalise all administrative procedures under its purview (over 3 000) by 2022 and to promote data aggregation including on climate, farmland, cultivated crops and geographical information on a data platform. Combining these data will make them more easily useable and support the development of “smart agriculture”, with the aim of improving productivity and efficient use of land and other natural resources. There is also potential to boost productivity in the forestry sector. Around two-thirds of all land is forest in Japan, although much is in mountainous regions. Innovations in forestry using automation (such as automated logging and transportation) and ICT technologies (“smart forestry”) can cut administrative costs and improve felling productivity while improving safety.
The digital transformation of health care
With an ageing and shrinking population, improving labour productivity in social security and medical and long-term care is key to limit pressure on budgets. Digitalisation and AI technology potentially offer benefits by improving the quality and efficiency of such services. Big data analysis of personal health and medical data will be useful for the effective prevention of diseases and adjusting health and long-term care services to meet the needs of the population. This could help reduce the observed high rate of unnecessary treatments, procedures and therapies. In this context, better use of timely and high-quality data could improve public sector cost efficiency (OECD, 2019[15]). This calls for investing in capacities and governance frameworks for enhanced access to and sharing of data in the public sector (OECD, 2019[16]), as well as adopting principles and frameworks to govern their use (OECD, 2021[17]). Furthermore, digital devices, such as real-time health check sensors and care robots, as well as tele-medicine can improve productivity while reducing pressure on health care workers, by allowing them to concentrate on where their interventions are most needed. For example, care robots in nursing homes make for shorter and more flexible working hours for staff, and reduce the psychological burden they bear (Eggleston, Lee and Iizuka, 2021[18]). Expanding the use of these approaches, however, requires the acceptance of the medical staff and the public. The Japanese government has set a (conservative) target to improve productivity per hour by more than 5% by 2040 through the introduction and use of robots, ICT equipment and health data (MHLW, 2019[19]).
To utilise digital tools, getting the regulatory framework right is essential. However, the readiness to use medical data in Japan was limited (Figure 2.8). In addition, some medical databases, such as the National Database of Health Insurance Claims and Specific Health Check-ups (NDB) and the Diagnosis Procedure Combination (DPC) Database, had been constructed separately and are not linked. The failure to ensure data governance frameworks that enable access to and sharing of data in the public sector could limit the scope for big data analysis. Therefore, efforts are underway to link some of these databases. The government began analysing how to link the NDB and the long-term care database in October 2020. In addition, the government is now aiming to link several health record systems (such as the NDB and DPC databases) with anonymized IDs (name, sex, birthday, health insurance policy numbers, etc). An additional aim is to link these databases with personal information such as residence and income, as well as improve the connection between health and long-term care databases. Making progress on coverage and data integration is important. The room to improve public service productivity is considerable. One estimate suggests that doubling investment to manage health sector data and information would more than pay for itself (OECD, 2019[15]). However, data integration should be carried out in such a way as to ensure that sensitive information is only released to those who need to have access and that publicly available data cannot be used to identify individuals or firms. For example, the Census Bureau in the United States since 2007 has used noise infusion procedures to ensure confidentiality while still allowing disaggregated data to be used by researchers.
At the same time, use of the national personal ID (My Number) card as a health insurance card was rolled out nationally in October 2021. This allows easier identification in hospitals and doctors can access personal medicine records or specific health check-up records more easily online. Wider use of the My Number card could deliver additional benefits by connecting other databases, improving the quality of health care, cost effectiveness and ultimately welfare. For example, it will ease the declaration of medical costs when filing income tax returns. Even though the take-up of the My Number card has been slow, the coverage of My Number card issuance has reached 35%, as of August 2021, six years after its launch. The Japanese government is continuing to encourage more people to apply through various measures, which has led to an increase by more than 15% over a year earlier. To achieve the target that almost all citizens will be using the My Number cards by March 2023, more incentives will be necessary.
Figure 2.8. Japan has made progress in using electronic health records
Electronic Health Record data governance and technical/operational readiness, 2016
Note: “Japan 2021 preliminary” was calculated by the secretariat.
Source: OECD Survey of Electronic Health Record System Development and Use, 2016; Oderkirk (2017) “Readiness of electronic health record systems to contribute to national health information and research”, https://dx.doi.org/10.1787/9e296bf3-en.
Local governments and the smart city
Many public services for citizens are provided by local governments. In this light, the digital transformation and reforms at local government level are critical for the public’s well-being and quality of life. To solve urban and regional problems and create new value by utilising advanced technology and management, the Japanese government has been supporting smart city programmes. Between 2017 and 2020, 200 projects have been launched across 160 areas, covering education, monitoring/safety, disaster prevention, mobility/logistics, energy/environment, finance or public health. The evidence so far suggests that local governments can harness the power of new digital technologies with some, such as Aizuwakamatsu (Box 2.2), making significant progress already in creating a smart city.
However, many projects remain at the stage of “proof of concept”. Only 23 projects were rolled out into actual service during 2020. The current target is that 100 programmes will be in use by 2025. The experience so far has highlighted several challenges. These include ensuring the profitability of services, participation of citizens in sharing and using data, and lack of human resources to sustain and develop new services. As many local areas are rapidly ageing and the older population is less comfortable with digital tools, digitalisation might leave some groups behind. It also raises other concerns. To enhance citizens’ participation and understanding while safeguarding data security and privacy, it is important to establish clear incentives and to explain and make visible how personal data will be secured and used. For these purposes, traditional facilities and interactions with government agencies are still available. For example, in rural areas, staff in Japan Agricultural Cooperatives support (often elderly) farmers in conducting digital administrative procedures, or local government officials conduct visits to train the elderly in the use of ICT devices. Support for the use of ICT technology and services or the access to digital services could also be provided in certain public facilities, such as libraries for instance, as is done in some other OECD countries.
Most of the smart city projects are developed individually and are not interconnected. This works in an early proof-of-concept phase, but could raise R&D and organisational costs later on. The current projects and their future development should be evaluated with an aim to disseminate good practice and warn of potential pitfalls. Sharing experiences of successes and challenges, as well as the applications, modules or data for digital services that have already been developed would also help improve cost effectiveness when deploying digital services in other areas. Along these lines, the government developed the “Smart City Reference Architecture” in 2020, including a standard design framework of smart cities and examples of components, such as business model, database and operating systems, in each service area. In addition, the government produced a guidebook to introduce and implement new smart city projects. More efforts along these lines, sharing experience and tools, will be needed as projects begin to be implemented.
Box 2.2. Aizuwakamatsu: digitalisation in a historical city
Aizuwakamatsu-shi (City) is located in Fukushima-ken (Prefecture), surrounded by mountains and a big lake. The city is known as a Samurai city, having witnessed a fierce battle during the transition from the Edo era. The city population of 120 thousand people is ageing relatively fast. Nonetheless, this historical city is developing as a leading smart city in Japan. The initiative started in 2011, soon after the Tohoku (Great East Japan) Earthquake. The city has introduced a management operating system with an information platform, dubbed “Aizuwakamatsu Plus”, which provides tailored information on more than 10 digital services, such as transportation, medical care, child care, and tourism. An office complex, “Smart City AiCT”, built in 2019, now has 31 companies and over 400 employees. One key element of the initiative is human resource development. The University of Aizu, specialized in education for ICT technologies has been training more than 1000 data scientists, and collaborating with private companies to address the challenges faced by the city. The city uses an “opt-in” model: residents can choose if they want to provide personal information in exchange for digital services. The city officials explain to the residents how personal data are handled, and the benefit of the services. Currently, around 12,000 people have registered with “Aizuwakamatsu Plus” and this is expected to reach 50,000 by 2030, as the city establishes trust while offering human-centred and personalized services. Foreign tourists can benefit from digital services by using the “Visit Aizu” app which provides AI-supported recommendations according to their nationality and preferences.
The Digital Agency and beyond
The Japanese government has prioritised the digital transformation and improving the productivity and efficiency of public service provision. However, significant challenges remain. As mentioned above, the inter-connectivity of databases among central and local governments, and between departments is underdeveloped. The use of private data requires further investment in research, development and deployment. Current ineffective data management procedures and resources reduce the effectiveness of administrative procedures, especially during natural and other emergencies. Public sector officials are often not familiar with designing nor utilising digital tools. A better and more joined-up government should be capable of establishing the governance mechanisms to co-ordinate and integrate public sector organisations for the digital transformation of Japan, including the availability of common tools, standards and platforms to enable a coherent digital transformation. This includes strengthening the data governance approach to enable data sharing across levels of governments, which can improve targeting of social benefits and public good provision, which caused problems during the pandemic.
To address these issues, the Japanese government established a new “Digital Agency” in September 2021, with a staff of 600 officials, including 200 ICT engineers. Although Japan has been working to implement the Data Exchange Standard for Administration, challenges remain in developing a public sector data policy that frames access, sharing and reuse of high-quality data across public sector organisations (OECD, 2020[20]). The Digital Agency operates in three areas: designing the digital architecture, changing rules and regulations, and establishing digital public goods. The Digital Agency, through making and implementing action plans, has power over other ministries to supervise their actions for digitalisation and to allocate budget for the ministries’ ICT systems. One of the key tasks will be to establish a base registry to achieve interconnectivity and efficient storage and use of data. The new agency will then be able to combine and integrate the national government’s information systems. This in turn will allow an integrated, user-oriented approach to reforming administrative services and operational systems. In addition, the government will make arrangements so that citizens can use their current national personal ID cards (My Number Social Security cards and Tax Number cards), to complete administrative procedures.
