Chapter 1. Japan’s demographic transition and the productivity challenge1

This chapter provides an overview of regional trends and challenges facing Japan. It begins with a look at the demographic and macroeconomic context, reflecting in particular on the implications of population ageing and decline for economic policy, especially efforts to enhance productivity. It then turns to an analysis of regional growth performance, inter-regional disparities and other socio-economic outcomes. This is followed by consideration of the economic policy response to demographic change at different territorial scales, from the national to the local, and an examination of the spatial consequences of Japan’s restructuring challenge. The primary focus of the discussion is on the interactions among demographic change, productivity performance and the evolution of the country’s settlement pattern.



Japan’s future prosperity depends on its success in addressing two enormous and inter-related challenges, which have tremendous implications for the evolution of its economic geography. The first is a process of demographic change that is without precedent in human history in terms of its speed and scale: Japan’s population is both ageing and shrinking at an extraordinary rate. The second challenge concerns productivity. After more than two decades of lacklustre economic performance, Japan is in the midst of an ambitious attempt to revive growth, involving changes to monetary, fiscal and structural policies. The link between these two challenges is clear. Demographic change is making productivity performance all the more important in view of the rapidly rising old-age dependency ratio. At present, the working-age population is falling by about 1% per year, and the rate of shrinkage will eventually approach 1.7% per year, so that even productivity growth of 2% or more will deliver very low aggregate or per capita growth. There will simply be no way to sustain high living standards and quality public services in a “super-ageing” Japan unless the country is able to achieve much higher rates of productivity growth than it has experienced in recent decades. Given the public service-delivery challenges that demographic change entails, the productivity of services, in particular, is a central concern.

Japan is not alone in facing both population ageing and population decline. OECD population projections suggest that at least 6 OECD countries will have populations more than 10% below their peak by mid-century and 12 will have elderly dependency ratios2 in excess of 50%. These processes are further advanced in Japan and they are unfolding faster there than elsewhere, so Japan’s response to demographic change is of great relevance to many other OECD members.

The spatial implications of these two challenges are the focus of this review. The dynamics of fertility, ageing and population decline differ greatly between rural and urban areas, as well across cities of different sizes. Most areas face population decline and some are threatened with depopulation, and the pace of ageing is often fastest in those experiencing the sharpest population decline. The productivity challenges are similarly diverse. In some places, the key will be unlocking economies of agglomeration in order to re-establish the dynamism of Japan’s major cities as engines of growth and innovation. However, agglomeration is far from the only basis for growth, and many of Japan’s smaller towns and rural areas will need strategies to ensure their own prosperity and viability in the decades to come. Service delivery challenges, in particular, will vary according to ageing and density trends.

This chapter examines the broad outlines of Japan’s demographic and productivity performance beginning in each case with the national context and then focusing on the spatial aspects of the problem – changing settlement patterns and regional economic performance. It then considers the broad outlines of the policy response to these twin challenges. The chapters that follow then explore the specific policy issues that arise in light of this analysis. Chapter 2 looks at national-level spatial planning and territorial governance, particularly the government’s emerging strategy to address these problems within the framework of the so-called “Grand Design for National Spatial Development 2050” and the new National Spatial Strategy. Chapter 3 focuses on the major metropolitan areas and Chapter 4 on the revitalisation of smaller towns and cities and of rural areas.

The demographic context

Japan’s population is both shrinking and ageing very rapidly

Japan’s total population rose rapidly during much of the 20th century, roughly tripling between 1900 and 2000, when it reached 126.9 million. By the end of the century, however, population growth had slowed dramatically, and the population peaked in 2010 at just over 128 million before beginning what is projected to be a sustained and increasingly steep decline (Figure 1.1). At the same time, the population that remains is ageing very rapidly, with the share of the population aged 65+ rising from under 5% in 1950, when the OECD average was about 7.7%, to around 26% in 2014, the highest such figure in the world.3 The share of over 80s in the population has risen even faster, from just 0.44% in 1950 to 7.3% in 2013, well above the 4.1% figure for the OECD area in 2011. By 2013, Japan’s median age had reached 45.9 years, against a world average of 29 (Cabinet Secretariat, 2015) and an OECD median estimated at 38.7 for that year.

Figure 1.1. Japanese population and age structure, 1950-2050

Source: OECD (2015a), “Historical Population Data and Projections (1950-2050)”, (accessed 16 October 2015).

As will be seen, both these processes – population ageing and population decline – are unfolding across the whole of Japan, but both the rates of change and the impacts vary greatly from place to place, resulting in significant changes to both labour markets and the settlement pattern. On current projections, the government anticipates that Japan’s population will fall by about 23-24% between 2010 and 2050, with the elderly (age 65+) share of the population standing at roughly 40% at the end of the period. The Ministry of Land, Infrastructure, Transport and Tourism (MLIT) projects that more than 60% of the inhabited grid squares in the country will lose more than half their population by 2050; 19% are expected to become uninhabited. By contrast, only 2% are projected to experience population growth (Ministry of Land, Infrastructure, Transport and Tourism, 2014). If fertility does not increase, the projections point to a sharper decline, to perhaps 86 million in 2060 and 43 million early in the next century.4

Population decline and population ageing in Japan reflect the interaction of two factors: fertility and longevity (Figure 1.2).

  • The main factor underlying Japan’s rapid demographic decline is low fertility. The total fertility ratio (TFR) in Japan fell sharply in the early 1950s, from a peak of almost 4.5 during the post-war baby boom. From the mid-1950s until about 1974, it fluctuated near or just above replacement rates (2.00-2.16), but after that it began a sustained decline, dropping to 1.26 in 2005 before recovering somewhat to 1.42 in 2014. Kono (2011) observes that, even if fertility returned to replacement levels immediately, the population decline would continue for a further 50-60 years before stabilising at around 110 million.

  • The collapse in fertility has coincided with an extraordinary increase in longevity. Life expectancy at birth rose from just under 51.68 years in 1947 to 83.48 years in 2013 and is now the highest in the OECD for both men and women. Such a rapid increase in life expectancy enabled Japan to sustain overall population growth long after the birth rate had fallen off. Though an extraordinary achievement in terms of human welfare, such longevity, when set alongside low fertility, implies a rapidly rising old-age dependency ratio.

Figure 1.2. Longevity and fertility in Japan
Life expectancy at birth and the total fertility ratio

Source: OECD (2015b), “Population”,, (accessed 16 October 2015); information provided directly by the Ministry of Health, Labour and Welfare (MHLW).

In principle, migration constitutes the third major driver of population dynamics in a country. In Japan, however, its impact has been negligible. Both emigration from and immigration to Japan have been very low since 1945.5 Japan’s foreign-born population was just under 1.7% of the total in 2010, far below the levels found in most OECD countries (Figure 1.3). Although it experienced labour shortages during the high-growth decades of the late 20th century, Japan chose not to rely on foreign labour. Instead, it pursued automated production, buttressed by specific forms of subcontracting and by employment systems that emphasised multi-skilling and teamwork rather than narrow specialisation (Fujimoto, 2013a, 2013b).

Figure 1.3. Share of foreign-born residents in total population, 2010

Source: OECD (2015b), “Population”,, (accessed 16 October 2015).

Rapid population ageing is common across the OECD (OECD, 2015c) – the elderly dependency ratio has risen in all OECD countries over recent decades – but it is more advanced in Japan than elsewhere (Figure 1.4), owing to ultra-low fertility over an extended period, exceptional longevity and much more restrictive approaches to immigration. Where Japan is now leading, many others will follow (Figure 1.5). Many countries are also experiencing depressed fertility levels and overall population decline. The drop in fertility in Japan has been far more pronounced than in many western countries, but it is not unique in East Asia: the People’s Republic of China, Korea and Chinese Taipei and have all experienced sharp declines from very high to very low fertility, over even shorter periods (Kono, 2011).6 Delayed marriage, falling infant mortality, and anti-natalist policies in some countries have all played a role, as has rising female labour force participation in the absence of the institutions, policies and social conventions needed to support women combining careers and child-rearing. Policy has been slow to respond in part because pro-natalist policies have traditionally been all but unknown in East Asia, which was long seen as “a region of limited land, scarce natural resources and overpopulation” (Kono, 2011:42).

Figure 1.4. Elderly population and total population in OECD countries, 1995 and 2012, %

Source: OECD (2015d), Regional Statistics (database), (accessed 10 September 2015).

Figure 1.5. Population estimates and projections by age group, 1950-2100

Source: OECD (2015c), Ageing in Cities, OECD Publishing, Paris,; OECD calculations based on United Nations Department of Economic and Social Affairs, Population Division (2010), World Population Prospects: The 2010 Revision, United Nations, New York, (accessed 30 September 2014).

Japan’s population is on the whole highly concentrated…

With 127.3 million inhabitants in 2013, Japan is the second most populous country in the OECD; in terms of surface area, it ranks ninth. This makes it the third most densely populated OECD member, with 341 inhabitants per square kilometre, more than double the OECD average. Most Japanese live in places that are far denser than the average, though. Almost 58% of Japanese live in prefectures with population densities in excess of 500 persons per square kilometre; nearly 45% are in prefectures where densities exceed 1 000 persons per square kilometre – comparable to the densities found in urban areas like Seattle, Rotterdam or Bristol. Five of the 20 densest functional urban areas (FUAs) in the OECD are in Japan. Altogether, just over half the population live in FUAs that (even with rural hinterlands included) occupy about 6% of the national territory.

Not surprisingly, therefore, Japan, ranks high on indicators of concentration of population; it was eighth on the OECD’s population concentration index in 2010 (OECD, 2014).7 In fact, this probably understates the case, because the index is computed at the level of TL3 regions, without reference to where those regions are in a country. In OECD countries with higher levels of concentration, like Australia and Canada, the largest centres of population tend to be distant from one another and spread across the national territory. In Japan, by contrast, the largest concentrations are clustered in the centre of the main island of Honshū. Among the larger OECD countries, probably only Korea and the United Kingdom might be comparable. The large conurbations centred on Tokyo, Osaka and Nagoya are home to almost half of Japan’s population, though they occupy only 5.2% of the national territory and occupy a stretch along the Pacific side of the island that is only about 500 km long.

…and the concentration process continues

The data for 1970-2010 show how Japan’s population has continued to concentrate on the Pacific side of the island of Honshū (Figure 1.6). The population growth has been fastest around Tokyo and, to a lesser extent, around Osaka, Nagoya and Kyoto. A number of prefectures are less populous than they were in 1970. This has resulted in growing overall concentration of population. Since the largest cities tend to have lower birth rates, this has been a result of migration rather than natural increase: the prefectural-level data on internal migration and population densities for the period from 1990 to 2012 show that people were moving from less dense to more dense places over the period and that this trend was significantly more pronounced after 2000 than before (Figure 1.7).8 A far larger proportion of Japanese regions are now experiencing net outflows than is the case in other large OECD countries, and only a handful (slightly more than 20%) have been experiencing net inflows. The gross flows are not dramatic, though. In any given year, only about 1.8-2.0% of Japanese move between prefectures, which is close to the OECD average for movement among TL3 regions. However, in context of demographic decline, such shifts are keenly felt in sending regions, particularly given that it is younger people who are most likely to move.

Figure 1.6. Population change by prefecture, average annual growth, 1970-2010

Note: This map is for illustrative purposes and is without prejudice to the status of or sovereignty over any territory covered by this map.

Source: OECD (2015d), Regional Statistics (database), (accessed 10 September 2015).

Figure 1.7. Initial population density and population change, 1990-2012
Japanese TL3 regions (prefectures)

Source: OECD (2015d), Regional Statistics (database),, (accessed 16 October 2015).

…yet the rural population is unusually remote

High concentration of population entails both costs and benefits, but it is complicated in Japan’s case by the fact that, largely owing to its topography, Japan’s rural population is unusually remote from the major centres of population. The four main islands of the archipelago (Hokkaidō, Honshū, Shikoku and Kyūshū) include a good deal of mountainous terrain that host a large number of very small settlements (hamlets and villages) with limited accessibility, and a significant minority of Japanese live on the 419 smaller inhabited islands. As a result, Japan, though highly urbanised, has more citizens living in remote rural regions (7% of the population), as defined by the OECD’s regional typology (Box 1.1), than in rural areas close to cities (5%). This is a characteristic Japan shares with only five other countries.9 As the population shrinks, therefore, the challenges facing rural Japan, particularly in terms of service delivery and connections to the rest of the economy, may prove to be far harder to address than they would be in a country where rural dwellers were concentrated closer to the major cities.

Box 1.1. The OECD regional typology

The OECD classifies TL3 regions as predominantly urban, predominantly rural and intermediate. This typology, based on the percentage of regional population living in rural or urban communities, is not as fine-grained as many national definitions but it allows meaningful comparisons among regions of the same type and level. Since national definitions vary, comparisons based on national figures can mislead. The regional typology is based on three criteria.

The first identifies rural communities according to population density. A community is defined as rural if its population density is below 150 inhabitants per square kilometre (500 inhabitants for Japan to account for the fact that its national population exceeds 300 inhabitants per square kilometre).

The second criterion classifies regions according to the percentage of population living in rural communities. A TL3 region is classified as predominantly rural, if more than 50% of its population lives in rural communities and predominantly urban, if less than 15% of the population lives in rural communities. If the share of population in rural communities is between 15% and 50%, it is categorised as intermediate.