However, the goal is not just to establish a new agency and achieve system interconnectivity, but to improve citizens’ well-being and quality of life by raising productivity in the public and private sectors. To achieve this goal, the central and local governments need to be actively supporting the digital transformation. This will require more than just investing in new digital technologies: complementary actions will be needed with respect to administrative reforms, regulatory reforms, work-style labour market reforms and business process reengineering.
Through the digital transformation and related reforms, the government’s ambition is to reduce the operating costs of information systems by 30% by 2025. Success here could make further reductions of administrative costs for the private sector feasible. Compliance with regulation is onerous in many sectors (Morikawa, 2020[21]). There is ample room for administrative reform to improve the businesses environment, thereby boosting productivity (Haidar and Hoshi, 2015[22]). The Digital Agency and organisations such as the Council for the Promotion of Regulatory Reform should work together to reduce compliance costs and other regulatory burdens that may impede the digital transformation.
Furthermore, central and local governments will be able to provide proactive (or “push-type”) services, becoming more user-driven by anticipating citizens’ needs based on the access, sharing and reuse of data - e.g. when they become eligible or when action is required. For example, some local governments, such as Aizuwakamatsu, already provide relevant information automatically to citizens, such as information about childcare facilities, regular vaccinations, health checks and special benefits to families with young children. To achieve high cost-efficiency for these new services, the authorities should set concrete targets and milestones as they move beyond proof-of-concepts trials to broader implementation, and also train their officials to acquire new skills including through training in the private sector.
While the government can take actions on its own to digitise its activities, it can also help improve framework conditions in the private sector and for households. For example, it can unify the multitude of different approaches in a consistent regulatory framework for dealing with personal privacy (such as GDPR in the European Union), or issue guidelines to clarify hours and conditions for teleworking. At the same time, there exist risks that individuals can be identified when databases are linked. In such cases, using techniques such as noise infusion will be required to ensure anonymity. A level playing field is needed for the re-use of data in public, private and academic projects. Greater use of the new digital technologies would be supported by laying out the advantages for sharing personal data and using digital tools. Recent evidence confirms that digital tools lead to lower fees and shorter and easier procedures (Cabinet Secretariat and Ministry of Internal Affairs and Communications, 2021[10]).
The digital transformation and related reforms in the central and local governments should be inclusive. For the vulnerable and people who may not be able to readily use digital technologies, such as many elderly people, low-income households and people with disabilities, easy-to-use software, user-support, and a safety net to use digital tools and services are necessary. At the same time, if the digital transformation aggravates existing inequality or job displacement, support for affected people and retraining should be provided.
The promises of digitalisation for businesses
The impact of the digital transformation is being felt throughout economies as lower costs for data transmission, storage and analysis have expanded applications beyond industrial processes. Digital technologies are automating tasks in both manufacturing and services. Firms are making use of big data and innovative data analysis to improve their performance in supplying goods. And as government begins to use digital tools more effectively, new and better data will become available for businesses to exploit. While there is potential, there is no guarantee that firms will pursue a digital transformation without investment and a supportive policy environment. In their absence there is a danger that the least-productive firms will lag behind further as more productive firms adopt new technologies (Berlingier et al., 2020[23]). And indeed, a divergence between mainly large firms that are highly productive and weaker, often smaller, firms have emerged, threatening to further polarise the economy (OECD, 2019[24]).
A recent report assessing digital competitiveness across countries ranked Japan only as mediocre (IMD, 2020[25]). This is mainly due to management being insufficiently nimble to react to emerging challenges, inadequate complementary investments in research and development and intangibles (including training and organisational reforms) and a regulatory framework that is less conducive to digitalisation. In part, management flexibility is constrained by entrenched employment practices, especially seniority wages and high costs of employee dismissal, which is discussed above and in previous OECD Economic Surveys (Jones and Seitani, 2019[26]). A second issue is a lack of skilled personnel to make best use of new technologies. This is particularly the case in small enterprises and outside the main urban areas. This section examines the concerns constraining investment in both tangible and intangible assets to exploit the potential for the digital transformation to lift productivity.
Weak investment has held back productivity
The level of productivity is relatively weak compared with the OECD average and has failed to converge to higher levels in recent decades (Figure 2.9). Partly, the weak performance occurred as employment shifted out of manufacturing into business services. In addition, weak investment growth has also contributed to the productivity slowdown over the longer run (Goldin et al., 2021[27]). The factors holding back productivity growth, such as low investment in many firms and weak business dynamism, have implications for the digital transformation. The possibilities created by digitalisation to incorporate business services and manufacturing have been exploited to a limited extent (Box 2.3). Realising the opportunities of digitalisation across the economy will likely require business services to become more productive.
The slowdown in productivity has been widespread across OECD economies after the mid-2000s. Nonetheless, in a number of countries the manufacturing sector (especially machinery and electrical equipment in Japan) continued to experience strong productivity growth (Figure 2.9, Panel C). In Japan, after a long malaise, productivity growth picked up in the construction sector. In other sectors of the Japanese economy productivity slowed, declining in a few and stagnating elsewhere. Utilities exerted a negative drag over the decade after the mid-2000s, reflecting the shutdown of nuclear power stations following the 2011 Fukushima disaster. The distribution of labour productivity growth rates by industry between 2010 and 2018 reveals that most industries had very low growth rates with a sizeable share of slightly negative growth, giving the distribution a marked skew (Panel D). Restricting the sample to manufacturing reveals the distribution was less skewed and growth rates somewhat stronger. Labour productivity growth in business services was weak in comparison with other advanced economies (Panel A) and in the ICT sector slightly negative. By contrast, ICT productivity rose rapidly in other advanced economies over the same period (Baily, Bosworth and Doshi, 2020[28]).
Figure 2.9. Labour productivity is relatively weak
Note: In panel B, productivity is defined as GDP per hour worked; In panel C, productivity is defined as real gross value added per hour worked, except for Japan and United States, where it refers to real gross value added per person employed. Business services excludes real estate, data for Canada in business services refer to 2010-2017, data for Japan and United States refer to 2010-2019.
Source: OECD Productivity Statistics (database); and JIP database.
Rapid capital stock growth at the beginning of the 1990s gave way to a slump following the global financial crisis. Private sector non-residential investment subsequently picked up somewhat accompanying the structural reforms aimed at boosting productivity growth. The growth rate of the capital stock has fallen once more with the onset of the pandemic (Figure 2.12). Firm-level empirical work suggests that the adoption of digital technologies boosts firm productivity (Gal et al., 2019[29]). It shows that stepping up the adoption of digital technologies - such as cloud computing, enterprise resource planning and customer relationship management - can lift multi-factor productivity. It also suggests that more productive firms are better able to transform investments in digitalisation into higher levels of productivity.
Weak productivity growth partly stems from pronounced differences between large and smaller firms. Labour productivity in smaller firms is only around 40% of that in large enterprises (OECD, 2020[30]). Differences in investment account for some of these differences. Investment in advanced and ICT technologies has been concentrated in large incumbent firms whereas diffusion to small and medium-sized enterprises or through new entry has been limited (Fukao et al., 2015[31]). While the digital transformation promises to boost productivity, the difficulties encountered in diffusing and adopting advanced technologies are serious impediments.
Investments in intangible assets are increasingly important for productivity growth and are complements to investments in ICT equipment. In particular, “digital capital” that encompasses investment in R&D, software and organisational capital appears to be an important determinant of firm performance (Tambe et al., 2021[32]). For example, empirical evidence suggests that investments in organisational capital are important for (new) Japanese firms’ growth (Hosono, Takizawa and Yamanouchi, 2020[33]). However, investment in intangibles also appears to be heavily skewed, with large enterprises undertaking the majority of such investments in Japan.
Box 2.3. Business services in exports
Japan is around the OECD average in the importance of digital services in total service exports (Figure 2.10). To a much larger extent than other countries these digital services are mainly for intellectual property use. The share of ICT-related services has been amongst the lowest in the OECD.
Figure 2.10. Digital-related service exports are largely intellectual property products
As a percentage of total services exports, 2019
Source: OECD calculations based on EBOPS 2010, https://www.oecd.org/sdd/its/EBOPS-2010.pdf; and WTO, Trade in Commercial Services, November 2021.
Business services can also contribute to total exports through their contribution to manufacturing exports. Digitalisation and new business models have created opportunities for services to be embedded in manufactured products, a phenomenon known as “servicification”. Manufacturing firms increasingly rely on services as inputs, during production, and as outputs bundled with goods (Miroudot and Cadestin, 2017[34]). This is not just that a higher share of value added originates in the service sector, but a shift to more productive activities that are more tightly linked to the consumer.
Servicification often accompanies the outsourcing by major firms of their inputs and the globalisation of production networks. However, the contribution of services to manufacturing exports in Japan in the past has been comparatively low (Figure 2.11). Some empirical evidence suggests that servicification is weak in manufacturing firms and that they remain largely dependent on manufacturing value added (Chun, Hur and Son, 2021[35]).
Figure 2.11. The contribution of services to manufacturing exports is low
Value added share of manufacturing gross exports by industry inputs, 2018
Note: Data refer to total economy covering ISIC Rev.4 sectors 01-99, manufacturing refers to sectors between 10-33, and services refers to business services covering sectors 45-82, other refers to the remaining sectors. Domestic value added refers to direct and indirect, as well as re-imported, value added content of exports.
Source: OECD, Trade in Value Added 2021 edition.
Figure 2.12. Growth of the private capital stock has slowed
Net capital stock of fixed assets year over year growth rate, chain linked
Source: Cabinet Office.
Some structural factors hold back the digital transformation outside the larger firms. Relatively few workers are employed in ICT task-intensive occupations. This has a bearing on small firms, especially in more remote locations, as the lack of ICT-trained workers constrains their ability to invest in new technologies. As discussed below an older workforce and low job turnover can hinder diffusion. In addition, small enterprises are often very small and with elderly owners without a designated successor, which tends to reduce incentives to invest in new technology. The median age of SME managers has risen from 47 years in 1995 to 66 in 2015, limiting managerial flexibility in adapting to new digitalisation opportunities.