The third criterion is based on the size of the urban centres. Accordingly, a region that would be classified as rural on the basis of the general rule is classified as intermediate if it has an urban centre of more than 200 000 inhabitants (500 000 for Japan) representing no less than 25% of the regional population. A region that would be classified as intermediate on the basis of the general rule is classified as predominantly urban if it has a urban centre of more than 500 000 inhabitants (1 000 000 for Japan) representing no less than 25% of the regional population.

Predominantly rural regions are sometimes further subdivided into remote rural regions and rural regions close to a city on the basis of the driving time needed for at least half of the population in a region to reach a populated centre of 50 000 or more inhabitants.

Source: OECD (2013a), Regions at a Glance 2013, OECD Publishing, Paris,

The demographic transition Japan is now experiencing has tended to reinforce this polarisation of the settlement pattern between high-concentration urban areas and remote rural places. Japan’s intermediate and rural regions experienced declining population during 1995-2013, as did most of its major urban areas (see Chapters 3 and 4 for details). Moreover, although both proximate and remote rural regions lost population overall, the contraction was more severe in rural remote regions, which recorded an average annual change in population of -0.45%, than in rural regions close to cities (-0.34%). While a small number of rural communities have devised strategies for attracting new inhabitants and have begun to grow again (see Chapter 4), the nation-wide picture is overwhelmingly one of rural-to-urban migration.

Rural areas are also ageing faster

Ageing has also proceeded further in rural areas. Examining Japanese prefectures in light of the OECD regional typology, which classifies regions into predominantly urban (PU), intermediate (IN), predominantly rural and close to a city (PRC) or predominantly rural remote (PRR), clarifies a number of things about the pattern of ageing. First, as one might expect, the elderly dependency ratio of the population rises with rurality: in 2012, it was lowest in PU regions, higher in intermediate and proximate rural regions (in that order) and highest in rural remote regions (Figure 1.8). The same pattern holds true with respect to the ratio of elderly to working-age population, except that the differences were far greater, owing to the fact that urban places tend to have lower fertility and rural places higher. This may be expressed more simply when inverted: the working-age share of the population tends to be highest in predominantly urban regions and lowest in remote rural regions. This is in fact the typical pattern for OECD countries, but the gap is extremely wide in Japan: only three other OECD members have greater differentials in the dependency ratio between predominantly rural and urban regions (Figure 1.9). Moreover, the rural-urban gap in dependency ratios appears to be growing more pronounced. The rise in the elderly dependency ratio between 1990 and 2012 was lowest in PU regions, somewhat greater in intermediate regions and highest in rural regions.

Figure 1.8. Ageing indicators for Japanese prefectures

Note: The elderly dependency ratio is defined as the ratio of elderly to working-age people. PU=predominantly urban; IN=intermediate; PRC=predominantly rural, close to a city; PRR=predominantly rural remote. The coloured lines above each series show the average values for each category.

Source: OECD (2015e), “Metropolitan areas”, OECD Regional Statistics (database),, (accessed 10 September 2015).

Figure 1.9. Elderly dependency rates, 2012
Country averages and averages for urban and rural regions within countries

Note: The elderly dependency ratio is defined as the ratio of elderly to working-age people. Latest available year 2011 for Australia and the United States.

Source: OECD (2013b), How’s Life? 2013: Measuring Well-being, OECD Publishing, Paris,

Because the regional typology can only classify prefectures by predominant settlement pattern, it is worth looking also at Japan’s functional urban areas (FUAs), as defined in OECD (2012a). The picture here is the same: in 2010, the elderly share of the population was 6.5 percentage points lower in the 76 FUAs with populations of 50 000 or more than in places outside of these urban areas, although elderly dependency ratios were rising strongly in both FUA and non-FUA areas, up almost six percentage points over the course of a decade.

Box 1.2. Urban data sources for this review

For statistical purposes, OECD countries define cities in different ways; even the definition of urbanisation and the urbanisation rate is not uniform across countries. In most OECD economies, the available official statistics are based on cities as administrative units. However, this can create problems with respect to comparisons across – or even within – countries, since the relationship between a city defined by its administrative boundaries and a city defined in terms of urban construction, settlement patterns and labour market flows can vary quite radically. In some cases, the two more or less coincide; in others, a large conurbation – effectively, the city defined as a socio-economic entity rather than an administrative one – can encompass hundreds of separately constituted municipalities (OECD, 2015f). For this reason, some countries define statistical metropolitan areas, based on settlement patterns and economic (usually commuting) flows. In an effort to foster greater comparability within and across countries, the OECD, in collaboration with EC and Eurostat, has developed a new approach to classifying urban areas with the aim to better monitor urban development within and across countries. The method adopted is used to define the functional urban areas (FUAs) referred to in this review; it is described in detail in Chapter 3 (Box 3.1). The OECD’s Metropolitan Database covers the (275 convergence here) FUAs in the OECD area with populations in excess of 500 000. These FUA data form the basis of much of the analysis in Chapters 2 and 3 of this review. However, the OECD-defined FUAs do not correspond to existing statistical units in any OECD country, so the time period and range of variables in the database are limited. It is therefore necessary to supplement the FUA data with other sources.

This review therefore considers developments in urban Japan in three ways:

  1. In presenting overall trends and challenges, it relies chiefly on prefectural-level data, because they cover the whole country and because of the greater coverage they offer in terms of both time and range of indicators. At times, it considers developments in rural and urban areas on the basis of the typology presented in Box 1.2 above, particularly when making cross-country comparisons.

  2. In looking at the three main metropolitan areas, centred on Tokyo, Osaka and Kyoto, the review looks at times at the Metropolitan Areas defined by Statistics Japan. These do not correspond to the OECD-defined FUAs, which are based on functional labour markets. Because they are constructed on the basis of prefectures, they permit the use of the rich body of prefectural-level data to analyse these metropolitan areas.

  3. Where reference is made to the term FUAs, the OECD Metropolitan Database is employed. From a methodological point of view, this is the best available definition of cities in Japan but, as noted above, the data coverage is limited. In Chapter 2, the micro-data analysis presented involves the use of the FUA boundaries in conjunction with individual-level data.

Although each of these definitions differs from the others, the findings presented here tend to reinforce one another, giving confidence that the overall picture of urban trends does not depend on the selection of one definition or another.

The consequences of demographic change

Japan’s demographic transition will create new economic, social and service-delivery challenges

The economic consequences of demographic change are myriad and complex (Table 1.1). The anticipated effects include shrinking domestic markets and the loss of economies of scale in some activities, as well as the fiscal impact of increased age-related spending and rising dependency ratios. On the positive side, some anticipate that population decline in the OECD’s most densely populated large country could create opportunities for more space-intensive activities, more flexibility in land use, less congestion, lower housing costs and, in some respects, less environmental pressure. Many outcomes are uncertain, since what Japan is experiencing now is without precedent in recorded history: such a contraction of population in the absence of a major epidemic or prolonged war has never occurred anywhere, let alone in conjunction with the rapid population ageing that is now under way.

Table 1.1. Possible consequences of demographic change

Potential benefits and opportunities

Potential costs and challenges

Population ageing

High life expectancy.

Demand for goods and services and new market opportunities (“silver economy”).

Rising burden of pensions and age-related services.

Shrinking labour force relative to population.

Less entrepreneurship and innovation.

Less demand for “non-silver” goods and services.

Population decline

Less congestion.

Opportunities for space-intensive activities.

Decreasing environmental pressure.

Flexibility in land use.

Lower housing costs.

Loss of tax base.

Shrinking labour force.

Smaller domestic market.

Challenges to efficient service delivery.

Source: Author’s elaboration.

One thing that does seem to be clear, however, is that the economic consequences of ageing are not a given: they are in large measure determined by policy choices. In a wide-ranging cross-country study, Oliveira Martins et al. (2005) concludes that many of the apparent negative economic consequences of ageing result not so much from the ageing process itself as from the sometimes perverse interactions between ageing and existing distortions, most notably labour markets and pension systems that encourage early withdrawal from the labour force. The economic impact of ageing thus depends in large measure on how longevity gains are managed. Well-articulated and complementary reforms to support, among other things, healthier ageing, longer careers, more efficient healthcare provision, etc., are more likely to offset the impact of ageing and manage longevity than piecemeal approaches that treat particular problems in isolation.

Some economic impacts, however, are clear and are already being felt. The most obvious and immediate are a shrinking labour force and rising elderly dependency ratios. While total population grew until 2010, the working-age population has been shrinking in absolute terms for two decades and as a share of the population for somewhat longer. Moreover, the share of older cohorts in the working-age population has been rising steadily, a harbinger of further increases in the old-age dependency ratio to come. By 2013, there were fewer than 2.5 people of working-age for every one above the age of 65, down from more than 11 in the 1950s. By 2050, this ratio is projected to fall below 1.33. Fortunately, Japan has unusually high employment rates – 71.7% of the working-age population was employed in 2013, as against an OECD average of 65.3%. Even so, in 2013 there were only 1.77 working-age people in employment for every person above the age of 65. This compares to a 2011 OECD average of just under three.10 This, moreover, considers only the old-age dependency rates – not the (shrinking but still significant) share of children in the population. While these figures also omit the contribution made by older Japanese who remain active – and there are many of these – it remains the case that Japanese workers must shoulder an increasing burden if the country is to sustain the growth of per capita living standards with a rapidly growing inactive population.

Falling population densities will also create challenges for efficient service delivery, especially – but not only – in rural areas. It is already the case that, in many rural areas across the OECD, especially the more remote ones, certain services are either unavailable or are available only at considerably higher cost and/or lower quality than in urban locations (OECD, 2010a). In part, this simply reflects the obvious potential of larger places to support both a wider variety of services and a greater number of entities that provide any particular service. Distance from major urban centres makes all forms of connectivity more expensive; it also imposes a similar burden within rural areas because of extensive geography and low population densities. This is particularly the case in remote rural areas of Japan, above all in the mountains and on small islands. Urban densities can also facilitate connectivity, whereas the dispersion of people in rural regions tends to make it harder to achieve. While some technologies (ICT) have reduced the distance-related challenges facing rural regions, most of the ways rural people exchange goods, services and ideas are still subject to distance penalties.

Even setting aside transport costs, small – and falling – populations in many places will make it harder to achieve economies of scale in the production of many goods and services, including public services. Shrinking cities and towns will also face important challenges, particularly where declining demand makes it difficult to maintain urban infrastructures or to operate them efficiently. Under-use of water and wastewater infrastructure, for example, can make for environmental and health hazards; grid-bound infrastructure cannot easily be thinned out in response to population loss in many cases, because the integrity of the grid as whole must be maintained and a certain level of usage must be maintained for the system as a whole to function (Hoornbeek and Schwarz, 2009).

Demographic change may also affect entrepreneurship and innovation

One important issue concerns whether and to what extent the ageing of the population and the workforce will make the Japanese economy less dynamic when it comes to innovation and entrepreneurship – the latter, in particular, being an area in Japan’s recent performance that has been lacklustre (OECD, 2015g). The issue is much debated in countries with rapidly ageing populations. On the demand side, population ageing presents enormous opportunities for innovators and entrepreneurs in such fields as medicine, construction, transport and finance, to name but a few. The question is how demographic change will affect the supply of entrepreneurship and innovation. The evidence from many countries suggests that older individuals are less likely to become entrepreneurs, aside from self-employment, and that they are less likely to found businesses based on process or product innovation (Wadhwa, Freeman and Rissing, 2008; Wadhwa et al., 2009; Botham and Graves, 2009; Colovic and Lamotte, 2012; Stangler and Spulba, 2013; Kautonen, 2013). On the positive side, several studies point to higher success rates for older entrepreneurs, whose greater experience, higher initial wealth and more extensive networks in many cases help them to succeed. Third-age entrepreneurs leading high-growth firms are not less innovative than younger high-growth entrepreneurs, though older people are less likely to found such high-growth ventures (Accenture, 2012; Wadhwa Freeman and Rissing, 2008; Kautonen, 2013; Botham and Graves, 2009). Haga (2014) highlights the potential of third-age entrepreneurship in rural settings, specifically with reference to Japan.

These findings are not wholly surprising. The opportunity cost of time increases with age, and one would expect this to make older individuals less attracted to activities that yield rewards further in the future and with higher levels of associated risk, as opposed to activities that generate immediate and riskless returns, such as salaried employment (Lévesque and Minniti, 2006). In a recent study for the OECD, Kautonen (2013) highlights the distinction between “push” and “pull” factors motivating entrepreneurs. They may be pushed out of the traditional labour market by such factors as age-discriminatory practices in recruitment, promotion and training or poor labour market prospects following a redundancy. On the other hand, they may be drawn into entrepreneurship because it offers greater potential rewards (whether pecuniary or otherwise), a more flexible alternative to conventional employment and/or a better work-life balance. With a shrinking and ageing labour force, the push factors are likely to grow weaker over time. Kautonen concludes that the propensity of individuals to seek to own and grow a business rises with age up to about 40 and then declines – rather sharply after about 50. By contrast, an individual’s propensity to opt for self-employment, or even to be pushed into it, rises with age.

Kautonen’s most critical findings for Japan touch on two points. First, there are enormous variations across countries in third-age entrepreneurship, which is particularly widespread in such countries as Iceland, the United States. Secondly, these differences are at least in part the product of public policies. In other words, central and local governments can do much in order to affect the relationship between ageing and entrepreneurship. Chapter 4, in particular, will look at the ways in which the Japanese authorities can promote entrepreneurship and innovation at all ages, especially outside the largest cities.

The implications for public finances are worrying

Demographic change has enormous implications for public spending. Indeed, it is one of the major factors underlying Japan’s public finance problems. Pensions, health-care and long-term care, which are mainly focused on the elderly, now account for about 90% of public social spending, which, in turn, is the fastest-growing expenditure item in the state budget. Since 1990, social security-related expenditures have almost tripled, while all other major categories of spending11 except debt service have shrunk. Social security spending in 2015 accounted for 32.7% of total expenditure and 55% of general expenditures,12 up from 17.5% and 32.8%, respectively, in 1990. These figures chiefly concern current spending and they must, of course, be considered alongside the impact of population ageing on tax receipts, which the finance ministry already recognises as a substantial cost to the budget.

Demographic change will also entail substantial expenditure on infrastructure, for several reasons. First, as will be seen, the authorities anticipate significant investment in new infrastructures and in the upgrading of existing infrastructures to help adjust the country’s settlement pattern to a far smaller population. Secondly, many facets of public infrastructure are to be adapted to the needs of much older population. And, thirdly, even networks for which demand may fall will entail some costs. Downsizing network infrastructures can be very costly, but failure to shrink them efficiently can also create problems and can be very costly in the long run. For example, the under-use of water networks can lead to corrosion of pipes and possible contamination (Hoornbeek and Schwarz, 2009). Many areas will confront the costs associated with such under-used infrastructures, as well as housing and building vacancies. More generally, they will confront rising costs for service delivery in places where population declines erode economies of scale (i.e., the fixed costs of service delivery, which are often substantial, will be shared across fewer clients). Finally, there is likely to be expenditure in some areas to address the environmental costs of depopulation: in rural places where local communities are responsible for things like the management of forests and watercourses, the abandonment of some settlements could otherwise entail significant environmental damage (Odagiri, 2012, 2015).

The economic context

Growth and productivity performance has long been disappointing

Economic growth has been sluggish for roughly a quarter-century. By 2014, per capita income had fallen from roughly equal to the average of the top half of OECD countries in the early 1990s to a level 14% below that average (Figure 1.10). The collapse of the asset price bubble in the early 1990s was followed by an extended period of corporate restructuring and a banking crisis. While the worst of this process unfolded during the mid-to-late 1990s, Japan’s performance has remained weak since the turn of the century. Over the period 1995-2013, real growth of GDP per capita averaged just over 0.8%, the third-lowest growth rate in the OECD, and in 2014, the economy slipped into recession for the fourth time since 2008 (OECD, 2015h). Though the economy rebounded again, growth remained subdued, with the OECD in mid-2015 projecting real GDP growth of just about 0.7% for the year.

Figure 1.10. Gross domestic product per capita
USD, current prices and current PPPs

Source: OECD (2015i), OECD Compendium of Productivity Indicators 2015, OECD Publishing, Paris,

The good news here is that Japan’s productivity performance, though not outstanding, has been somewhat better than its growth performance. During 1995-2013, Japan ranked third from the bottom among OECD economies in the growth of GDP per capita. However, it ranked 16th in terms of the growth of labour productivity over the period (Figure 1.11) – hardly spectacular but still relatively close to the OECD average. In the most recent period (2007-13), productivity growth was even a little above the OECD average. However, it was far lower in 2007-13 than in previous periods, so the change in Japan’s relative position was a result of greater deterioration elsewhere following the crisis rather than an acceleration of productivity growth in Japan. There is a paradox here, insofar as Japan invests heavily in many of the factors that should drive productivity growth – particularly human capital and research and development – and yet it has recorded relatively poor productivity performance for a quarter-century. Overall, Japanese productivity in 2013 was 4.7% below the OECD average. What makes this especially worrying in a shrinking labour market is that the gap in terms of GDP per hour worked was far greater, at almost 14.9%; it was partially offset by the greater number of hours worked (OECD, 2015i).

Figure 1.11. Growth of labour productivity, 1995-2013
GDP per hour worked, total economy, percentage change at annual rate

Source: OECD (2015i), OECD Compendium of Productivity Indicators 2015, OECD Publishing, Paris,

Demographic change makes productivity growth more important than ever

Reviving productivity growth is perhaps Japan’s most urgent economic priority. If an ageing economy is to sustain continued prosperity, then output per worker must rise faster than would be required in a healthier demographic context. The increasing importance of productivity for growth – and the growing drag effect of demographic change – is illustrated in Figure 1.12, which presents a slightly unusual decomposition of real GDP growth in OECD countries over 1999-2011.13 Instead of simply contrasting the contributions of labour input and output per worker, it splits the labour-supply component into two: one that depends on individuals’ activation decisions and the state of the labour market (i.e., employment and labour force participation) and another that reflects the combined effects of ageing and migration (change in the working-age population). As can be seen, Japan is the only OECD country in which there is a substantial negative effect of demography – equivalent to just over one-half a percentage point of GDP growth. This negative contribution is also, as noted above, set to increase, because the rate of shrinkage of the working-age population is projected to accelerate over the coming decades. Between 2015 and 2050, the working-age population is expected to decline by an average of 1.2% per year on average and in some periods by up to 1.7%.

Figure 1.12. Decomposition of real GDP growth for selected OECD countries, 1999-2011

Note: “Change in activation” reflects only the combined effect of changes in the participation rate and changes in the employment rate. The expansion or contraction of the working-age population is encompassed in the “demographic change” variable.

Source: OECD (2015j), OECD Economic Outlook: Statistics and Projections (database), (accessed 16 October 2015).

In productivity terms, then, Japan will have to run faster and faster simply to maintain its position in relation to other economies. To sustain a constant rate of growth of GDP of, say, 2% per annum (the government’s current target for 2020) over the period to 2050, and assuming a constant employment rate, Japan will need to achieve an average annual rate of growth of about 3.3% in GDP per worker. Of course, the other things being equal assumption in this example is wholly unrealistic, but it serves to highlight the magnitude of the purely demographic contribution to growth performance. In reality, increases in labour force participation and longer careers must surely be part of the solution, as well as labour market reforms to increase labour force participation – and, in particular, full-time employment – among women. At present, the male labour force participation rate is about 21 percentage points above that of women. OECD (2015h) observes that if female employment rates were to converge to those of men over the next 20 years, the labour supply would decline by only 5%, as compared to a projected decline of 17% (Figure 1.13). The opportunity cost of existing gender imbalances is all the greater, given that young women in Japan tend to be more educated than young men.14

Figure 1.13. The potential impact of female employment on labour supply


1. Assuming that the labour force participation rate for men remains constant from 2011 to 2030.

Source: OECD (2014), Japan: Advancing the Third Arrow for a Resilient Economy and Inclusive Growth, Better Policies, OECD Publishing, Paris,

Unemployment has remained low but inequality has increased

Despite sluggish growth for more than two decades, Japan continues to experience relatively high activation rates and low unemployment. In 2013, the OECD’s estimates of harmonised unemployment rates15 showed Japan at just 4.0%, lower than all but two OECD countries, while its activation and employment rates were well above the OECD averages for men and women in all age groups. Older (55-64) people had particularly high employment rates by OECD standards. Overall, Japan’s ratio of employment to working-age population, at 71.7%, was 6.4 percentage points above the OECD average. This is good news, and it is likely to have reduced considerably the level of social stress caused by this prolonged period of economic stagnation. Strikingly, though, the gap between men and women, at 18.3 percentage points, was wider than the OECD-wide differential (16 points), though Japanese women still had employment rates well above the OECD average.

Continued high employment rates and low unemployment have not been sufficient to prevent rising income inequality and poverty. The Gini coefficient for interpersonal income (net of taxes and transfers) reached 33.6 in 2009, the tenth-highest in the OECD area and somewhat above the OECD average, while the relative poverty rate16 in 2012 was the sixth-highest in the OECD area. There is, however, comparatively little variation in levels of inter-personal inequality across Japanese regions: the interpersonal Ginis for all Japanese TL2 regions fall within about three points of the national average, and the country ranks 17th among the OECD-28 in terms of the variation across regions in the Gini coefficient for personal incomes. Moreover, the inter-regional variation in Ginis post-taxes and transfers is less than half that observed before taxes and transfers are taken into account, which suggests that the fiscal system plays an important role in smoothing out spatial inequalities, even if it is less effective in reducing inter-personal inequality (OECD, 2015h).

A further consequence of Japan’s long-term low-growth trajectory has been increasing labour market dualism (CJF, 2014). Japan’s labour market is segmented between regular and non-regular workers. Non-regular workers are primarily part-time, fixed-term and dispatched workers (i.e. workers sent from private employment agencies). The share of non-regular workers rose for almost three decades, from 16.4% in 1985 to a peak of 38.2% of total employment in 2012; it fell back somewhat to 37% in 2014. Non-standard contracts offer firms greater flexibility and enable them to hold down costs, but they also have a number of troubling effects. OECD (2015h) notes that non-standard workers not only have lower wages and less job security, they typically also receive less firm-based training and are less well protected by the social safety net (many are not covered by employment insurance, despite the greater risk of unemployment). The evidence also suggests that, once employed in non-regular jobs, workers have little chance to move into regular employment: the second-tier labour market is, for many, a trap (OECD, 2013c).

Labour market dualism is also reflected in subjective measures of life satisfaction. While Japan scores better than most OECD countries on many dimensions of well-being, such as life expectancy and education, it scores near the bottom of the OECD on measures of work-life balance, a dimension of well-being directly linked to labour market institutions and practices, as well as health status and reported life satisfaction (OECD, 2013b). Moreover, the happiness level reported by non-regular workers is significantly below that of regular workers and the self-employed (OECD, 2015h). The work-life balance issue also has implications for productivity. The evidence suggests that very long working hours are indeed detrimental to labour productivity: up to a certain level (perhaps 48-50 hours per week), output appears to be roughly proportional to hours; above that threshold, output rises at a rapidly decreasing rate as hours increase (Pencavel, 2014). The long-hours culture of many countries thus confers no productivity advantage and imposes high costs in terms of accidents and injuries (Dembe et al., 2005; Rho, 2010; Ricci et al., 2007), as well as life satisfaction.

Labour market institutions and practices seem to contribute to demographic decline

The evidence also suggests that labour market segmentation and work-life balance contribute to Japan’s demographic situation. Roughly 70% of non-regular workers are women, while men account for an almost identical proportion of regular employees. Strikingly, Japan has among the lowest rates of employment for university-educated women in the OECD, despite the fact that the female employment rate overall, at 63%, is five points above the OECD average. OECD (2015h) cites analyses finding that, after adjusting for type of job and educational attainment, the wage gap between full- and part-time workers is 31% for men and round 48% for women. The income of a household headed by a regular worker aged 45-49 was more than four times that of a household headed by a non-regular worker. The lower income of non-regular workers also discourages marriage and hence reduces the fertility rate. Women with children find it particularly difficult to sustain regular, full-time employment, owing to the country’s long-hours culture – Japan ranks third, after Turkey and Mexico, in the proportion of employees working 50 hours a week or more on average. Moreover, commutes are often very long, employment practices can be relatively inflexible for regular workers and available child-care is lacking.17 Although the authorities have in recent years taken steps to reduce the barriers to combining parenthood with employment, labour market institutions still tend to force a choice between career and child-bearing. The very low fertility rate is to some extent a reflection of that choice (CJF, 2014).

A look at prefectural-level data confirms two important facts about this situation. First, as can be seen from Figure 1.14, there is a clear and consistent positive relationship between fertility and female labour supply in terms of hours worked in Japanese prefectures. There is also a consistent and positive relationship between fertility and female labour force participation. The issue is not a choice between policies promoting high fertility and low employment, as opposed to low fertility and high employment. It is possible to achieve both high fertility and high female employment. Indeed, whereas there was a broadly negative relationship between female employment and fertility in OECD countries in 1980, by the mid-2000s the relationship was positive, suggesting that women are likely to have more children where structures are in place to support combining family and work rather than forcing a choice between them (OECD, 2007). The second point is that both fertility and female labour force participation are consistently higher in predominantly rural regions than in urban areas: in 2008, female labour force participation rates were on average 13 percentage points higher in PR prefectures than in predominantly urban ones, while fertility rates were almost 12% higher. Intermediate regions fell between the two on both counts. For a variety of reasons to be explored in Chapter 3, Japanese cities are particularly inhospitable to combining careers with child-rearing. Yet economic pressures are drawing more young people, particularly young women, to the cities – and they are being drawn disproportionately to Tokyo and the other very large metropolitan areas, which tend, in turn, to have the lowest fertility (Cabinet Secretariat, 2015).

Figure 1.14. Fertility and female labour supply, Japanese prefectures, 2011

Source: OECD (2015d), Regional Statistics, (database), (accessed 10 September 2015); information provided directly by Official Statistics Japan (Government of Japan).

A generation of underperformance has put public finances under strain

Japan’s public finance challenge has been building up over a very long period, as government spending has risen sharply since 1990, due to population ageing and frequent fiscal stimulus packages, while revenues have failed to keep pace. Gross government debt had reached 226% of GDP by 2014, and net debt reached 129% of GDP; both figures are the highest in the OECD. The debt continues to climb, with the primary budget deficit hovering around 8% of GDP during 2008-13, before declining to roughly 7% in 2014. Sluggish growth and rising social spending have meant that total revenues have largely stagnated, while total expenditures have climbed steadily. Persistent deflation has aggravated the growth of debt ratios by reducing nominal GDP. Public finances came under further pressure as a result of the 2011 Great East Japan Earthquake, the worst disaster in Japan’s post-war history.