Furthermore, business process outsourcing, including of ICT tasks, remains relatively underdeveloped. A large share of outsourcing of IT functions, for example, is to affiliated companies. As a result, the available resources for small firms outside these networks can be limited. Concerns about digital security can also become a barrier to pursuing digitalisation without sufficient expertise. The combined effect makes digitalisation more complex and costly for many small firms. However, as already seen, Japanese firms are making extensive use of cloud computing, suggesting that the business process outsourcing market could develop further. Some cities have recognised these difficulties and are endeavouring to improve the enterprise ecosystem (Box 2.4).
Box 2.4. Fukuoka: transforming government and promoting business
The city of Fukuoka has moved fast in promoting digitalisation, recognising the importance of creating a vibrant business environment to encourage the digitalisation of businesses. Fukuoka is a major city on the island of Kyushu. With an airport and a seaport, it is an important regional transportation hub. More recently, it has garnered attention by becoming the city with the highest business start-up rate in Japan. The digitalisation of government and the promotion of start-ups has been a deliberate policy aim since 2012, with the aim of making Fukuoka a dynamic yet liveable city.
The city has been at the forefront of the digital transformation in the provision of public services. The city identified around 3 800 procedures that can now be completed without a stamp (hanko) and promoted conducting administrative tasks online or even in convenience stores, rather than by visiting government offices. The authorities recognise that vulnerable groups could be left behind in the push for digitalisation. As a result, in-person facilities and training in the new means of interaction are provided.
As part of the so-called “silicon sea belt” linking areas of east Asia specialising in semiconductors, Fukuoka hosts semiconductor design companies and institutions with large-scale integration (LSI) researchers. The city is promoting the semiconductor sector with the Fukuoka Institute of System LSI Design Industry playing a central role in these efforts.
Fukuoka city is trying to create a business environment to encourage start-ups, including by attracting foreigners. This involves creating supportive organisations (such as business incubators), easing regulatory procedures, agreeing on special treaties for start-up visas and giving some corporate tax advantages in agreement with the central government. As a result, it is the only economic zone for start-ups in the country.
Government support for digitalisation of small firms in other countries also confronts reluctance to invest in new technologies and shortages of qualified personnel. For example, the approach adopted to support investment in robots in France recognised that training of managers and their workforces was essential and so offered a mixture of financial and technical assistance (Faquet and Malardé, 2020[36]).
Spending on R&D can support the development of new products and production processes that can exploit the possibilities opened up by the digital transformation. Spending on R&D is large in comparison to other OECD countries, but relatively little is invested in the ICT sector (Figure 2.13). Over 80% of enterprise R&D spending is in manufacturing, with over one quarter accounted for by the automotive sector alone. Partly as a result, spending is heavily concentrated in large enterprises. Small and medium-sized enterprises play a very small role, lacking the capacity to undertake cutting-edge research. Just 10% of R&D spending is done by small and medium-sized enterprises, whereas the average across the OECD is closer to one-half.
Despite relatively robust spending on R&D and a sizeable registration of patents, measures of patent quality suggest many patents are not as used in follow-on research and this has deteriorated over time (Bahar and Strauss, 2020[37]). By creating a “patent thicket” of intellectual property rights, incumbent firms may deter new entrants attempting to commercialise their inventions and discourage investment (van Zimmeren, 2009[38]). This is a potentially important issue for the digital transformation with the development of new products and processes spanning several intellectual property domains. The competition authorities may need to monitor whether improper strategic use of patenting is hindering the digital transformation. If evidence of abuse does emerge, competition authorities should consider appropriate investigation and enforcement (Shapiro and Lemley, 2019[39]).
Public support to encourage R&D spending in firms facing financial constraints is available. Overall public spending on R&D in Japan is somewhat lower than the OECD average, with the composition heavily skewed towards tax support rather than direct funding (Figure 2.14). Tax incentives have been found to be effective in promoting R&D spending (Hall and Van Reenen, 2000[40]), although they may also support R&D that would have been undertaken in any case. Furthermore, tax incentives appear to promote experimental development rather than research (Appelt. S. et al., 2020[41]). Small firms’ investment can be particularly responsive to these incentives as they are more likely to be credit constrained (Kobayashi, 2014[42]). The effect can be augmented by making tax incentives refundable or allowing them to be carried forward given that new and small firms may have insufficient tax liability. A carry-forward provision was abolished in 2015. As a result, the implied tax subsidy rate for a loss-making SME is well below the OECD average, whereas the rate for a profit-making small firm is around the OECD average (OECD, 2020[43]). In Japan, the share of small companies in overall R&D tax subsidies and also direct financing is very small, and they receive less than 10% of total government support for business enterprise expenditure on R&D (OECD, 2020[44]). Against this background, altering the structure of support could help target different objectives and groups of firms. For example, reinstituting the tax carry-over for new and young firms or making the credit refundable would help them invest in their early stages. Increasing direct grants to new firms would help target support. Other countries, such as New Zealand, restrict tax incentives to small and medium-sized enterprises.
Policy to support investment in ICT tends to target key sectors or technologies. For example, the “Strategic Information and Communications R&D Promotion” programme gives incentives for collaboration between ICT firms, universities and local governments. Supporting firms that have less capacity to adopt new technologies by forging partnerships with universities and research institutes can overcome constraints within the firm. The lack of skilled personnel within firms need not be a limiting constraint. The share of firms collaborating with universities or research institutes for innovation is around the OECD average, though larger firms, particularly in manufacturing, appear to have better linkages (OECD, 2020[45]). Empirical evidence suggests that small technology-based firms can exploit such university research networks as a source of knowledge (Fukugawa, 2012[46]). The second area of government support is through the promotion of innovation hubs focussing on the internet of things with the aim of accelerating the adoption and diffusion of these technologies, particularly in small and medium-sized enterprises and start-ups (OECD, 2019[47]).
More widespread investment in ICT appears pressing. An assessment of the digital transformation on the business sector noted that many firms have antiquated core digital systems and are not investing in updating them (METI, 2018[48]). This leads to the so-called “2025 cliff problem” whereby 60% of all systems will be legacy systems and will cause JPY 12 trillion losses by 2030. In addition, companies and government struggle with differences in ICT systems, which hinder information sharing and mergers of companies (to wit past large bank mergers) and require costly investment to harmonise operations. This situation is aggravated by the shortage of trained personnel, as discussed below. In order to address these investment shortfalls, accelerated tax depreciation allowances for investment in ICT could be considered. However, this may come at a large revenue cost and would thus need to be targeted at the firms most in need of upgrading their information and communication technology. As a result, administration costs would rise and a regular assessment would be required to check whether policy objectives are being met and whether other means to achieve the same ends offer better value for money.
Figure 2.13. Business R&D spending is high, but not in the ICT sector
Source: OECD, Research and Development Statistics database; and OECD, Main Science and Technology Indicators.
Evidence from elsewhere in the OECD suggests that developing the provision of digital services can offset the need for digital investment. Small firms, in particular, can view digital investment and digital services, such as offered by cloud computing, as substitutes (Andres et al., 2020[49]). For example, Spain in 2015 began promoting cloud computing and other digital services for small and medium-sized enterprises, with support on offer for both the demand and the supply side (Serrano Calle, Pérez Martínez and Frias Barroso, 2016[50]). In Italy, during the pandemic the Digital Solidarity programme facilitated access for the self-employed and small enterprises to digital services available from large private sector companies including access to mobile data and cloud computing (OECD, 2021[51]). In this light, some of the shortfall in investment can be met by promoting the development of digital business services and ensuring small and medium-sized enterprises have access to these resources.
Figure 2.14. Public support of R&D spending is largely tax based
Government support for business R&D, as % of GDP, 2018
Note: Subnational tax support for BERD is included in tax support for BERD in Canada, Hungary and Japan. For Spain and the United States, these data are unavailable.
Source: OECD R&D Tax Incentives Indicators.
Low business dynamism holds back innovation
A final feature of investment is the relative lack of new firms that subsequently grow. New entry and growth also support competition and resource reallocation. Entry rates are among the weakest in the OECD and while they have increased somewhat over time, they remain comparatively low (Figure 2.15). Entry and exit rates are similar in digitally-intensive sectors to the rates elsewhere in the economy. Furthermore, since many of the firms that exit are small this leads to a shift of the workforce to larger firms and towards the service sector. Recognising the importance of a vibrant business sector, the Japan Revitalisation Strategy of 2013 set the target of raising entry and exit rates to 10% by 2020.
The lack of business dynamism stultifies the “productivity ladder", whereby growing more-productive firms draw in workers from lower-productivity firms, which appears stronger in other countries. This has consequences for workers with relevant skills outside large firms, limiting their opportunities to make a better match and rewarding their investment in skill acquisition. As a result, wage dispersion, reflecting underlying firm productivity, is pronounced and workers have limited opportunities to improve wages by moving to new firms.
The policy environment for starting up a new company is relatively straightforward for licensing. The licensing regime is lean and a “one-stop-shop” provides the necessary information on licensing requirements and can issue licenses and permits. Furthermore, a “silence is consent” rule reduces waiting times for approvals. However, the administrative burden on setting up a business is more cumbersome than the OECD average (Figure 2.16), which is mainly driven by the number of procedures that are required to register a new business. As noted above, the government is trying to reduce the requirements for hanko (a physical seal). The cost of producing a hanko is not insignificant, at JPY 10,000 to 20,000 (USD 90 to180). For example, from February 2021, it is no longer mandatory to register a carved company seal when applying online for a registration of incorporation. In an increasingly digital business environment switching to digital seals would permit greater flexibility. Already companies are offering digital seals for individuals to stamp documents accessed on the cloud, reflecting the custom of stamping documents rather than a legal necessity to use a physical stamp. Demand for digital seals has risen during the pandemic as individuals teleworked and wanted to avoid proximity.