The impact of such high debt on government interest payments has hitherto been mitigated by exceptionally low interest rates. Indeed, the effective interest on government debt in early 2015 was the lowest in the OECD at 0.9%. Net interest payments were thus limited to just 1.2% of GDP, well below the OECD average of 1.6%, despite Japan’s very high debt levels (OECD, 2015h). Around 90% of government debt is domestically held, limiting the government’s external debt to less than USD 1 trillion in mid-2014, far below the levels of the United States (USD 6.6 trillion) and the euro area (USD 4.2 trillion). However, this relatively benign state of affairs will not continue indefinitely, leaving Japan vulnerable to rising interest rates in future, even as demographic change continues to add to the strain on public finances. A loss of confidence in Japan’s fiscal sustainability could result in a run-up in long-term interest rates, with negative implications for the financial sector, fiscal sustainability and growth.

This danger is one of the major constraints on the government’s ability to use fiscal stimulus to prevent growth from stalling. It is also a crucial factor conditioning the response to population ageing, since it ensures that the social and infrastructure spending needed to cope with demographic change will remain under pressure for some time. Financing the transition to a new demographic equilibrium will be extremely difficult.

The government has undertaken a three-pronged strategy to re-launch growth

In early 2013, the government launched a three-pronged strategy, known colloquially as “Abenomics” after Prime Minister Shinzō Abe, aimed at reviving growth. The “three arrows” of Abenomics comprise:

  • bold monetary policy, aimed at fighting deflation, even via such unorthodox measures as “qualitative and quantitative easing” and the establishment of a 2% inflation target

  • flexible fiscal policy, intended to support growth in the short term while establishing a credible path to fiscal sustainability of over the medium-to-long term and

  • a package of structural reforms to relax the supply-side constraints on growth and trigger the revitalisation of Japan’s regions which was unveiled in June 2013 and revised a year later, in an effort to boost real GDP growth to 2% per annum over the period through 2022.

The first two arrows initially delivered encouraging results. In 2013-14, nominal GDP increased for 6 consecutive quarters for the first time in over 20 years, as inflation picked up and yen depreciation gave a boost to exports. However, the economy tipped back into recession following a consumption tax hike in April 2014. This led to the postponement of plans for a further hike in 2015, ensuring that the primary deficit would remain large until at least 2017. An OECD (2015h) simulation suggests that, in the absence of further steps to reduce the deficit, the government debt ratio would exceed 400% of GDP by 2040, even on the optimistic assumption of average nominal GDP growth of 3% per year. The government thus faces the very difficult challenge of trying to balance growth and fiscal sustainability objectives over the coming years, in a context of population ageing and decline.

The third arrow of Abenomics, on which there has been less progress to date, is where the regional dimension becomes prominent. At the end of June 2015, the government approved a number of key policy documents, including a new basic strategy for job creation in small cities and towns, as well as rural areas. Known colloquially as “local Abenomics”, they focus on employment generation as the key to sustaining local economies and, in particular, to attracting younger people. The key policy priorities outlined in the strategy include strengthening the efficiency of the service sector in regions, promoting regional innovation, and supporting efforts linked to the “branding” of regions. The strategy aims to overcome the fragmentation that has often prevailed across policy sectors and across prefectural and municipal boundaries, fostering co-operation across regions and across the public-private divide to unlock regional potential. Public-Private Partnerships (PPPs) and Public Finance Initiative (PFI) will be central to these efforts.18

Inter-regional disparities and living standards

Economic disparities among Japanese regions remain low…

As noted above, inter-regional inequalities in income in Japan are exceptionally low by OECD standards: in 2010, Japan had the second-lowest level of inequality among regions in GDP per capita (Figure 1.15). Moreover, inequality across Japanese regions has been falling over time, even as it has been increasing in most other OECD countries (Figure 1.16). In some respects, this may seem quite surprising, given the concentration of population and activity in the three main metropolitan areas. In this, Japan resembles perhaps some of the Nordic countries, which also combine high concentrations of population and economic activity with low territorial disparities. Like them, Japan has traditionally been a highly centralised country with a strong commitment to territorial cohesion. It is possible that the extremely high urbanisation rate contributes to lower rural-urban inequalities; the highest differentials in the OECD area tend to be in countries with lower levels of urbanisation, possibly because the greater relative weight of the rural population raises the cost of policies aimed at improving service provision outside the cities or raising the living standards of rural dwellers (Figure 1.16). As will be seen, the data also suggest that there is a labour market equilibrium in Japan that serves to limit wage disparities across space. This is likely to be the more important factor in explaining such low spatial inequality.

Figure 1.15. Gini index of GDP per capita across TL3 regions, 1995 and 2010

Source: OECD (2015d), Regional Statistics (database), (accessed 10 September 2015).

Figure 1.16. GDP per capita in predominantly urban and rural regions relative to the national average

Source: OECD (2015d), Regional Statistics (database), (accessed 10 September 2015).

Limited territorial disparities in per capita GDP are mirrored in the data for poverty rates for Japan’s major (TL2) regions. As noted above, poverty rates in Japan are comparatively high, but the variation in poverty rates across regions is smaller than the OECD average and well below the variation seen across regions in such large OECD countries as Germany, Italy, Turkey, and the United States.

…but the picture with respect to non-monetary aspects of well-being is more complex

Japan’s performance on the non-pecuniary dimensions of well-being is rather more mixed, as can be seen in Table 1.2. On indicators such as health, safety and jobs, Japan ranks very high among OECD countries and also has very low levels of inter-regional inequality. By contrast, it performs less well – and exhibits greater spatial disparities – on access to services, education and environmental quality.

Table 1.2. Performance and dispersion on regional well-being indicators in Japan

Dimension of well-being

Rank among OECD countries on this dimension

Rank among OECD countries concerning inequalities on this dimension (higher rank = greater inequality)

Access to services


















Source: OECD (2015g), “Regional well-being”, OECD Regional Statistics (database), (accessed 10 September 2015); for detail on the indicators, see OECD (n.d.), Regional Well-being website, (accessed 1 September 2015).

Japanese prefectures perform comparatively well when it comes to CO2 emissions (Figure 1.17, top panel). This reflects the fact that the country’s CO2 emissions per capita have long been below the OECD average, although by 2012 they were converging with it. Japanese emissions per capita rose 13.1% from a low point in 2009, while the OECD average slightly declined. Here again, the demographic factor has played a role, as aggregate emissions rose by somewhat less than the per capita figure. So, too, has economic performance: sluggish growth has helped to limit emissions, but the emissions intensity of the economy (CO2 emissions/unit of GDP at constant prices) rose steadily until the 2008-10 period, when it stabilised. It fell more than 6% in 2011-12, though it was still 4.6% higher in 2012 than in 2000, and it is not clear whether this trend has been sustained since.

Figure 1.17. CO2 and NO2 emissions, OECD TL3 regions

Source: OECD (2015d), Regional Statistics (database), (accessed 10 September 2015).

With respect to NO2 emissions, Japanese regions overall perform less well with respect to other OECD regions (some are very near the top of the distribution) and there is far greater disparity in performance (Figure 1.17, bottom panel). While both CO2 and NO2 emissions contribute to global environmental damage, the local effects of the latter are also quite serious, making high concentrations an important challenge for regions and cities. Short-term NO2 exposures, ranging from 30 minutes to 24 hours, have been linked with adverse respiratory effects including airway inflammation in healthy people and increased respiratory symptoms in people with asthma. Studies also show a connection between breathing high short-term NO2 concentrations and increased visits to emergency rooms and hospital admissions for respiratory issues, especially asthma. NO2 exposure concentrations are of particular concern for susceptible groups, including asthmatics, children, and the elderly. Moreover, emissions that lead to the formation of NO2 generally also lead to the formation of other harmful nitrogen oxides. Emissions control measures leading to reductions in NO2 can generally be expected to reduce population exposures to all gaseous nitrogen oxides. This can also help reduce the formation of ozone and fine particulate matter, both of which pose significant public health threats.

Japanese regions also perform less well in respect of particulate matter (PM), perhaps the most immediate threat to public health stemming from air quality. Exposure to small particulate matter can affect both lungs and heart. Fine particles typically contain microscopic solids or liquid droplets that are so small that they can get deep into the lungs and cause serious health problems, including: premature death in those with heart or lung disease, heart attacks, irregular heartbeat, aggravated asthma, reduced lung function, and respiratory symptoms such as irritation of the airways, coughing or difficulty breathing. They are also a main cause of reduced visibility (haze) in many places, and they can settle on ground or water, potentially making lakes and streams acidic; changing the nutrient balance in coastal waters and large river basins; depleting soil nutrients; damaging forests and farm crops; and reducing biodiversity. Data for PM are not available for small (TL3) regions across the OECD, but the data for large (TL2) regions show that most Japanese regions are in the middle-to-upper reaches of the distribution when it comes to fine particulate matter, or PM2.5 (Figure 1.18). Indeed, only Tōhoku and Hokkaidō fall in the lower half of the distribution. While the worst-affected region (Southern Kantō) ranks only 56th – making it far from the worst performer in the OECD area – the fact is that Japan’s population is heavily concentrated in the areas with the highest levels of PM2.5 (apart from Hokuriku). Altogether, more than 95% of Japan’s population lives in regions with average PM2.5 levels above the WHO standard (10 μg/m3 annual mean).

Figure 1.18. Exposure to PM2.5 in OECD TL2 regions

Source: OECD (2015g), “Regional well-being”, OECD Regional Statistics (database), (accessed 10 September 2015).

Trends in regional economic performance

As seen above, Japan’s growth performance, both in per capita terms and in aggregate, has been relatively weak since the early 1990s. This section examines the regional patterns that underlie this national performance, looking at both per capita and aggregate performance.

The growth of GDP per capita has been weak but broadly consistent across space

There has been a modest reduction in divergence in terms of GDP per capita since the mid-1990s. As can be seen in Figure 1.19, Tokyo even lost ground in per capita terms, albeit only slightly, during the decade to 2010. Owing to changes in the system of national accounts, data from before 2000 are not directly comparable, but an analysis using the old series, which runs from 1995 to 2009, reveals more or less the same picture. Concentration of population and activity has not so far resulted in greater inter-regional disparities, with roughly as many regions gaining ground as losing, and by remarkably similar magnitudes. Moreover, regions of all types (predominantly urban, intermediate and rural) may be observed along the entire spectrum. In fact, per capita GDP growth was higher in predominantly rural than in predominantly urban or intermediate regions over the period. However, growth across all regions was relatively weak, and no Japanese prefecture managed to achieve even the average level of growth recorded by all OECD regions over the period.

Figure 1.19. Trends in relative GDP per capita by prefecture, 2001-10
Ratio of prefectural GDP per capita to the national average (National average = 1)

Source: OECD (2015d), Regional Statistics (database), (accessed 10 September 2015).

Repeating the somewhat unorthodox decomposition of per capita GDP growth presented above at prefectural level (Figure 1.20) highlights the variation in the relative importance of productivity, labour-supply decisions and demographics across Japanese prefectures. In particular, it throws into sharp relief the far greater impact of demographic decline on intermediate and rural regions than on predominantly urban ones. The imperfect correlation between trends in productivity, defined as GDP per worker, and the growth of GDP per capita is also striking. Output per worker has grown in only four of Japan’s twelve predominantly urban regions. This need not imply that all other cities are performing badly. In some cases, poor productivity may be a side effect of migration: newcomers to cities tend to be – initially, at least – less productive than incumbent residents. As a result, rapid inflows of population often tend to reduce per capita and per worker growth rates even in very dynamic cities. Yet they are beneficial both for a country and for the affected individuals, as long as the new arrivals are more productive in the cities than they would have been in sending regions. However, in four predominantly urban regions,19 output per worker has fallen even against the backdrop of stagnant or declining labour input. This is very bad news indeed.

Figure 1.20. Decomposition of GDP growth by prefecture, 2000-10

Note: “Change in activation” reflects only the combined effect of changes in the participation rate and changes in the employment rate. The expansion or contraction of the working-age population is encompassed in the “demographic change” variable.

Source: OECD (2015d), Regional Statistics (database), (accessed 10 September 2015).

The pattern observed in the three big metropolitan areas (MAs) defined by Statistics Japan – specifically, those based on Tokyo (Kantō), Osaka (Keihanshin) and Nagoya (Chūkyō)20 – is noteworthy. The “core” prefectures of Tokyo, Osaka and Aichi (where Nagoya is located) all recorded exceptionally poor productivity performance, whereas a number of the better-performing regions were adjacent to them, within the same MAs: Ibaraki, Yamanashi, Tochigi and Gunma in Kantō; Kyoto, Shiga and Wakayama in Keihanshin; and Mie (the best-performing prefecture in Japan) in Chūkyō. In short, there have been significant divergences of performance within the MAs; the reasons for these are examined in more detail in Chapter 3.

Aggregate growth, by contrast, has been far more concentrated

Strikingly, while the growth of GDP per capita has been relatively consistent across regions, even leading to some reduction in inter-regional disparities, aggregate growth has been highly concentrated (Figure 1.21). At first glance, the degree of prefectural-level concentration seems to have gradually diminished over time – evident in the way the curve shifts gradually away from the origin in successive periods. However, this is mainly due to the declining contribution of Tokyo Prefecture, as well as the rising shares of a handful of dense, nearby prefectures. It definitely does not reflect any nation-wide spread of growth drivers. Overall, the growth contribution of the top ten prefectures rose from 68.6% in 1995-99 to 72.6% in 2003-07. The contraction that followed the crisis was even more concentrated: just ten prefectures accounted for 85% of the drop in output. What the data point to, then, is not a deconcentration of growth on any large scale but rather a rebalancing of growth within and among the three major MAs defined by Statistics Japan. Figure 1.22 shows that the combined growth contribution of these MAs has risen over time – they accounted for almost 75% of aggregate growth in the run-up to the crisis and 80% of the contraction that followed. This is well above their share of total GDP at the start of the period (roughly 62.5%). The contribution of Kantō fell somewhat over the period, while that of Chūkyō rose significantly. As will be seen, the analysis of functional urban areas in Chapter 3 presents a broadly similar picture.

Figure 1.21. Contributions to growth by prefecture, 1995-2010

Source: OECD (2015d), Regional Statistics (database), (accessed 10 September 2015).

Figure 1.22. Growth contributions of the main metropolitan areas

Source: OECD (2015d), Regional Statistics (database), (accessed 10 September 2015).

Productivity performance has varied widely across prefectures

Figure 1.23 shows the degree of variation in the relative growth rates of productivity and GDP over the decade to 2010. While the average rate of (aggregate) GDP growth barely exceeded 2% per annum in just 3 of Japan’s 47 regions, labour productivity growth exceeded 2% in no fewer than 15 prefectures, including some that recorded little or no growth overall, and was at or above 3% in 3. This highlights the degree to which increasing concentration of economic activity is being driven by population change (differences in migration and ageing) rather than faster productivity growth in the big cities. The major cities’ share of economic activity appears to be rising simply because that is where more and more people are.

Figure 1.23. GDP and labour productivity by prefecture, 2001-10

Source: OECD (2015d), Regional Statistics (database), (accessed 10 September 2015).

Table 1.3 presents an overview of growth and productivity performance by region type. It suggests several observations. First, while GDP per capita growth was similar across different types of regions in the run-up to the crisis, rural regions were far less affected by the downturn. Secondly, rural regions consistently record the highest levels of growth in output per worker. Thirdly, intermediate regions likewise do better than urban ones in terms of productivity, but in many places this has been insufficient to sustain regional prosperity. The impact of ageing and migration has been such that GDP per capita has fallen despite rising productivity.

Table 1.3. GDP growth by type of region: Japan and the OECD
Average annual real growth, %

A. GDP per capita










Predominantly urban














Predominantly rural







B. GDP per worker










Predominantly urban














Predominantly rural







Note: Averages for Japanese and all OECD TL3 regions are not weighted.

Source: OECD (2015d), Regional Statistics (database), (accessed 10 September 2015).

The economic policy response to demographic change

To a great extent, the focus of this review is on policies to address the spatial/territorial dimension of the two-fold challenge of productivity and demographic change. However, economy-wide policies will have a vital role to play if Japan is to meet this challenge. They will also to some extent condition the prospects for success of regional and local initiatives. This section therefore presents a brief overview of economy-wide fiscal and structural policy priorities for restoring the public finances, boosting productivity and adapting the economy to demographic change.

Much remains to be done to address the public spending consequences of ageing

Achieving public finance sustainability is a critical priority, not least via reforms to expenditure items heavily influenced by ageing (pensions, healthcare, long-term care, etc.) but also by improving effectiveness and efficiency of service provisions in regions (merging healthcare centres for improving efficiency, utilising so-called small stations in villages to work as service hubs and leading healthy life in regions, reducing public investments in sparsely populated areas, re-examining the demands of large infrastructure, etc.). While the pension reforms of the last decade did much to ensure the long-term sustainability of pension provision, there is still more to do, and OECD (2015h) argues for raising the pension eligibility age, in order to reduce the fiscal burden while raising the labour force participation of older persons (Sutherland, Hoeller and Merola, 2012). Pension eligibility ages for both men and women are still low compared to Japan’s life expectancy of 83.48 years.21

In contrast to pensions, health and long-term care spending is expected to continue rising. Between 2000 and 2012, healthcare spending rose from 7.6 to 10.1% of GDP, as spending rose by more than 2% per year while nominal GDP fell. Rising costs are driven mainly by spending on pharmaceuticals and very long hospital stays. Average hospital stays in Japan are the longest in the OECD, almost four times the average. This is linked to long-term care; the average stay for curative care in 2012 was just over half the average hospital stay (still the longest in the OECD), because only about half of hospital patients needed healthcare (OECD, 2015h). Reliance on hospitals to address what are in fact long-term care needs is expensive and reflects the lack of resources for home-based care and/or specialised long-term care institutions. Japan has been expanding the provision of long-term care (LTC) for the elderly for some years, but ensuring that provision keeps pace with population ageing has proved an enormous challenge (OECD/European Commission, 2013; Matsuyama, 2015; Asai, 2015).

The burden of providing LTC for the elderly is set to grow as the population ages, but its spatial impact is difficult to foresee. On the one hand, rural areas are ageing faster. There are, moreover, economies of scale in elderly care, and care at home is easier when the elderly are not widely dispersed, as they can be in rural places. However, the cost of both labour and space tends to be higher in cities, particularly the large ones, and providers in Tokyo and other major cities complain that systems of public support for LTC do not adequately reflect cost differentials between places (Matsuyama, 2015). Some rural places have tried to maintain population levels by specialising in welfare services for older people (variously known as, e.g. Silver Areas or Welfare Villages). The Japan Policy Council (JPC) recently argued for moving more of Tokyo’s elderly to rural regions for nursing care, on the basis that over the next decade there would be a shortage of around 130 000 beds for LTC in the Greater Tokyo area over the coming decade. Indeed, labour supply appears to be a key constraint in many places – in the cities, because better pay is available in other activities and in rural areas because they often struggle to attract younger workers to staff LTC services. The JPC proposal has met criticism in some of the proposed “destination regions”: though the Council suggests that they have spare capacity to provide additional health and nursing care, some of them face a chronic shortage of nursing-care providers and see the Council’s proposal as an attempt to dump the capital’s burden on them (Asai, 2015).

Healthcare and long-term care are, of course, only one aspect of the problem, since much can be done to promote healthier ageing. Indeed, much is already being done, and this is reflected in Japan’s extraordinary levels of healthy life expectancy, which the WHO estimates at 75 years in 2012. Healthy ageing is also an area in which there is much that regions and cities can do, as is evident in the recent experiences of Japanese cities like Toyama and Yokohama (OECD, 2015c). However, super-ageing brings its own problems: while people in their 60s and 70s are now healthier than ever, the rapid growth of the very elderly – over-80s are projected to constitute more than 16% of the population by 2050 – will entail costs of its own, particularly for dealing with diseases like Alzheimer’s and dementia.

Structural reforms particularly need to address labour supply and fertility…

OECD (2015h) sketches a range of structural reforms that could enhance innovation and competition, thereby strengthening productivity performance, while mitigating the problem of declining labour supply. The most important concerns here include increasing labour force participation, particularly among women and older workers (enhancing the provision of training to such workers, reducing inefficient overtime work, etc.). Immigration could address some skills gaps or specific bottlenecks, but overseas immigration on a scale sufficient to change the demographic picture substantially is extremely unlikely and probably not feasible. That makes policies to boost fertility all the more important.

Above all, there is a need to press ahead with measures that help Japanese families to combine parenthood with child-care; otherwise, the recovery in fertility on which current strategies hinge may not be realised. This concerns in the first instance the position of women in the labour force, an issue with implications for gender equity as well as economic efficiency: the potential productivity gains that are forfeited when women are forced to choose between child-rearing and full participation on the labour market also entail significant costs for the women themselves. These costs go far beyond the wages forgone when their children are born. The evidence suggests that long periods outside the workforce or in non-regular employment result in lower wages over the rest of the working life. Human capital deteriorates during such spells, while opportunities for advancement are missed.22 The prospect of lower lifetime earnings, in turn, tends to depress the returns to investment in human capital and may thus encourage women to invest in skills less than they otherwise would.23

While some steps have been taken to address these challenges and some improvement has been observed in recent years (Figure 1.24), more needs to be done. The total fertility rate (TFR) is still extremely low, and it is not yet clear whether and to what extent the progress seen since 2005 reflects cohort effects:

  • The older cohorts of working-age women are, other things being equal, less likely to be employed than younger women, so as time passes, cohorts with lower labour force participation rates pass out of the working-age population, while the younger cohorts coming up behind them are more likely to be employed.

  • Demographic transitions in many countries have gone through a J-curve that reflects cohort effects. While the main shift was from high to low fertility, there was also a tendency of younger cohorts to start having children later. As Figure 1.24 illustrates, this transition led to a deep dip in the curve initially followed by a partial recovery.

Figure 1.24. Female labour force participation and the fertility rate

Source: OECD (2015h), OECD Economic Surveys: Japan 2015, OECD Publishing, Paris, based on data provided directly by the Ministry of Internal Affairs and Communication (MIC) and the Ministry of Health, Labour and Welfare (MHLW).

In light of these considerations, it is too soon to argue that the recovery in Japan’s fertility rates will continue, let alone how far it will go.

The authorities plan a significant expansion in child-care and after-school places for older children in the coming years, but measures are also needed to make the tax and benefit systems neutral with regard to work decisions by secondary earners in households: at present, there are significant incentives (e.g. high tax thresholds) for second earners to limit their earnings (OECD, 2015h). CJF (2014) argues for a weighted support system for birth, nursing and education that would give greater support for the third child and children born after that, somewhat similar to arrangements in place in France. However, it is also important to recognise that pro-fertility/pro-family policies are not concerned solely with women’s working conditions. It would also be advisable to move ahead with more flexible arrangements for fathers, such as proposals for paternity leave and related benefits. More broadly, as Cabinet Secretariat (2015) acknowledges, for men and women alike, there is a need to address the employment practices of firms and public-sector entities alike to reduce excessive working hours and improve work-life balance.

There is also considerable scope for local-level measures to make cities, in particular, more family-friendly, in terms of both urban form and the availability of childcare facilities, etc. These will be considered in Chapter 3.

…as well as innovation and entrepreneurship

Innovation and entrepreneurship have a critical role to play if Japan is to meet the productivity challenge outlined above. Yet its recent performance in these areas has been lacking:

  • Japan spends heavily on inputs to innovation, particularly research and development (R&D) and education, but the return on that investment in terms of total factor productivity (TFP) growth has been modest, at best (OECD, 2015a). Gross expenditure on R&D (GERD), at 3.4% of GDP in 2012, was the fifth-highest in the OECD, and more than three-quarters of this was financed and conducted by the business sector, which tends to have the greatest impact on TFP. Yet TFP growth, which averaged less than 0.5% per year during 1995-2010, has been far below the levels of countries with similarly high levels of GERD, such as Korea (over 1% average TFP growth over the period), Finland (above 1.5%) and Sweden (around 2%). Japanese companies remain world leaders in some critical high-tech sectors, such as robotics, automobiles and electronic components, but this has been insufficient to sustain better productivity growth overall.

  • Japan consistently scores poorly on measures of entrepreneurship. The entry and exit rate over 2004-09 was just about 4.5% on average, less than half the rate recorded in the United Kingdom and the United States (OECD, 2015h). The government’s Revitalisation Strategy aims to raise that figure to 10% in the coming years. On the 2015 Global Entrepreneurship Index (GEDI) Japan ranks 33rd of 130 countries rated and 26th among OECD countries, despite exceptional strength on some dimensions of entrepreneurship, such as technology absorption and process innovation. Significantly, Japan’s main weaknesses on the index’s different dimensions concern attitudes and skills rather than infrastructure or institutions.

These weaknesses are closely linked, of course, insofar as many innovations, particularly disruptive innovations, do not come from established large companies – they need entrepreneurs to bring them to the market.

OECD (2015h) canvasses a range of economy-wide reforms that can help to strengthen both innovation and entrepreneurship in Japan. These encompass reforms to framework conditions for innovation, measures to strengthen product-market competition, changes in the national science and technology system, steps to strengthen the role of venture capital-backed firms in both innovation and entrepreneurship, and reforms to stimulate greater dynamism in the small and medium enterprise (SME) sector. The current government has been increasingly active in this sphere, as well, taking steps to reorganise the central-level institutions responsible for innovation policy and introducing a range of new programmes and measures.24 There is still much to do, including: addressing skill shortages in science and engineering, enhancing entrepreneurship education, upgrading universities and strengthening their links with both domestic firms and international partners, attracting more foreign students to Japan and encouraging more Japanese to study abroad. Public support for R&D could also be reshaped in order to benefit small firms and start-ups rather more than at present: the current reliance on tax credits tends to benefit large companies more than SMEs, which are subject to a lower corporate tax rate, or start-ups, which often have no profit tax liabilities to offset. One key challenge will be to change attitudes to entrepreneurship, which is not widely regarded as a desirable career path, owing to a perceived cultural preference for lifetime employment in large companies and public-sector institutions, with seniority based wages and high levels of job security (Makinen, 2015).

These are economy-wide challenges, but as will be seen in Chapters 3 and 4, there is much that can and should be done by cities and regions to improve their performance. This is particularly true with respect to providing education in entrepreneurship. De Backer, K. and S. Miroudot (2013) reports that only 18% of Japanese believe that school had “helped to develop a sense of initiative and a sort of entrepreneurial attitude”. This was the lowest share reported for any OECD country and far below the OECD-wide average of 52%. Only 20% thought that school had provided them with the skills and know-how to run a business. The authorities are aware of this challenge, and the Ministry of Education, Culture, Sports, Science and Technology has recently stepped up its promotion of entrepreneurship education, but its programmes are primarily aimed at university students and graduates. A growing body of evidence suggests that such entrepreneurial skills can and should be fostered even in primary and secondary schools, with the emphasis not on business skills per se (which can indeed be acquired later) but on creativity, entrepreneurial know-how, responsibility, risk-taking, problem solving, and team-working (European Commission, 2013).