Figure 2.15. Firm entry rates are particularly weak
Average entry and exit rates, 1998-2015
Source: OECD (2019) Measuring the Digital Transition: A Roadmap for the Future.
New businesses typically face difficulties as they move from incubation to the growth stage, with cash flow being one of the key constraints. This can pose difficulties given that much of the firm’s assets at these early stages are intangible. Bank preferences for collateral and personal guarantees become a potential obstacle. To some extent, this can be overcome by venture capital funds. At present, the size of the venture capital market is comparatively small (Figure 2.17). In the past, most venture capital investments were in start-up and other early-stage phases. The more developed venture capital markets typically make a larger share of investment in later-stage ventures. Experience from elsewhere in the OECD (such as Cooperative Venturing in the United States or the Scottish Angel Capital Programme) suggests that developing the network of “business angels” who not only invest in start-ups, but who can give advice and access to their networks, can help thicken venture capital markets (OECD, 2020[52]).
Government support in Japan for small and medium-sized firms (not necessarily new ones) takes several forms. Funds can be channelled through credit guarantee corporations (Japan Federation of Credit Guarantee Corporation and Japan Finance Corporation). However, such subsidies can be easier to access for small and medium-sized enterprises with established relationships with banks. Efforts are also needed for new firms and the government offers income tax advantages for business angels (the Angel tax system) investing in start-ups. Furthermore, the Japan Patent Office introduced the IP Business Valuation Report and the IP business proposal in 2014 to help financial institutions better understand intellectual property issues when lending to small and medium-sized enterprises. This has supported the development of this part of the market based on intangibles, which can be important for new firms. Some banks, such as Chiba Bank, are now making loans backed by intellectual property.
Figure 2.16. Administrative burdens for setting up a business are relatively large
Countries ranked in ascending order from most (0) to least (6) competition friendly
Note: For limited liability companies and personally-owned enterprises.
Source: OECD 2018 PMR database.
The Small and Medium Enterprise Agency takes the lead in addressing the difficulties experienced by small and medium-sized enterprises. The agency runs several programmes attempting to support small firms, their growth and avoid their failing. One issue is that small enterprises are often very small and the rate of self-employed who are also employers, as a rate of entrepreneurs, is the lowest amongst OECD countries (Figure 2.18). This is true for both male and female entrepreneurs. They include promoting mergers and acquisitions through tax incentives, which may help consolidate managerial resources. A second approach is promoting business succession through a deferral and exemption framework for inheritance and gift taxes. The policies aiming to prevent firm failures may in the process hinder new entry and keep afloat less productive firms and as such a balance needs to be struck. Empirical evidence suggests that productivity growth is lower in (unlisted) family-owned firms in Japan (Morikawa, 2013[53]). Consequently, policy needs to prioritise targeted support to promising small and medium-sized enterprises, with enhancing entrepreneurship, especially for females. In other countries, such as the Netherlands, the support for small firms is embedded in a wider framework encouraging the entry of new firms and supporting them as they grow (Box 2.5).
Business dynamism can also be boosted by facilitating exit of less productive firms. When entry and exit complement one another, the effect can be substantial and contribute to lifting overall productivity growth. Alternatively, the survival of low-productivity firms can hamper the reallocation of resources and inhibit entry of new firms. Evidence from across the OECD suggests that the “congestion” created by zombie firms hampers productivity growth, especially for young firms (Adalet McGowan, Muge, Andrews and Millot, 2017[54]). Empirical evidence from Japan increasingly suggests that more productive small enterprises tend to disappear and the less productive firms survive, such that firm exit has a negative effect on productivity overall (Ikeuchi et al., 2020[55]). Part of this is due to large firms acquiring promising start-ups or other small firms. Even so, some high-productivity firms appear to be failing, contributing to the overall weak productivity performance.
The corporate insolvency regime appears comparatively well designed, keeping costs of failure relatively low and imposing relatively few barriers to restructuring (Adalet McGowan, Andrews and Millot, 2017[56]). In practice, however, poorly performing firms have tended to survive. For the individual entrepreneur, the limited protection of personal assets and the heavy use of (real estate) collateral and personal loans by the banking system can create incentives for rolling over loans (OECD, 2017[57]). In addition, firm failure in practice can lead to personal bankruptcy and loss of the entrepreneur’s home. In addition to the loss of the home, fear of social stigma amplifies the risk of declaring bankruptcy. In other countries, or states within countries, wealth exemptions (such as the family home) from personal bankruptcy proceedings appear to support higher entry and exit rates. However, financing may become more difficult as a result.
Figure 2.17. The venture capital market is relatively small
Venture capital investments, 2019 as per cent of GDP
Source: OECD Venture Capital Investments database.
Figure 2.18. Entrepreneurship in Japan is quite low
Share of self-employed who are employers in 2019 or latest
Note: Chile, Mexico, Korea, Australia are in 2017. Israel is in 2016.
Source: OECD Gender Data Portal.
Box 2.5. Developing business ecosystems in the Netherlands
In the Netherlands, the StartupDelta and subsequently the TechLeap.NL programmes have been running since 2015 promoting the development of regional start-up initiatives, and fostering linkages between these regions. The programmes are public-private partnerships bringing together key actors in government (such as the ministries of economic affairs and education, culture and science) universities, businesses and business associations.
A number of services help start-ups access key resources, including through establishing searchable knowledge databases, creating a platform to facilitate business-to-business linkages for innovation partnerships and investment; and support for research collaboration with universities and companies with R&D capacity. A start-up box guides start-ups dealing with public services and the support they offer. One success was lobbying to reduce the processing time for R&D tax deduction applications for new firms to just one month. The programmes also work with government and are indeed part of national entrepreneurship policy, creating a forum to highlight emerging bottlenecks and promote policies that will help start-ups and scale-ups. This includes attracting key talent, such as skilled ICT workers, including from abroad.
A separate initiative run by a consortium of Dutch universities, also initiated in 2015, aims to strengthen entrepreneurship education. This initiative initially received support from the government. The Dutch Centre for Entrepreneurship (DutchCE) helped bring together existing smaller projects and they benefited from the pooling of capacity to promote research and knowledge sharing. The effect has strengthened the entrepreneurship training available in higher education institutes. The network of centres works to ensure the valorisation of research by transferring knowledge to entrepreneurs, business and government bodies. The network also works with StartupDelta and TechLeapNL. However, this initiative has struggled as entrepreneurship is not a core university subject and some centres have either closed or refocussed their activities. This highlights the importance of aligning university incentives with the business sector’s demand for skills.
Source: (OECD, 2020[52]).
Entry and exit can be boosted by ensuring competitive pressures are strong. There is a trade-off in digital sectors between economies of scale and scope and competitive pressures. This is particularly so when economies of scale are important (such as with two-sided platforms) or the costs of meeting regulatory requirements are significant (such as biomedical products). In these cases, firm dynamics may alter as innovations created in small or new entrants are taken to market by large firms with an already developed customer base or capacity to meet regulatory requirements. At present entry and exit rates are not discernibly different for highly digital sectors and the rest of the economy. However, the competition authorities should remain vigilant in policing incumbents’ actions to deter entry, which can stifle innovation (Baker et al., 2020[58]). The Fair Trade Commission has revised the different sets of merger guidelines to be able to identify actions that are detrimental to competition. How competition policy needs to adapt to digitalisation is an active area of debate and best practice is evolving (OECD, 2020[59]).
Following the start of the pandemic, various subsidies, concessional loans and loan guarantees shielded enterprises from the consequences of revenue collapses and helped with rental payments or business restructuring. Other subsidies were designed to help enterprises invest to minimise the impact of the pandemic on business, either in digital equipment to allow remote working (a capped subsidy of 50% of installation costs) or in protective equipment to allow continued working in enclosed spaces. These measures prevented the collapse of businesses most affected by the containment measures and behavioural responses. In addition, other subsidies aimed at preserving worker attachment, such as by extending the coverage of the Employment Adjustment Subsidy, potentially reducing the threat of scarring for many workers. One consequence is the very low rate of bankruptcy, which raises questions of whether crisis-period support could prop up unviable businesses. Empirical evidence suggests that firms performing more poorly before the pandemic were more likely to receive grants and concessional loans (Hoshi, Kawaguchi and Ueda, 2021[60]). While the difference in take-up may not be surprising, it does raise the possibility that support has kept some enterprises afloat.
Firms that have taken on loans may limp on particularly given the revealed reluctance of individuals to close firms or declare bankruptcy. Some sectors may face lingering difficulties due to changes in behaviour (such as fewer individuals using face-to-face services in the hospitality sector). In this case, a desire to extend support to prevent large-scale collapse of enterprises in a given sector could prolong support. This runs the risk of hindering resource reallocation to more productive enterprises. As a result, support measures should become more targeted as the recovery gains traction and move away from loan guarantees propping up ultimately unviable firms. Particularly for small enterprises, moving from debt to equity is difficult due to the large number of firms involved and their typical reliance on bank lending. The European Commission’s Temporary State Aid Framework from the start of 2021 allowed some loans up to a ceiling to be converted into grants in certain conditions.
Providing the skills for the digital transformation
In a highly digitised society, ICT equipment becomes an indispensable tool and ICT skills become almost as fundamental as reading or numeracy. Skills amongst the Japanese population are highly developed. The share of the population with a university degree is amongst the highest in the OECD and measures of adult literacy and numeracy are among the best in the OECD (Figure 2.19). Digital problem-solving amongst adults is comparatively good with 42% of adults possessing strong problem-solving skills, as against 32% on average OECD-wide. However, around one quarter of adults lack basic ICT skills, above the 19% OECD average. To some extent, this reflects that Japan is ageing earlier than other OECD countries. Younger adults are generally more literate and numerate with a larger share having completed tertiary education and having greater familiarity with digital technologies than previous cohorts, suggesting that some weaknesses will diminish over time. That said, digital skill proficiency of younger cohorts is only around the OECD average (Figure 2.19).
While education attainment is high, the changing landscape of jobs and tasks will require new skills to be developed. For example, the spread of automation threatens existing jobs or demands significant changes in how some jobs are performed. The number of such jobs is estimated to be relatively large (56%) in comparison with the rest of the OECD (47% on average) (Nedelkoska and Quintini, 2018[61]). In part, the elevated risk reflects the slower roll-out of digital technologies in some parts of the economy. Those most at risk are workers with lower levels of educational attainment, the young and low paid and those in part-time jobs and working in smaller enterprises.
The impact of digitalisation on workers is uncertain. The fear of robots displacing employment found in some studies (Acemoglu and Restrepo, 2020[62]) appears to be a lesser concern in Japan (Adachi, Kawaguchi and Saito, 2020[63]). In part, this may be due to the export orientation of major automobile manufacturers. Adoption of robots lowered costs and by boosting exports created more demand for labour. Another factor is that some evidence points to robots being complements to workers who possess fewer skills (temporary and part-time employees and the elderly) while raising productivity and average wages (Dekle, 2020[8]). Further automation in the context of an ageing and shrinking population is therefore not necessarily a threat to workers.
Against this background, ensuring the skills are available to make the most of the digital transformation will require action both for the inflow and the existing stock of workers. First, the share of graduates in STEM disciplines is modest, suggesting the skills needed to harness digital technologies fully are not widespread. Second, upgrading skills and reskilling of the existing workforce requires addressing institutional features that hinder adaptation to new challenges.
Figure 2.19. The share of university graduates is high, but digital skills are around average
Note: Share of adult working-age population with tertiary educational attainment. 'No digital skills' includes adults who have had no computer experience, failed the ICT core test or opted out of taking the test. 'Limited digital skills' includes adults scoring below or at level 1 of proficiency in problem solving in technology-rich environments.
Source: OECD Education at a Glance database.
The inflow of workers with relevant skills
The development of relevant skills depends on compulsory education. And education can also benefit from the digital transformation. Better integration of digital teaching technologies could provide individually-tailored education and reduce administrative burdens for teachers. The share of schools with sufficient ICT resources is limited, partly due to responsibility lying at the sub-central government level and differences across local governments in providing digital resources. For example, some prefectures, such as Saitama, have already developed initiatives to expand computer use. The provision of computers per student is well below the average in OECD countries (Figure 2.20). Japanese students have less opportunity to use computers in schools, and less time to do so. Furthermore, access appears to depend more on where the student lives and the school she goes to than on average across the OECD (Figure 2.21). Time spent on the internet is relatively low, both at and outside of school. At home, the relatively limited use was not due to a lack of access to the internet. On the other hand, relatively few Japanese students have access to a computer at home (OECD, 2021[64]) (Figure 2.22).
Figure 2.20. Computers are comparatively less available in schools
Number of computers per student as reported by the principal
Note: This analysis is restricted to schools with the modal ISCED level for 15-year-old students.
Source: OECD, PISA 2018 database, Table V.B1.5.6.
To lift basic skills, ICT literacy and promote creativity, the Japanese government has been changing the education curriculum to include programming courses, and launched the GIGA (Global and Innovation Gateway for All) School initiative in 2019. A survey carried out around that time suggested only 12% of schools had enough computers for all students (JAPET, 2020[65]). The GIGA project aims to build an ICT environment to create a learning environment that leverages the best of new technologies and traditional approaches to pedagogy. A large part of the resources being made available are devoted to making sure every student in compulsory education has access to a computer or tablet connected to high-speed broadband (around 13 million students in 35 thousand schools). At the same time, the Ministry of Education is mindful that opportunities provided by ICT should also be available to those who face difficulties in using these technologies and additional resources are available for students needing assistance.
While the GIGA School initiative started before the spread of COVID-19, the shock revealed the heterogeneity of digital ability across areas and schools. It also highlighted the importance of expanding computer availability - by inter alia facilitating distance learning and online classes - and the government decided to accelerate GIGA implementation. By the end of March 2021, nearly all school districts (96.5%) were providing one computer per student. Most students can now use ICT networks at school, including the internet.
The budgets in the fiscal years 2019 and 2020 provided around JPY 230 billion per year. Yet effective provision is still not complete. To enhance resilience against a background of pandemic risks, the digital infrastructure and software need to be improved rapidly so that all students can access online-communication and education, with the service tailored to the student’s own needs and abilities. More fundamentally, the associated investments to exploit the full functionality of these devices are insufficient. Teachers on average are the least prepared to use ICT resources in the OECD (Figure 2.23). Schools also report that they are among the least well equipped in the OECD to help teachers use digital devices, to give teachers adequate time to prepare lessons integrating digital technologies, to provide online learning support and to have enough qualified technical support staff (OECD, 2020[66]). In short, the laudable push to provide ICT equipment needs to be accompanied by efforts to ensure teachers can integrate them into their classrooms and teaching methods.
Figure 2.21. ICT access is quite dependent on school type and location
Access to the Internet and a computer that can be used for schoolwork at home, by student and school characteristics
Note: The socio-economic profile is measured by the PISA index of economic, social and cultural status (ESCS). Japan does not have separate data for rural areas or villages.
Source: OECD (2021[1]), PISA (2018).
Figure 2.22. Students in Japan spend less time on the Internet and using computers
Digital use of students at school and home
Note: Results based on students' self-reports. Ranges show highest and lowest of OECD countries.
Source: OECD (2021[1]), PISA (2018).
Figure 2.23. Schools are not well prepared for the use of ICT tools
Percentage of students in schools whose principal agreed or strongly agreed with the following statements:
Source: OECD calculations based on PISA 2018.
These deficiencies held back distance learning and online classes replacing in-person learning during the pandemic. At university level, however, most courses after an initial delay moved online. In addition, local governments, such as Kumamoto city, have been ensuring distance learning was possible for students given fears of natural disasters disrupting education (Kang, 2021[67]). This suggests that the obstacles are not insurmountable and complementing equipment availability with training and plans on usage will allow better integration of ICT into compulsory education.
At university level, there are also weaknesses in teaching graduates relevant skills. Notably, the share of graduates from STEM disciplines is around the OECD average (22% versus 23%) but just 7% of female graduates are from STEM disciplines (Figure 2.24). This is amongst the lowest in the OECD, despite Japanese female students on average outperforming male students in almost all OECD countries in the PISA mathematics tests. Indeed, just 2% of female students are graduating from natural sciences, mathematics and statistics, versus 5% on average in the OECD.
Redesigning university courses may help attract more potential students. In some cases, the perception was that the courses were too academic and divorced from practical application, contributing to the low take-up of STEM courses. Studies suggest that mentoring and exposure to women engaged in STEM research can help improve positive attitudes to STEM disciplines amongst female students (Kijima, Yang-Yoshihara and Maekawa, 2021[68]). Understanding the pronounced gender differences in enrolment and addressing the factors identified appears to be a prerequisite to ensuring that more women take up STEM disciplines. In addition, it is essential to ensure there is no discrimination preventing women taking STEM courses, such as requiring female students to obtain higher marks in university entrance examinations.
According to the OECD Priorities for Adult Learning dashboard, Japan performs poorly with respect to the alignment of education to labour market needs (OECD, 2019[69]). Some 89% of employers report difficulties in hiring workers, and close to 70% of workers report needing training to cope with tasks on their current job. Skill shortages are noted for legal workers, management, finance and insurance specialists, social welfare specialists and care service workers (OECD, 2021[70]). In addition, digital skills are in shorter supply and small and medium-sized enterprises face larger difficulties in hiring workers with these skills. Japan’s 2017 Growth Strategy began to address this with the focus on skills development (Government of Japan, 2017[71]).
Figure 2.24. There are few STEM graduates
Share of students graduating from STEM disciplines, % in 2019
Note: For Japan data for Information and Communications Technologies graduates are not given but allocated to different fields of study. The picture does not change substantively when omitting this field of study from the comparison.
Source: OECD Education at a Glance, 2020.
Other countries are also working to ensure the provision of needed skills. For example, the Converge Challenge programme run by Scottish universities with the involvement of businesses and other interested groups aims to use universities’ intellectual assets to develop the expertise being amassed by students into new high-growth businesses (OECD, 2020[52]). This programme provides a mix of training and competition to hone the potential entrepreneurs’ ability to commercialise their new product or service.
Training existing workers in new skills
The second approach to building up skills needed for the digital transformation is training the existing workforce. In part, this is needed due to the expected impact of digitalisation on the nature of the work. But another imperative as mentioned above is the potential for digitalisation to be disruptive and require workers to move occupations and retrain. In this regard, training of existing workers needs to occur both within business and outside.