While OECD (2015h) emphasises the need for inter-ministerial co-operation, as well as the support of public institutions and the private sector, to promote entrepreneurial education, this is a field in which there is great scope for regional and local action and experimentation, particularly with respect to the design and delivery of curricula, an issue to be explored in the chapters to come. Strengthening links among key institutions in the innovation process, particularly universities and businesses, can also benefit from a place-based approach, as the experience of many OECD countries attests (OECD, 2011).

A shrinking Japan, seeking to regain its economic dynamism, cannot afford to turn inward

The period of fast growth in Japan was founded on its export dynamism. The country was extremely successful in integrating itself into the post-war world economy. Its prosperity today derives in no small measure from its central position in some key global value chains (GVCs), particularly autos and electronics. Its estimated GVC participation index in 2009 was close to that of Germany and above those of other large economies, such as Canada, France, Italy, the United Kingdom and the United States (De Backer and Miroudot, 2013).25 Japan’s central role in GVCs as a producer of higher value parts and components used in industries across the globe was a major reason for the speed with which the economic consequences of the 2011 Great East Japan Earthquake were felt around the world. Japanese firms reported production slowdowns in their affiliates abroad and slowdowns also occurred in some foreign industries like automotive and electronics that rely heavily on inputs from Japan. The earthquake and tsunami disrupted the activities of both direct and indirect suppliers, leading in some cases to the complete disruption of international supply chains (De Backer and Miroudot, 2013).

Nevertheless, in important respects, Japan has failed to keep pace with globalisation in recent decades (Cabinet Office, 2011). Its import penetration ratio is well below the OECD average, thanks in large measure to significant non-tariff barriers, notably in the form of product-market regulation.26 Its level of intra-industry trade is also exceptionally low compared to other large economies, and while it has been an active investor overseas, the stock of inward foreign direct investment (FDI) relative to GDP was the lowest among OECD economies in 2012. It also hosts the lowest stock of foreign workers as a share of the labour force among OECD countries for which data are available (OECD, 2012b). This is not just a question of money: foreign investors and workers bring ideas, as well as money, and FDI-induced spillovers can play a significant role in stimulating technology transfer and innovation (Ewe-Ghee, 2001, and Savvides and Zachariadis, 2005). Japan’s agriculture and services sectors are far less integrated into GVCs than its powerhouse manufacturing industries (De Backer and Miroudot, 2013), and this must be a concern given the growing role of services in both the Japanese economy and the growth of international trade in services. Japan’s share of world exports fell from almost 10% in 1992 to only 4% in 2012, while its share of high-tech exports from OECD economies dropped by roughly half (OECD, 2015h).

Deeper internationalisation is critical to Japan’s future prosperity. Access to imported inputs and inflows of FDI and expertise can reduce costs and also enhance productivity through technology spillovers. Lowering non-tariff measures would better enable Japanese firms to access potential efficiency gains and opportunities for innovation by further developing global supply chains. Jones and Yoon (2006) emphasise the need for a comprehensive approach, involving changes in a number of different policies, and they identify four broad priorities: i) reducing barriers to FDI and imports through multilateral trade negotiations and regional trade agreements; ii) relaxing product market regulations, notably in the service sector; iii) fully opening the market for corporate control (mergers and acquisitions) to foreign firms; and iv) relaxing restrictions on the inflow of foreign workers, including those in non-technical occupations.

The government recognises the need to address these issues and is strongly committed to increasing Japan’s integration into the world economy, not least in aiming to double the flow of inward FDI into Japan over the coming years (Cabinet Secretariat, 2015). In two areas, at least, the government’s internationalisation priorities have important regional and local dimensions: the internationalisation of SMEs, which are the backbone of many local economies in Japan, and the internationalisation of education.

SMEs, which account for over 99% of firms, 74% of employment and over 50% of value added in Japan, rely almost exclusively on domestic demand. Only around 0.2% of them are exporters, and exports typically constitute just 7% of sales in manufacturing SMEs, as opposed to 28% for larger firms (EIU, 2010). Reliance on domestic demand leaves Japanese SMEs extremely vulnerable to domestic economic and demographic conditions: the domestic market is, after all, shrinking. The SME sector thus needs to become more internationalised. SMEs were long treated by policy as a disadvantaged group requiring generous state support – SME policy was, in many respects, social policy – but they have been defined as a source of growth since 1999. The Revitalisation Strategy emphasises accelerated restructuring of the SME sector, and measures to help SMEs export and expand overseas are central to that effort (SMEA, 2014). Regional and local efforts to mobilise and support the internationalisation of SMEs are needed to supplement these efforts; the potential for action in cities and regions will be explored in Chapters 3 and 4.

Education is further field in which Japan can do more to reap the benefits of globalisation. The Institute of International Education estimates the share of foreign students in Japanese universities at only 3.8%, in 2013, less than half of the OECD average of 8%, and the number of branch campuses of foreign universities in Japan fell from around 40 in the early 1990s to 4 at present (OECD, 2015h). The flow of Japanese students abroad has likewise been limited in recent years, although the trend has changed. According to IIE (2014), the number of Japanese students in overseas higher education institutions fell from just over 82 000 in 2004 to a low of around 40 000 in 2010 (just over 1% of the total). By 2013, however, the number had risen more than 50% to over 69 000 (Ministry of Education, Culture, Sports, Science and Technology, 2015) – still less than 2% of the total, but a dramatic rise over three years, particularly against the backdrop of a 6.8% drop in total student numbers. These modest numbers are particularly striking when set alongside a UNESCO (2012) estimate that international student mobility rose by 78% in 2000-10. Further internationalisation of education will help Japan to boost performance and develop globally-competitive talent.

While the recovery in the number of students going abroad is encouraging for Japan’s future, the numbers of foreign students coming to Japan are in some ways more striking. Once seen as a “source” country for students seeking to study abroad, Japan has in recent years been trying to reposition itself as a study abroad destination, part of the government’s larger drive to promote the internationalisation of Japanese higher education. The goal of its Global 30 programme is to bring in 300 000 international students by 2020; the government has also set a target of 120 000 Japanese students studying abroad in the same year. There are some encouraging signs on this front. Six Japanese universities have entered the top 100 of the QS World University Rankings (for 2014/15), up from 4 in 2009, and foreign student numbers have begun rising again following the disruption caused by the Great East Japan Earthquake. In a 2012 roundtable discussion of ways to attract foreign students, Japanese experts identified a number of issues, including:

  • a lack of pathways for students from other Asian countries to come study and work in Japan

  • the lack of an official “foreign student support system”

  • inadequate government, industry and private sector collaboration with respect to the recruitment of students post-study to the workforce, and

  • a perceived lack of confidence when it comes to promoting the country.

Much is being done to address these shortcomings, including the government’s “skilled migration approach,” which promotes the employment of international students in Japan after their studies, efforts to expand the range of English-taught courses in Japanese universities and more aggressive recruitment efforts in Asian countries, particularly China. Moreover, the government has also committed to financial support for universities that expand their study-abroad programmes.27 There are also plans to provide scholarships to help high school graduates take part in short-term overseas study programmes prior to beginning their university studies in Japan. There is considerable scope for regions and cities to take advantage of these opportunities, particularly in promoting themselves as potential destinations for study. While many foreign students will doubtless gravitate to Tokyo, such factors as cost of living, environmental quality, historical and cultural amenities, and strong academic specialisations should enable some regions and cities to promote themselves effectively. Potential actions in this sphere will be considered in the chapters that follow.

Internationalisation is particularly important for Japan’s innovation systems

Strengthening international business links and the internationalisation of education should do much to facilitate the deeper global integration of Japan’s innovation system. Recent years have seen a growing trend towards “open innovation in global networks”, in which firms collaborate with external partners at home and abroad (OECD, 2008). This trend has largely passed Japan by. Japanese firms have not embraced open innovation to the same extent as their foreign peers, reflecting concerns about losing technology to competitors (Motohashi, 2013). As OECD (2015h) shows, the result is that Japan lags other OECD countries on a range of indicators when it comes to cross-border innovation and knowledge creation activities. This is clearly reflected in the relatively weak innovation connections between Japanese regions and foreign partners. Japan’s most innovative regions are not so deeply integrated into global innovation networks, and the trend between 1995 and 2010 was towards decreasing international collaboration (Figure 1.25). OECD (2011) finds a couple of Japan’s leading regions connected to the periphery of international green patenting networks, but overall integration was low. In 2008-10, all Japanese regions ranked below the OECD median in terms of the share of co-patents filed with foreign co-inventors. Indeed, only 28.1% of Japanese patents filed during the period involved co-inventors in more than 1 prefecture, and only 2.7% involved foreign co-inventors. The corresponding figures for all OECD TL3 regions were 34.4% and 17.8%, respectively. While Japan is home to some strong regional innovation systems, they appear to remain very Japan-centred and even, to a surprising degree, disconnected from each other. They have much to offer the rest of the world and much to gain from it, and efforts to integrate Japan’s innovative regions into global networks will be crucial in the years ahead.

Figure 1.25. Regional patterns of co-patenting
Per cent of co-patents (X axis) and per cent of foreign collaborations (Y axis) in the top 40 regions with the highest patent applications, 2008-10 and compared to their values in 1995-97

Note: Average number of co-patent applications over the period 2008-2010. The number of foreign co-inventors is defined as the number of co-inventors that reside/work in a TL region outside national borders.

Source: OECD (2011), Regions and Innovation Policy, OECD Reviews of Regional Innovation, OECD Publishing, Paris,

The spatial consequences of Japan’s restructuring challenge

The spatial effects of demographic change are difficult to anticipate

Predicting the spatial consequences of demographic change in Japan is difficult, since no country has ever experienced what is now unfolding there. There is widespread fear, in particular, of an over-concentration of population in the main urban centres, leaving rural areas and smaller towns and cities under-populated, aged and impoverished. Indeed, there are already policies in place that are explicitly aimed at reducing the population growth of Greater Tokyo, in particular (see below). The concentration of economic activity and population in the centre of the country is unusually great for a country of Japan’s size and it appears on some indicators to have been increasing in recent years. At present, the Ministry of Land, Infrastructure, Transport and Tourism (MLIT) anticipates that concentration in Tokyo will increase further, with Greater Tokyo being the only major region of the country expected to gain population between now and 2020, after which it, too, will begin to lose population. The capital region is currently experiencing net inward migration flows of about 100 000 per year, with 15-29 year-olds constituting the vast majority of new arrivals.

This is of some concern to the authorities, not least because Tokyo has by far the lowest fertility rate in Japan (1.13 in 2013, as against 1.43 nationally). And Tokyo is just the most extreme case: more generally, young adults appear to be moving to the major cities, where fertility rates overall tend to be lower. Low metropolitan fertility is generally seen as linked to high living costs, lack of affordable housing and the pressures of work, but it may also be linked to gender mismatches. Ishikawa (2011) cites data showing that migrant flows from the periphery to the major metropolitan areas are mostly young men, whereas young women who move tend to migrate to larger cities within their prefectures of origin. Few women move from prefecture to prefecture. The result is a complicated pattern of spatial gender imbalances in origin and destination areas. Male-female ratios, especially among young adults, are very high in eastern Japan but much lower in the western regions. Within prefectures, though, they tend to be higher in mountainous and farming areas and lower in urban places, including prefectural capitals. The data suggest that these imbalances are indeed related to lower marriage rates.

Policy makers face a dilemma with respect to the degree of concentration that may be desirable

Closely related to this is the question of whether and how the government should attempt to shape the settlement pattern as depopulation proceeds. The productivity imperative would seem to point to a need for greater concentration in an effort to realise the potential productivity benefits of agglomeration, achieve economies of scale in infrastructure and service provision and sustain the global competitiveness of Japan’s major cities, particularly the Tokyo-Nagoya-Osaka urban mega-region. Firms in regions that lose a lot of population – and especially places where population decline coincides with rapid ageing – will confront smaller and less resilient labour markets, as well as problems in finding efficient suppliers and generating links with other businesses. Low-population locations will struggle to attain sufficient critical mass to provide basic public and private goods efficiently, since the fixed costs will have to be borne by a smaller number of clients and consumers (Sáez, Ayuda and Pinilla, 2011).

The productivity-focused logic of concentration, however, must be balanced against the need for a sustainable settlement pattern. The authorities are rightly concerned about the environmental deterioration that can occur in abandoned locations, as well as about the equity implications of leaving a substantial portion of the population living in places where depopulation trends may destroy the economic and social fabric of local communities. There is also a fear that over-concentration could leave the country even more vulnerable to both economic shocks and natural catastrophes. This is a particularly real concern in view of Tokyo’s vulnerability to earthquakes; many observers believe the capital will experience a major quake in the coming decades. Such fears have been reinforced by analyses such as that contained in the so-called Masuda report, an analysis of the spatial and economic consequences of demographic change prepared by Hiroya Masuda, former Governor of Iwate Prefecture and former Minister of Internal Affairs and Communications. The report, prepared for the Japan Policy Council, created a sensation with its stark warning that 896 local governments – roughly half the total - risked “extinction” by 2040 (see Masuda, 2014, for an English-language summary).