The Survey of Adult Skills estimated that in the early 2010s, 35% of adults in Japan engaged in work-related training in a given year, somewhat below the 39% OECD average (OECD, 2021[70]). In comparison with other countries, the share of workers receiving firm-based training is also comparatively small (Figure 2.25). The relative weakness of training becomes important because workers tend to remain at their firm and job tenure is amongst the longest in the OECD. Indeed, job-to-job flows are muted in comparison with other OECD countries, although they have increased modestly over time (Figure 2.26). In France and Italy, for example, the estimated job-to-job flows as a share of total employment are around 8-10% per annum, roughly the rate of the much more mobile youth in Japan (Berson, De Philippis and Viviano, 2020[72]). The 2015 Survey of Job Changers also revealed that job change was much more likely in smaller businesses, especially in the wholesale and retail sectors, and amongst those at earlier stages of their careers. This constellation of labour market features, in the context of a declining labour force, puts a greater onus on training the existing work force rather than relying on hires to address skill needs. The government has recognised this and accompanying the 2018 Work Style reforms they promoted job change regardless of age. In addition, existing policy supports give vouchers to co-fund job training or professional training that the worker co-finances.
Figure 2.25. Firm-based training is comparatively underdeveloped
As a percentage of workers in each sector group, 2012 or 2015
Source: OECD (2019b), calculations based on OECD Survey of Adult Skills (PIAAC).
Training is also important in view of skill mismatches. These are relatively large with many workers having higher-level qualifications than needed for new entrants to their jobs. This is more noticeable for workers with post-secondary non-tertiary education. In addition, many are employed in jobs unrelated to their field of study. This is more common in small firms and when the worker is a non-regular worker as well as for workers who graduated from humanities, language and arts disciplines (OECD, 2021[70]). Furthermore, jobs may not make full use of workers’ literacy and numeracy abilities. This constellation of training needs means policy should strive to reach beyond workers on a regular contract.
For regular workers, the expectation of lifetime employment, seniority wages and intakes of new graduates rather than mid-career hires creates stronger incentives for firms and workers to invest in training contracts (Jones and Seitani, 2019[26]). The training budget per worker reflects this. Larger firms, which have more lifetime employment contracts, typically spend more on training relative to total employment costs (Figure 2.27). The training budget relative to compensation has at best stood still since the beginning of the century. A second feature is that over the longer term, the number of workers receiving training appears to have declined. Whereas in the early 1990s, over half of regular workers received training, this dropped to closer to 40% by the early 2010s (Hara, 2019[73]). The share of non-regular workers receiving training is even lower. According to the Basic Survey of Human Resource Development the share of firms offering training to non-regular workers is about one-half of the share offering training to regular workers.
Public support is available for worker training. The Human Resource Development Support Grant is a major way the government encourages firm-financed worker training. While it is available for both regular and non-regular workers, the firm needs to submit a plan for converting non-regular workers receiving this training into regular workers, which may undermine its attractiveness for some employers. A subsidy is available for firms granting workers education and training leave but it is limited to full-time employees who are contributing to the employment insurance system. However, take-up is relatively small (9% of employees) with the major barriers a lack of awareness about the different programmes and difficulties in finding replacement staff.
Figure 2.26. Job-to-job flows have trended up modestly
Percentage of workers changing jobs over the past year
Note: February or first quarter data.
Source: "The Special Survey of the Labour Force Survey" from 1984 to 2001, "Labour Force Survey (Detailed Tabulation)" since 2002.
A growing share of the workforce is non-regular employees (Figure 2.28), notwithstanding a small decline in 2020. This type of employment is predominantly occupied by women and increasingly by workers over the age of 60. The split between regular and non-regular employment for younger age groups has remained reasonably stable since the early 2000s, though younger males appear slightly less likely to be in regular employment than at the beginning of the century. The picture is more nuanced for women as the number in regular employment has risen in recent years, reflecting the impact of the labour market reforms aimed at increasing female labour force participation (OECD, 2019[24]).
The rising share of elderly workers on non-standard contracts arises in part due to a downside of the lifetime employment system and its associated seniority wage system. The impact of seniority on wages is amongst the strongest in the OECD. Ever rising wage costs have induced firms to dismiss (less-productive) elderly workers. Firms have used mandatory retirement schemes to achieve this, typically at the age of 60, which the government imposed to stop even earlier forced retirements. This policy is evolving and firms have since 2013 needed to ensure workers remained employed until age 65. From 2021, firms are required to make efforts (though not guarantee) that workers can remain in employment until age 70, although not necessarily within the firm itself. The public Japan Organisation for Employment of the Elderly, Persons with Disabilities and Job Seekers (JEED) supports these efforts by giving advice and providing grants to firms providing employment to workers wishing to continue working beyond the age of 65.
This system obliging firms to keep older workers employed has led to workers being switched to non-regular contracts, thereby allowing compensation to be cut. In order to prevent sharp income shocks for the elderly, some firms have been moving to task-based pay, which would reduce the incentives seniority wages give to fire older workers. Subsidies are available for firms creating such a pay schedule for their employees. The share of firms using merit-based pay schemes has been gradually rising over time. In addition, the 2018 Work Style reforms aimed to reduce the differences between regular and non-regular workers, as well as capping permissible overtime hours. The government also issued guidelines targeting labour market flexibility and creating conditions that would support job changes at all ages. Reforms in 2020 established quotas of mid-career worker hires for large firms (rather than relying almost exclusively on graduate hires).
Figure 2.27. Firm spending on training is more generous in larger firms
Training costs as a percentage of total labour costs by firm size
Note: The survey only includes workers in firms with over 30 employees. Before 2016, non-corporate organizations and multi-service corporations were not included.
Source: MHLW Basic Survey of Human Resource Development.
The elderly often being on non-regular contracts are also less likely to receive training. However, training older workers appears to have a positive effect. Some empirical evidence suggests that training either just before or just after retirement not only increases re-employment chances but also the probability of being re-employed with a regular contract. However, only around 14% of full-time employees regain regular full-time employment after compulsory retirement (Sato, 2017[74]). The majority of workers move into non-regular employment, mainly part-time employment. The transition to non-regular employment is also often accompanied by a change in occupation or sector, entailing a loss in firm or sector-specific human capital.
With seniority wages still widespread, labour market fluidity relatively underdeveloped for mid-career hires, and ongoing pressures from an ageing workforce, training needs to be addressed to an increasingly elderly workforce, many of whom are often on non-regular contracts. However, firms, particularly small ones, appear less likely to finance continuing education or training for such workers. A larger share of workers in non-standard employment increases the problems in retraining and upskilling. Some evidence suggests that non-regular workers are less likely to engage in training that augments their earning capacity than the training undertaken by regular workers (Yokoyama, Kodama and Higuchi, 2019[75]). Against this background, the training available needs to be aligned with the skills demanded by firms and workers need to be sufficiently informed about the potential outcomes of their training choices.
The other workers most at risk of lack of training are those in small enterprises. These firms find it less easy to juggle a worker’s absence and may face bigger financial hurdles in providing training. Recognising this, the Human Resource Development Support Grant is more generous for small firms. Government also provides subsidies for SMEs that establish certified training centres. The Korean approach of involving large enterprises, employer associations and universities in training centres may exploit scale efficiencies and better forge linkages between smaller enterprises and universities (OECD, 2020[76]).
At present, public institutions play a limited direct role in worker training (around 5% of training participants for workers or jobseekers (OECD, 2021[70]), as policy relies more on supporting firm-based training either within the firm or by external public or private providers. This approach tends to prioritise firm-specific skills and may not be suitable to meet the needs of workers more exposed to reskilling needs or moving to new employment induced by the digital transformation. The polytechnic centres in each prefecture are the main providers of adult training. They tend to provide technical training related to manufacturing, which does not lead to a degree. The current 46 polytechnics were established by the aforementioned JEED. As seen after the Tohoku earthquake in 2011, when instructors travelled to the affected regions to give needed training, these institutions can be very responsive to changing demands. Provision of adult learning in universities has been relatively limited but more recently the government has begun to promote closer linkages between universities and the business sector. For example, the “Brush Up Program for Professionals” certifies qualifying adult education courses at universities and other higher education institutes. In 2019, over one third of these courses were eligible for benefits (OECD, 2021[70]).
Figure 2.28. Non-regular employment has trended up through the late 2010s
Source: Statistics Bureau of Japan, Labour Force Survey statistics.
The digital transformation is creating employment opportunities and demands for skills that were unforeseen even recently, but also offers new opportunities for providing training. For example, distance learning has had a small footprint in training. Before the pandemic, only around a quarter of universities offered distance-learning options. Distance learning promises more flexibility in training provision, which may make it easier for firms and workers to participate. Initiatives are already underway to disseminate training through massive open online education.
However, for workers with less ability to undertake distance learning, provision of face-to-face training will need to be retained to prevent low-wage individuals from falling behind. Efforts are also needed to expand the reach of adult education. As noted above, training is currently less available to non-regular workers and those in smaller enterprises. In addition, those with lower levels of educational attainment and women are less likely to participate in adult learning, reflecting that they tend to be employed in jobs less likely to involve training. Attempts to expand coverage need to bear in mind the costs and programmes should embed evaluation into the design. As such, if trial projects reveal strengths or weaknesses, the programme can then be scaled up or down, or adjusted.
Digitalisation can also provide workers and jobseekers better information on the training available. Such information is available but scattered across different government portals, including the site encouraging adult learning (Manapass), the Ministry of Health, Welfare and Labour site listing courses eligible for training benefits, the public employment service’s website for jobseekers, the Job card website for vocational training and an occupational information website. Bringing this information together would help workers and firms navigate training opportunities.
While greater provision of training opportunities offers part of the solution to training needs, the skills acquired also need to be relevant. The alignment of education and business needs is complicated beyond foundational skills because education systems need to forecast skill demands given the training lag (both of the instructors and then students/learners). The current system of curriculum design gives a role to the Ministry of Education, Culture, Sports, Science and Technology through the accreditation of educational institutions that offer suitable training for business professionals to brush up their knowledge. Besides pedagogical aspects of training to qualify for this certification, one requirement is to incorporate systematically the opinions of businesses in the courses offered, which should ensure coherence with local labour market requirements. The Ministry of Health, labour and Welfare is also active in the certification of private vocational education providers. However, the number of institutions certified has been modest, partly driven by the cost of the certification process. Other OECD countries (Austria, France, Korea, Switzerland) provide stronger incentives for certification by limiting the receipt of public funds to these institutions or by ultimately making it mandatory (Slovenia).