For these reasons, policy makers have recently put increasing emphasis on the revitalisation of “non-metropolitan Japan”, including both rural areas and small/medium towns and cities. Many such areas have struggled for decades, as growing population concentration led to depopulation even before fertility dropped to very low levels. In an environment of declining total population, the competition for people is even more acute and their position is even more vulnerable. The erosion of a small community’s human capital may quickly become self-reinforcing, if promising individuals leave because they see that the departure of others with high human capital reduces the opportunities available locally. This, in turn, can have a knock-on effect on asset prices, prompting disinvestment by asset owners keen to exit before prices drop too far. This redistribution mechanism has evident effects on the generation of local fiscal revenues and thus on the investment capacities of local authorities, which add to the risk of a sustained downward spiral. Moreover, the experience of some countries suggests that “winding down” remote communities and relocating their inhabitants closer to facilities and economic activities can be expensive (OECD, 2010b). In any case, states are territorial entities, and most are to a greater or lesser extent concerned with the occupation and exploitation of the national territory, for security or other reasons. Few governments view the depopulation of large swathes of their territory with indifference.

The tension between the logic of concentration and the need for a sustainable settlement pattern cannot be definitively resolved – it must be managed. The authorities are right to be concerned with the global competitiveness of the country’s major cities. However, it would be very unwise to neglect the potential of non-metropolitan Japan. Pursuing both these goals with a shrinking population and over-stretched public finances require some trade-offs, but it is difficult to imagine abandoning either of them. Chapters 2 and 4 of this review will look closely at the kinds of policies that might support these objectives without creating unnecessary contradictions or making excessive demands on fiscal resources.

However, the risk of hyper-concentration may not be as great as it sometimes appears

The policy dilemma outlined above may nevertheless be somewhat exaggerated, insofar as there is no a priori reason to believe that Japan’s settlement pattern will actually be far more concentrated in a shrinking society than in a growing one. There will be fewer settlements and they will, on average, be smaller – that is an inevitable consequence of such large-scale population decline – but it is not at all clear why one might expect that population concentration will increase sharply. Indeed, one of the striking features of urban systems is the persistence observed in the structure of urban systems over time, even in the face of extremely powerful shocks. While the degree of spatial differentiation (concentration) has increased gradually over time, as cities have grown larger, the relative positions of cities in the distribution of city sizes are remarkably stable, owing to a combination of locational fundamentals and path dependence. The spatial structure of economies passing through economic revolutions (agricultural, industrial, informational), as well as major wars and other political upheavals, has been surprisingly persistent, and Japan is no exception (Davis and Weinstein, 2002). It is not clear why the demographic transition should lead to some more radical break with the past.

Despite varying degrees of population concentration across countries, national urban systems typically encompass cities across a range of sizes. Variety is beneficial in the size and character of places no less than in the availability of goods and services. Firms and households choose cities of different sizes as a function of their needs and resources, and there is no obvious reason for depopulation to change this:

  • In terms of functions, urban and rural activities tend to be complements, not substitutes. This is true with respect to the consumption and amenities opportunities they offer one another, as well as their production profiles. Cities need rural areas, and rural areas need cities. Moreover, because relatively small-scale successes can make a big difference locally in a rural setting, it is likely that many rural communities can and will find paths to a prosperous and sustainable future, even if they are smaller than before. This issue is considered in detail in Chapter 3.

  • Cities, by contrast, compete with one another, and larger cities typically enjoy important advantages when competing with respect to many urban functions. This suggests that the most difficult challenges may face not the rural areas but the small and medium cities. And since large cities are themselves likely to shrink, the suburban hinterlands of the major urban areas may likewise face particular problems.

The city-size distribution in most places tends to conform to the rank-size relationship known as Zipf’s law (Box 1.3), at least for cities above a certain size (it tends to break down at very small scales). Japan’s urban system is indeed a bit more concentrated at the top than Zipf’s law would predict – i.e., Tokyo, Osaka and Nagoya are indeed larger than one would expect (Figure 1.26)28 However, the difference is not dramatic and there was little change in urban concentration between the 2000 and 2010 censuses, as the population peaked and started to decline. Zipf’s law is merely an empirical observation: it has no obvious normative implications and it does not constitute the basis of for any policy recommendations. However, unless there are good reasons to suspect that population decline will lead to major deviations from Zipf’s law (for example, if it can be shown that the elderly population really will prefer to concentrate in large cities), there is little reason to imagine that Japan will experience a further dramatic increase in population concentration. The challenge, then, will not be preventing a huge influx into Greater Tokyo but devising policies to improve the performance and viability of cities of all sizes, as well as rural areas, in order to sustain growth and well-being.

Box 1.3. Zipf’s law and the urban hierarchy

In the context of urban studies, the term Zipf’s law refers to an empirical regularity concerning city-size distributions that has been observed and debated for over a century (Auerbach, 1913; Zipf, 1949): the population ranks of cities in various countries follow a power law of a specific type such that, under the hypothesis of a Pareto probability distribution, the log(rank)-log(size) relationship is linear, with a coefficient equal or close to -1. Put more simply, this implies that the largest city is twice as large as the second largest city, three times as large as the third and so on along the urban hierarchy. While the relationship tends to break down at very small scales, it holds remarkably well for many countries across a very wide range of city sizes (Gabaix and Ioannides 2004). The relevance of Zipf’s law in the context of city-size distribution is twofold. First, it relates to efforts to understand the distribution of population and human activity across space; Krugman (1996: 40) has argued that so stable regularity is “spooky” and that there should be a theoretical explanation for it. There is also the question of whether Zipf’s law implies some constraints in the pattern of urban growth –i.e., that the growth trajectories of individual cities could not change the overall city-size distribution (Duranton, 2007). Others raise the question of whether there are different levels of economic efficiency for different shapes of the urban systems (number of cities and their sizes) (Storper, 2013).

Sources: P. Veneri (2013), “On City Size Distribution: Evidence from OECD Functional Urban Areas”, OECD Regional Development Working Papers, 2013/27, OECD Publishing, Paris,; F. Auerbach (1913), “Das Gesetz der Bevölkerungskonzentration”, Petermanns Geographische Mitteilungen 59, pp. 74-76. Gabaix, X. and Y.M. Ioannides (2004), “The evolution of city distributions”, in J.V. Henderson and J.F. Thisse (eds.) Handbook of Regional and Urban Economics, Ch. 53, North Holland, Amsterdam, pp. 2 341-2 378; G. Duranton (2007), “Urban evolutions: the fast, the slow, and the still”, American Economic Review, Vol. 97/1, pp. 197-221; Krugman, P. (1996), The Self-Organizing Economy, Blackwell, Cambridge, MA; Storper, M. (2013), Keys to the City. How Economics, Institutions, Social Interaction, and Politics Shape Development, Princeton University Press, Oxford; Zipf, G. (1949), Human Behavior and the Principle of Least Effort, Addison-Wesley, Cambridge, MA.

Figure 1.26. City-size distribution in Japan

Note: City sizes are based on functional urban areas defined in the OECD Metropolitan database rather than on the basis of political-administrative frontiers; Veneri (2013) shows in a cross-national study of OECD metros that functional cities conform more closely to Zipf’s Law than administrative cities.

Source: OECD (2015e), “Metropolitan areas”, OECD Regional Statistics (database), (accessed 10 September 2015).

Ageing might be the one factor that leads to a sharp deviation from the traditional pattern of city-size distribution. It may be that the changing structure of the population will drive increased concentration in large metropolitan areas, above and beyond what one would otherwise expect. Indeed, this hypothesis underlies many recent initiatives, which are based on the expectation that the elderly will (or, in some cases, should) tend to prefer more compact urban living (OECD, 2015c). However, population ageing has been under way in many OECD countries for decades and has not triggered a large-scale shift in the elderly population to the cities. In many cases, rural and suburban areas are ageing faster. Moreover, in Japan as elsewhere, there are significant numbers of older people moving out of urban areas. The typical life-cycle migration pattern in some countries sees young singles moving to dense urban areas in search of jobs and urban amenities, before moving to the suburbs as they have children and then even further out of town as they retire (Detang-Dessendre, Goffette-Nagot and Piguet, 2004; Johnson, Winkler and Rogers, 2013). While this is not such a typical Japanese pattern, Ishikawa (2011) observes that people in late middle age and early retirement age (especially Baby-Boomers) are particularly prominent among reverse (centre to periphery) migrants. He estimates net migration of the 60-64 age group to the periphery at around 50 000 in 2010 – still relatively small as a proportion of the relevant cohorts, at least when compared with other OECD countries. By contrast, later-stage elderly (75+) are more likely to return to the metropolitan areas.

The policy challenges ahead are enormous and will need a multi-dimensional response

The foregoing suggests that some of the most extreme scenarios look very improbable – the country is unlikely just to contract “inwards” towards the great urban mega region in the centre of Honshū. However, the spatial consequences of Japan’s demographic shift will still be painful for many places. The country will in future have fewer and smaller settlements. Some will downsize; others may disappear. The baseline assumption, in the absence of strong reasons to expect the contrary, should be that the settlement pattern will remain rather diverse. It is also critical to bear in mind that in some places, downsizing may be the key to prosperity rather than barrier to it: in a number of OECD countries, struggling rural regions have found new sources of growth in less labour-intensive activities (OECD, 2010b). They do not lack the potential for a prosperous future, but that future will be based on fewer people living there. While such restructuring and downsizing can be painful, the long-term goal must surely be to keep people prosperous, not necessarily to keep places populated.

Demographic change will entail important policy responses and significant costs – adapting infrastructure, altering service provision and, in some cases, encouraging or even facilitating the relocation of some populations. Ensuring continued prosperity in such a context will require a co-ordinated approach encompassing spatial planning at various scales from the national to the local, infrastructure and service provision, regulatory reform, and policies for the promotion of innovation and entrepreneurship. The chapters to come will explore these in various dimensions. First, the discussion will look at the national context, focusing on the role of national-level spatial planning and territorial governance (Chapter 2). Then it will turn to the challenges of ensuring the continued dynamism and global competitiveness of Japan’s major cities (Chapter 3) before addressing the challenges associated with regional revitalisation – policies critical to sustaining the prosperity and even viability of many smaller cities and rural areas (Chapter 4).


Accenture (2012), The Seven Myths of Population Aging, Accenture Institute for High Performance.

Asai, F. (2015), “Panel: Old People in Tokyo Should Move to Rural Regions for Nursing Care”, Asahi Shimbun, 5 June.

Auerbach, F. (1913), “Das Gesetz der Bevölkerungskonzentration”, Petermanns Geographische Mitteilungen, No. 59, pp. 74-76.

Botham, R. and A. Graves (2009), The Grey Economy: How Third Age Entrepreneurs Are Contributing to Growth, NESTA Research Report, National Endowment for Science, Technology and the Arts, London, August.

Brezzi, M., L. Dijkstra and V. Ruiz (2011), “OECD Extended Regional Typology: The Economic Performance of Remote Rural Regions”, OECD Regional Development Working Papers, 2011/06, OECD Publishing, Paris

Cabinet Office (2011), New ‘Opening’ of Japan and Innovation, Cabinet Office Economic Report, Tokyo, available at:

Cabinet Secretariat (2015), “Regional Empowerment for Japan’s Growth: Overcoming Population Decline and Revitalizing Local Economies: Japan’s Long-term Vision and Comprehensive Strategy”, Headquarters for Overcoming Population Decline and Vitalizing Local Economy in Japan, Cabinet Secretariat, Tokyo.

CJF (2014), “Choice for the Future: Creating a Growth and Development Model Made in Japan: Transcending Rapid Depopulation and Super Aged Society”, Committee for Japan’s Future, Council on Economic and Fiscal Policy Expert Panel, Tokyo, May.

Colovic, A. and O. Lamotte (2012), “Entrepreneurs seniors et innovation: Une étude sur données individuelles”, Revue Française de Gestion, No. 227, pp. 127-41.

Davis, D. and D. Weinstein (2002), “Bones, Bombs, and Break Points: The Geography of Economic Activity”, The American Economic Review, Vol. 92/5, December, pp. 1 270-1 289.

Davies, R. and G. Pierre (2005), “The Family Gap in Pay in Europe: A Cross-Country Study”, Labour Economics, Vol. 12/4, August.

De Backer, K. and S. Miroudot (2013), “Mapping Global Value Chains”, OECD Trade Policy Papers, No. 159, OECD Publishing, Paris,

Dembe, A. et al. (2005), “The Impact of Overtime and Long Work Hours on Occupational Injuries and Illnesses: New Evidence from the United States”, Occupational and Environmental Medicine, Vol. 62/9, September, pp. 588-597.

Detang-Dessendre, C., F. Goffette-Nagot and V. Piguet (2004), “Life-cycle position and migration to urban and rural areas: estimations of a mixed logit model on French data”, Working Paper W.P., Vol. 04/03, GATE Groupe d’Analyse et de Théorie Économique, UMR 5824 du CNRS, April.

Duranton, G. (2007), “Urban evolutions: the fast, the slow, and the still”, American Economic Review, Vol. 97/1, pp. 197-221.

EIU (2010), SMEs in Japan: A New Growth Driver?, The Economist Intelligence Unit, London.

European Commission (2013), Entrepreneurship in the EU and beyond - Flash Eurobarometer 354, Report, available at:

Ewe-Ghee, L. (2001), “Determinants of, and the Relationship Between, Foreign Direct Investment and Growth: A Summary of the Recent Literature”, IMF Working Paper, No. 2001/175.

Fujimoto, T. (2013a), “The Future of the Japanese Manufacturing Industry: Overall Hollowing Out Will Not Happen”, Japan Spotlight, May/June.

Fujimoto, T. (2013b), “Innovation Management in Japan”, in M. Dodgson, D. Ganns and N. Phillips (eds.), The Oxford Handbook of Innovation Management, Oxford University Press, Oxford.

Gabaix, X. and Y.M. Ioannides (2004), “The Evolution of City Distributions”, in J.V. Henderson and J.F. Thisse (eds.) Handbook of Regional and Urban Economics, Ch. 53, North Holland, Amsterdam, pp. 2 341-2 378.