Beyond the standard curricula development and skill assessments for the provision of public vocational training, led by the Central Training Council, specific projects have been put in place to support the digital transformation, particularly in the development of IT skills. Training programmes now cover artificial intelligence, the internet of things, cloud computing and data science more generally. Indeed, the Artificial Intelligence Strategy from 2019 supports training to increase the numbers of skilled workers in innovative industries (Integrated Innovation Strategy Promotion Council Decision, 2019[77]). The strategy also recognises the importance of disseminating best practice, which will help different actors react to a fast-changing digital environment.
References
[62] Acemoglu, D. and P. Restrepo (2020), “Robots and jobs: Evidence from us labor markets”, Journal of Political Economy, Vol. 128/6, pp. 2188-2244, http://dx.doi.org/10.1086/705716.
[63] Adachi, D., D. Kawaguchi and Y. Saito (2020), Robots and Employment: Evidence from Japan, 1978-2017, REITI Discussion Paper 20-E-051, https://www.rieti.go.jp/en/ (accessed on 2 June 2021).
[54] Adalet McGowan, Muge, D. Andrews and V. Millot (2017), The Walking Dead?: Zombie Firms and Productivity Performance in OECD Countries, OECD Economics Department Working Papers No. 1372, http://www.oecd.org/eco/workingpapers. (accessed on 1 June 2021).
[56] Adalet McGowan, M., D. Andrews and V. Millot (2017), Insolvency Regimes, Zombie Firms and Capital Reallocation, OECD Economics Department Working Papers, No. 1399, http://www.oecd.org/eco/workingpapers. (accessed on 11 June 2021).
[49] Andres, R. et al. (2020), “Capital incentive policies in the age of cloud computing: An empirical case study”, OECD Science, Technology and Industry Working Papers, No. 2020/07, OECD Publishing, Paris, https://dx.doi.org/10.1787/4bedeb36-en.
[41] Appelt. S., M. et al. (2020), The effects of R&D tax incentives and their role in the innovation policy: mix: Findings from the OECD, OECD Science, Technology and Industry Policy Papers, No. 92.
[37] Bahar, D. and S. Strauss (2020), Innovation and the Transatlantic Productivity Slowdown, Brookings Institution, Washington D.C.
[28] Baily, M., B. Bosworth and S. Doshi (2020), “Lessons from Productivity Comparisons of Germany, Japan and the United States”, International Productivity Monitor, pp. 81-103, http://www.csls.ca/ipm/38/Baily_Bosworth_Doshi.pdf (accessed on 27 May 2021).
[58] Baker, J. et al. (2020), Joint Response to the House Judiciary Committee on the State of Antitrust Law and Implications for Protecting Competition in Digital Markets, Congressional and Other Testimony. 18., https://ssrn.com/abstract=3239248. (accessed on 25 June 2021).
[78] Banerjee, R. and B. Hofmann (2020), Corporate zombies: Anatomy and life cycle, http://www.bis.org (accessed on 1 June 2021).
[80] Bank of Japan (2020), The Bank of Japan’s Approach to Central Bank Digital Currency, Bank of Japan, https://www.boj.or.jp/en/announcements/release_2020/data/rel201009e1.pdf (accessed on 15 June 2021).
[23] Berlingier, G. et al. (2020), Lagging Firms, Technology Diffusion and Its Structural and Policy Determinants, OECD Science, Technology and Industry Policy Papers, No. 86, https://www.oecd-ilibrary.org/docserver/281bd7a9-en.pdf?expires=1625840544&id=id&accname=ocid84004878&checksum=26F9C6CF140192EA677F1AC5C6F2515D (accessed on 9 July 2021).
[72] Berson, C., M. De Philippis and E. Viviano (2020), Job-to-Job Flows and Wage Dynamics in France and Italy, Banque de France Working Paper Series no. 756.
[81] Boar, C. and A. Wehrli (2021), Ready, steady, go? – Results of the third BIS survey on central bank digital currency, BISPapers No 114, http://www.bis.org (accessed on 14 June 2021).
[10] Cabinet Secretariat and Ministry of Internal Affairs and Communications (2021), 行政手続等の棚卸結果等（令和２年度調査）( Inventory results of administrative procedures, etc. (Reiwa 2nd year survey) ), Cabinet Secretariat and Ministryof Internal Affairs and Communications.
[35] Chun, H., J. Hur and N. Son (2021), “Global Value Chains and Servicification of Manufacturing: Evidence from Firm-level Data”, Japan and the World Economy, Vol. 58, p. 101074, http://dx.doi.org/10.1016/j.japwor.2021.101074.
[8] Dekle, R. (2020), “Robots and industrial labor: Evidence from Japan”, Journal of the Japanese and International Economies, Vol. 58, p. 101108, http://dx.doi.org/10.1016/j.jjie.2020.101108.
[18] Eggleston, K., Y. Lee and T. Iizuka (2021), Robots and Labor in the Service Sector: Evidence from Nursing Homes, National Bureau of Economic Research, Cambridge, MA, http://dx.doi.org/10.3386/w28322.
[4] Esselink, H. and L. Hernández (2017), The Use of Cash by Households in the Euro Area, Banque de France, http://dx.doi.org/10.2866/377081.
[82] European Central Bank and Bank of Japan (2020), Balancing Confidentiality and Auditability in a Distributed Ledger Environment, STELLA – joint research project of the European Central Bank and the Bank of Japan, https://www.boj.or.jp/announcements/release_2020/data/rel200212a1.pdf (accessed on 15 June 2021).
[36] Faquet, R. and V. Malardé (2020), Numérisation des entreprises françaises, Treso-Economics.
[31] Fukao, K. et al. (2015), “Why Was Japan Left Behind in the ICT Revolution?”, Telecommunications Policy, Vol. 40/5, pp. 432-449, http://www.rieti.go.jp/en/ (accessed on 1 June 2021).
[46] Fukugawa, N. (2012), “University Spillovers into Small Technology-based Firms: Channel, Mechanism, and Geography”, The Journal of Technology Transfer, Vol. 38, pp. 415-431, http://dx.doi.org/10.1007/s10961-012-9247-x.
[29] Gal, P. et al. (2019), “Digitalization and Productivity: In Search of the Holy Grail - Firm Level Empirical Evidence from European Countries”, International Productivity Monitor, http://www.csls.ca/ipm/37/OECD.pdf (accessed on 31 May 2021).
[27] Goldin, I. et al. (2021), Why is Productivity Slowing Down?, Oxford Martin Working Paper Series on Economic and Technological Change, https://www.oxfordmartin.ox.ac.uk/future-of-work/ (accessed on 28 May 2021).
[71] Government of Japan (2017), 2017 Growth Strategy Japan.
[22] Haidar, J. and T. Hoshi (2015), Implementing Structural Reforms in Abenomics: How to Reduce the Cost of Doing Business in Japan, National Bureau of Economic Research, Cambridge, MA, http://dx.doi.org/10.3386/w21507.
[40] Hall, B. and J. Van Reenen (2000), “How Effective are Fiscal Incentives for R & D? A Review of the Evidence”, Research Policy, Vol. 29, pp. 449-469, http://www.elsevier.nlrlocatereconbase (accessed on 8 July 2021).
[73] Hara, H. (2019), “The Impact of Worker-financed Ttraining: Evidence from Early- and Mid-career Workers in Japan”, Journal of the Japanese and International Economies, Vol. 51, pp. 64-75, http://dx.doi.org/10.1016/j.jjie.2018.11.001.
[60] Hoshi, T., D. Kawaguchi and K. Ueda (2021), The Return of the Dead? The COVID-19 Business Support Programs in Japan., CARF, https://www.carf.e.u-tokyo.ac.jp/research/ (accessed on 1 June 2021).
[33] Hosono, K., M. Takizawa and K. Yamanouchi (2020), Firm Age, Productivity, and Intangible Capital, REITI Discussion Papers, https://www.rieti.go.jp/en/ (accessed on 31 May 2021).
[7] IFR (2021), Robot Race: The World´s Top 10 automated countries, International Federation of Robotics, https://ifr.org/ifr-press-releases/news/robot-race-the-worlds-top-10-automated-countries (accessed on 18 June 2021).
[55] Ikeuchi, K. et al. (2020), “Productivity Dynamics in Japan and the Negative Exit Effect”, Service Sector Productivity in Japan Discussion Paper Series DP20-001, Hitotsubashi University, http://sspj.ier.hit-u.ac.jp/ (accessed on 1 June 2021).
[25] IMD (2020), IMD World Digital Competitiveness Ranking 2020, IMD, https://www.imd.org/wcc/world-competitiveness-center-rankings/world-digital-competitiveness-rankings-2020/ (accessed on 3 June 2021).
[77] Integrated Innovation Strategy Promotion Council Decision (2019), AI Strategy 2019: AI for Everyone: People, Industries, Regions and Government.
[65] JAPET (2020), 第 12 回 教育用コンピュータ等に関するアンケート調査 報告書 令和 2 年 6 月 (12th Questionnaire Survey on Educational Computers, etc. Report Reiwa June 2), JAPET.
[26] Jones, R. and H. Seitani (2019), Labour Market Reform in Japan to Cope with a Shrinking and Ageing Population, OECD Economics Department Working Papers, No. 1568, http://www.oecd.org/economy/japan-economic-snapshot/ (accessed on 7 June 2021).
[67] Kang, B. (2021), “How the COVID-19 Pandemic Is Reshaping the Education Service”, in The Future of Service Post-COVID-19 Pandemic, Volume 1, Nature Publishing Group, http://dx.doi.org/10.1007/978-981-33-4126-5_2.
[68] Kijima, R., M. Yang-Yoshihara and M. Maekawa (2021), “Using Design Thinking to Cultivate the Next Generation of Female STEAM Thinkers”, International Journal of STEM Education, http://dx.doi.org/10.1186/s40594-021-00271-6.
[42] Kobayashi, Y. (2014), “Effect of R&D Tax Credits for SMEs in Japan: A Microeconometric Analysis Focused on Liquidity Constraints”, Small Business Economics, Vol. 42/2, https://www.jstor.org/stable/pdf/43552930.pdf (accessed on 8 July 2021).
[5] Kumur, R. and S. O’Brien (2019), Cash | 2019 Findings from the Diary of Consumer Payment Choice, Federal Reserve Bank of San Francisco, https://www.frbsf.org/cash/publications/fed-notes/2019/june/2019-findings-from-the-diary-of-consumer-payment-choice/ (accessed on 28 June 2021).
[48] METI (2018), Report on Digital Transformation (DX): Overcoming of ’2025 Digital Cliff’ Involving IT Systems and Full-fledged Development of Efforts for Digital Transformation, METI.
[19] MHLW (2019), 医療・福祉サービス改革プラン (Medical and welfare service reform plan), Ministry of Health, Labour and Welfare, https://www.mhlw.go.jp/content/12601000/000513536.pdf (accessed on 28 June 2021).
[34] Miroudot, S. and C. Cadestin (2017), “Services in Global Value Chains: Trade patterns and gains from specialisation”, OECD Trade Policy Papers, No. 208, OECD Publishing, Paris, https://dx.doi.org/10.1787/06420077-en.
[11] MLIT (2020), i-Constructionによる建設現場の生産性向上.
[14] MLIT (2020), 世界初！ 自動運転車（レベル３）の型式指定を行いました (World first! The model of the self-driving car (level 3) has been specified.), Ministry of Land, Infrastructure, Transport and Tourism.
[1] Morikawa, M. (2021), The Productivity of Working from Home: Evidence from Japan, RIETI, https://voxeu.org/article/productivity-working-home-evidence-japan (accessed on 21 June 2021).
[21] Morikawa, M. (2020), Evidence-based Regulatory Reform, RIETI, https://www.rieti.go.jp/en/columns/s20_0001.html (accessed on 25 June 2021).
[53] Morikawa, M. (2013), “Productivity and Survival of Family Firms in Japan”, Journal of Economics and Business, Vol. 70, pp. 111-125, http://dx.doi.org/10.1016/j.jeconbus.2012.11.001.
[61] Nedelkoska, L. and G. Quintini (2018), “Automation, skills use and training”, OECD Social, Employment and Migration Working Papers, No. 202, OECD Publishing, Paris, https://dx.doi.org/10.1787/2e2f4eea-en.
[64] OECD (2021), 21st-Century Readers: Developing Literacy Skills in a Digital World, PISA, OECD Publishing, Paris, https://dx.doi.org/10.1787/a83d84cb-en.
[70] OECD (2021), Creating Responsive Adult Learning Opportunities in Japan, Getting Skills Right, OECD Publishing, Paris, https://dx.doi.org/10.1787/cfe1ccd2-en.
[17] OECD (2021), Good Practice Principles for Data Ethics in the Public Sector, OECD Publishing, Paris, https://www.oecd.org/gov/digital-government/good-practice-principles-for-data-ethics-in-the-public-sector.htm.
[51] OECD (2021), The Digital Transformation of SMEs, OECD Studies on SMEs and Entrepreneurship, OECD Publishing, Paris, https://dx.doi.org/10.1787/bdb9256a-en.
[9] OECD (2020), Digital Government in Chile – Improving Public Service Design and Delivery, OECD Digital Government Studies, OECD Publishing, Paris, https://dx.doi.org/10.1787/b94582e8-en.
[20] OECD (2020), Digital Government Index (DGI): 2019 Results, OECD Publishing, Paris, https://doi.org/10.1787/14e1c5e8-en-fr.
[79] OECD (2020), Highlights from OECD Innovation Indicators 2019 How do firms innovate across the world?, OECD Paris, http://oe.cd/inno-stats, (accessed on 11 May 2021).
[45] OECD (2020), Highlights from OECD Innovation Indicators 2019 How Do Firms Innovate across the World?, OECD Publishing, Paris, http://oe.cd/inno-stats, (accessed on 11 May 2021).
[52] OECD (2020), International Compendium of Entrepreneurship Policies, OECD Publishing, Paris, https://dx.doi.org/10.1787/338f1873-en.
[30] OECD (2020), Japan: Business Dynamics, OECD Publishing, Paris, https://www.oecd.org/sti/ind/oecd-business-dynamics-insights-japan.pdf (accessed on 25 June 2021).
[44] OECD (2020), “Mapping Business Innovation Support (MABIS) Deliverable 1.1: R&D tax incentives reporting (Year 1) 1 Work package 1. Information and indicators of tax relief for business R&D expenditures OECD R&D tax incentives database, 2020 edition”, pp. OECD Publishng, Paris.
[6] OECD (2020), OECD Digital Economy Outlook 2020, OECD Publishing , Paris, https://www.oecd-ilibrary.org/docserver/bb167041-en.pdf?expires=1622449438&id=id&accname=ocid84004878&checksum=27F59F4C3E13AB6F41C91C3FEC33FBC4 (accessed on 31 May 2021).
[76] OECD (2020), OECD Economic Surveys Korea, OECD Publishing, Paris.
[66] OECD (2020), PISA 2018 Results (Volume V): Effective Policies, Successful Schools, PISA, OECD Publishing, Paris, https://dx.doi.org/10.1787/ca768d40-en.
[43] OECD (2020), R&D Tax Incentives: Japan, 2020 Design of R&D tax relief provisions, OECD Publishing, Paris, http://www.oecd.org/stiFormoreinformation,pleasevisit:http://oe.cd/rdtax (accessed on 8 July 2021).
[59] OECD (2020), The Concept of Potential Competition, OECD Publishing, Paris.
[15] OECD (2019), Health in the 21st Century: Putting Data to Work for Stronger Health Systems, OECD Health Policy Studies, OECD Publishing, Paris, https://dx.doi.org/10.1787/e3b23f8e-en.
[69] OECD (2019), How Future-ready is Japan’s Adult Learning System?, OECD Publishing, Paris, http://www.oecd.org/employment/skills-and-work/adult-learning (accessed on 9 June 2021).
[47] OECD (2019), ICT Investments in OECD and Partner Countries: Trends, Policies and Evaluation, OECD Publishing, Paris.
[24] OECD (2019), OECD Economic Surveys: Japan 2019, OECD Publishing, Paris, https://dx.doi.org/10.1787/fd63f374-en.
[3] OECD (2019), PISA 2018 Results (Volume I): What Students Know and Can Do, PISA, OECD Publishing, Paris, https://dx.doi.org/10.1787/5f07c754-en.
[2] OECD (2019), Skills Matter: Additional Results from the Survey of Adult Skills, OECD Skills Studies, OECD Publishing, Paris, https://dx.doi.org/10.1787/1f029d8f-en.
[16] OECD (2019), The Path to Becoming a Data-Driven Public Sector, OECD Digital Government Studies, OECD Publishing, Paris, https://dx.doi.org/10.1787/059814a7-en.
[57] OECD (2017), OECD Economic Surveys: Japan, OECD Publishing, Paris, https://www.oecd-ilibrary.org/docserver/eco_surveys-jpn-2017-en.pdf?expires=1624371312&id=id&accname=ocid84004878&checksum=A9EADDCD17B6BA27480E91AEF2374C67 (accessed on 22 June 2021).
[13] OECD (2016), OECD Territorial Reviews: Japan: 2016, OECD Publishing, Paris.
[12] Productivity Commission (2020), National Transport Regulatory Reform - Inquiry Report no. 94, Productivity Commission, http://www.pc.gov.au (accessed on 15 June 2021).
[74] Sato, K. (2017), The Effect of Training on the Employment of Older Workers after Compulsory Retirement in Japan, Panel Data Research Center at Keio University Discussion Paper.
[50] Serrano Calle, S., J. Pérez Martínez and Z. Frias Barroso (2016), “Spanish Public Policies towards the Promotion of Cloud Computing and Digital Services for SMEs”, Paper prepared for the 27th regional International Telecommunications Confrence, https://www.econstor.eu/bitstream/10419/148703/1/Serrano-et-al.pdf (accessed on 30 July 2021).
[39] Shapiro, C. and M. Lemley (2019), “The Role of Antitrust in Preventing Patent Holdup”, University of Pennsylvania Law Review, Vol. 168, https://perma.cc/Y97Y-KNM5]. (accessed on 22 July 2021).
[32] Tambe, P. et al. (2021), Digital Capital and Superstar Firms, Brookings, https://www.brookings.edu/research/digital-capital- (accessed on 31 May 2021).
[38] van Zimmeren, E. (2009), Patent Thickets and Refusals to License in the Life Sciences in Japan-Legal Remedies at the Interface between Patent and Competition Law, Institute of Intellectual Property, Tokyo.
[75] Yokoyama, I., N. Kodama and Y. Higuchi (2019), “Effects of State-sponsored Human Capital Investment on the Selection of Training Type”, Japan and the World Economy, Vol. 49, pp. 40-49, http://dx.doi.org/10.1016/j.japwor.2018.07.003.