Gray, M. and B. Chapman (2001), “Forgone Earnings from Child Rearing: Changes Between 1986 and 1997”, Family Matters, Vol. 58, Autumn.

Haga, K. (2014), “Innovation and Entrepreneurship in Aging Societies: Theoretical Reflection and a Case Study from Kamikatsu, Japan”, Paper presented to Innovation Forum VI “Crisis, Innovation and Transition”, University of Paris-Ouest, Nanterre-La Défense, 1-3 October.

Hoornbeek, J. and T. Schwarz (2009), “Sustainable Infrastructure in Shrinking Cities: Options for the Future”, Center for Public Administration and Public Policy, Kent State University, Kent, OH, July.

ICEF (2014), “Japan’s Globalisation Efforts Blossoming”, ICEF Monitor, 25 April 2012, (accessed 15 July 2015).

IIE (2014), “Project Atlas: Japan”, Institute of International Education, New York, (accessed 15 July 2015).

Ishikawa, Y. (2011), “Recent In-Migration to Peripheral Regions of Japan in the Context of Incipient National Population Decline”, in F. Coulmas and R. Lützeler (eds.), Imploding Populations in Japan and Germany: A Comparison, Koninklijke Brill, Leiden.

Izumi, N. (2015), “Japan’s Policies for Small and Medium Enterprises”, Ministry of Economy, Trade and Industry and Embassy of Japan in South Africa, March.

Johnson, K., R. Winkler and L. Rogers (2013), “Age and Lifecycle Patterns Driving U.S. Migration Shifts”, Carsey Institute Issue Brief, No. 62, Spring.

Jones, B. (2008), “Age and Great Inventions”, NBER Working Paper, No. 11 359, April.

Jones, R. and T. Yoon (2006), “Strengthening the Integration of Japan in the World Economy to Benefit more Fully from Globalisation”, OECD Economics Department Working Paper, OECD Publishing, Paris,

Joshi, H. and H. Davies (1992), “Child Care and Mothers’ Lifetime Earnings: Some European Contrasts”, CEPR Discussion Papers, No. 600, January.

Joshi, H. (1990), “The Cash Opportunity Costs of Childbearing: An Approach to Estimation Using British Data”, Population Studies, Vol. 44/1, January.

Joshi, H. and H. Davies (2002), “Women’s Incomes over a Synthetic Lifetime”, in Ruspini, E. and A. Dale (eds.), The Gender Dimension of Social Change, The Policy Press, Bristol, pp. 111-131.

Kashiwazaki, C. and T. Akaha (2006), “Japanese Immigration Policy: Responding to Conflicting Pressures”, Migration Policy Institute, 1 November, (accessed 15 July 2015).

Kautonen, T. (2013), “Senior Entrepreneurship”, Background Paper Prepared for the OECD Centre for Entrepreneurship, SMEs and Local Development, University of Turku, Turku, Finland.

Kono, S. (2011), “Confronting the Demographic Trilemma of Low Fertility, Ageing and Depopulation”, in Coulmas F. and R. Lützeler (eds.), Imploding Populations in Japan and Germany: A Comparison, Koninklijke Brill, Leiden.

Krugman, P. (1996), The Self-Organizing Economy, Blackwell, Cambridge, MA.

Lévesque, M. and M. Minniti (2006), ‘The Effect of Aging on Entrepreneurial Behavior’, Journal of Business Venturing, Vol. 21/2, pp. 177-194.

Makinen, J. (2015), “A Subculture of Entrepreneurship Hatches in Japan”, Los Angeles Times, 29 March.

Masuda, H. (2014), “The Death of Regional Cities: A Horrendous Simulation”, Japan Foreign Policy Forum, No. 18, available at:

Matsuyama, K. (2015), “Tokyo’s Elderly Turned Away as Nursing Homes Face Aid Cuts”, Bloomberg Business, 19 February.

Ministry of Education, Culture, Sports, Science and Technology (2015), “Japanese Study Abroad Situation”, Ministry of Education, Culture, Sports, Science and Technology, Tokyo, February, available (in Japanese) at:

Ministry of Land, Infrastructure, Transport and Tourism (2014), “Grand Design of National Spatial Development Towards 2050: Provisional Translation”, Ministry of Land, Infrastructure, Transport and Tourism, Tokyo, 4 July, available at:

Motohashi, K. (2013), “Innovation Policy Challenges for Japan: An Open and Global Strategy”, Asie Visions, No. 45, IFRI Center for Asian Studies, Paris.

Odagiri, T. (2015), “The Masuda Report Warns of the Disappearance of Country Life”, Opinion, No. 58, Meiji University, 20 March.

Odagiri, T. (2012), “Rural Regeneration in Japan”, Centre for Rural Economy Research Report, January.

OECD (2015a), Historical population data and projections (1950-2050),, (accessed 16 October 2015).

OECD (2015b), “Population”,, (accessed 16 October 2015).

OECD (2015c), Ageing in Cities, OECD Publishing, Paris,

OECD (2015d), Regional Statistics (database), http://dx.doi/org/10.1787/region-data-en (accessed 10 September 2015).

OECD (2015e), “Metropolitan areas”, OECD Regional Statistics (database), (accessed 10 September 2015).

OECD (2015f), Governing the City, OECD Publishing, Paris,

OECD (2015g), “Regional well-being”, OECD Regional Statistics (database),

OECD (2015h), OECD Economic Surveys: Japan 2015, OECD Publishing, Paris,

OECD (2015i), OECD Compendium of Productivity Indicators 2015, OECD Publishing, Paris,

OECD (2015j), OECD Economic Outlook: Statistics and Projections, (accessed 16 October 2015).

OECD (2014), Japan: Advancing the Third Arrow for a Resilient Economy and Inclusive Growth, Better Policies, OECD Publishing, Paris,

OECD (2013a), Regions at a Glance 2013, OECD Publishing, Paris,

OECD (2013b), How’s Life? 2013: Measuring Well-Being, OECD Publishing, Paris,

OECD (2013c), OECD Economic Surveys: Japan 2013, OECD Publishing, Paris,

OECD (2013d), Entrepreneurship at a Glance 2013, OECD Publishing, Paris,

OECD/European Commission (2013), A Good Life in Old Age? Monitoring and Improving Quality in Long-Term Care, OECD Health Policy Studies, OECD Publishing, Paris,

OECD (2012a), Redefining “Urban”: A New Way to Measure Metropolitan Areas, OECD Publishing, Paris,

OECD (2012b), Policies for a Revitalisation of Japan, Better Policies, OECD Publishing, Paris,

OECD (2011), Regions and Innovation Policy, OECD Reviews of Regional Innovation, OECD Publishing, Paris,

OECD (2010a), Strategies to Improve Rural Service Delivery, OECD Publishing, Paris,

OECD (2010b), OECD Rural Policy Reviews: Québec, Canada, OECD Publishing, Paris,

OECD (2008), Open Innovation in Global Networks, OECD Publishing, Paris,

OECD (2007), Babies and Bosses - Reconciling Work and Family Life: A Synthesis of Findings for OECD Countries, OECD Publishing, Paris,

OECD (n.d.), Regional Well-being website, (accessed 1 September 2015).

Oliveira Martins, J. et al. (2005), “The Impact of Ageing on Demand, Factor Markets and Growth”, OECD Economics Department Working Papers, No. 420,

Pencavel, J. (2014), “The Productivity of Working Hours”, IZA Discussion Paper, No. 8 129, April, available at:

Rho, H. (2010), “Hard Work? Patterns in Physically Demanding Labor among Older Workers”, Center for Economic and Policy Research, Washington, D.C., August.

Ricci, J. et al. (2007), “Fatigue in the U.S. Workforce: Prevalence and Implications for Lost Productive Work Time”, Journal of Occupational and Environmental Medicine, Vol. 49/1, pp. 1-10.

Roundtable (2012), “Round table on attracting foreign students”, The Japan Times, (accessed 16 July 2015).

Sáez, L., M. Ayuda and V. Pinilla (2011), “Public Intervention against Depopulation as a Local Policy: Justifications from Spain”, Documentos de Trabajo 2011-6, Universidad de Zaragoza Department of Applied Economics and Economic History, Zaragoza.

Savvides, A. and M. Zachariadis (2005), “International Technology Diffusion and the Growth of TFP in the Manufacturing Sector of Developing Economies”, Review of Development Economics, Vol. 9/4, November.

SMEA (2014), 2014 White Paper on Small and Medium Enterprises in Japan, Small and Medium Enterprise Agency, Tokyo.

Stangler, D. and M. Spulber (2013), “The Age of the Entrepreneur: Demographics and Entrepreneurship”, Paper for the I4J Summit, Los Angeles, March.

Storper, M. (2013), Keys to the City. How Economics, Institutions, Social Interaction, and Politics Shape Development, Princeton University Press, Oxford.

Sutherland, D., P. Hoeller and R. Merola (2012), “Fiscal Consolidation: Part 1. How Much is Needed and How to Reduce Debt to a Prudent Level?”, OECD Economics Department Working Papers, No. 932, OECD Paris,

UNESCO (2012), “Global Flow of Tertiary-level Students”, Global Education Digest 2011, available at:

Veneri, P. (2013), “On City Size Distribution: Evidence from OECD Functional Urban Areas”, OECD Regional Development Working Papers, 2013/27, OECD Publishing, Paris,

Wadhwa, V. et al. (2009), The Anatomy of an Entrepreneur: Making of a Successful Entrepreneur, Ewing Marion Kauffman Foundation, Kansas City, Mo., November.

Wadhwa, V., R. Freeman and B. Rissing (2008), Education and Tech Entrepreneurship, Ewing Marion Kauffman Foundation, Kansas City, Mo., May.

Zipf, G. (1949), Human Behavior and the Principle of Least Effort, Addison-Wesley, Cambridge, MA.


← 1. The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli authorities. The use of such data by the OECD is without prejudice to the status of the Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of international law.

← 2. Defined as the ratio of elderly to working-age people.

← 3. This contrasts with a figure of 15.5% for those OECD countries for which 2013 data are available. In 2011, the last year for which data on all OECD countries are available, Japan’s figure was 23.3% and the OECD-wide share was 14.9%.

← 4. Cabinet Secretariat (2015) projects that, even with a rise in the TFR to 1.8 in 2030 and 2.07 by 2040, the population would be not much above 100 million in 2050 and would fall to around 90 million by the end of the century.

← 5. Japan experienced substantial inflows of foreign workers from the late Meiji period until the Second World War. These consisted mainly of Chinese, who for a time formed substantial communities in the major port cities, and Koreans who entered Japan between 1910 and 1945, many of them conscripted labourers brought over during the war. After 1945, over 500 000 Koreans and much smaller numbers of Taiwanese and mainland Chinese remained in Japan. These and their descendants constituted the vast majority of foreign residents of Japan until well into the 1980s.

← 6. In 1970, South Korea recorded a TFR of 4.54, Chinese Taipei 4.00 and the People’s Republic of China (hereafter simply “China”) 5.44. By 2007 or so, Korea was at 1.26, Chinese Taipei at 1.10 and China, thanks to its one-child policy, at 1.6.

← 7. The Geographic Concentration index of population is defined as: where picture pi is the population share of region i, ai is the area of region i as a percentage of the country area, N stands for the number of regions and | | indicates the absolute value. The index lies between 0 (no concentration) and 1 (maximum concentration) in all countries.

← 8. The increase in concentration of population as measured by the TL3-level concentration index was not out of line with trends elsewhere in the OECD and was only slightly above the OECD average. However, this, again, reflects in part the construction of the index, which does not reflect the concentration in the Tokyo-Nagoya-Osaka mega-region.

← 9. Canada, Greece, Norway, Portugal and Switzerland. It is noteworthy that two of these countries have extremely remote northerly regions based on resource-extraction, while Greece also has plethora of small islands and Switzerland an extremely mountainous topography, which tends to favour fragmentation.

← 10. Data for some OECD countries are not available for 2012 and 2013.

← 11. Expenditure on public works did rise substantially in the 1990s but has since fallen back to below the level of 1990.

← 12. Non-interest central government spending (i.e. total expenditures minus national debt service, local allocation tax grants, etc.).

← 13. It should be noted that the period of coverage is not identical to that in Figure 1.11, so the two are not strictly comparable.

← 14. In 2011, 63% of women aged 25-34 had university degrees, as against 55% of men in the same age group.

← 15. The OECD harmonised unemployment rates are compiled for 34 OECD member countries and conform to the guidelines of the 13th Conference of Labour Statisticians of the International Labour Office (referred to as the ILO guidelines). In so far as possible, the data have been adjusted to ensure comparability over time. All series are benchmarked to labour force-survey-based estimates. The unemployment rates for the European Union member countries, Norway and Turkey are produced by the Statistical Office of the European Communities (Eurostat). For the remaining OECD countries, the OECD is responsible for collecting data and calculating unemployment rates.

← 16. Post-taxes and transfers, 50% threshold.

← 17. According to the OECD’s Social Expenditure Database, the ratio of family-related social expenditures to GDP was 0.96% in Japan in 2009, as against an OECD average of 2.3% and levels of 3.2 and 3.8% in France and Sweden, respectively. A Japanese estimate for 2011, showed a somewhat higher figure for Japan, at 1.35% (CJF, 2014).

← 18.;

← 19. Osaka, Hyogo, Miyagi and Nara.

← 20. Chapter 2 will rely chiefly on the OECD’s Metropolitan Database to analyse urban performance in terms of FUAs. However, because the coverage of the database is still limited, owing to the need to construct all the indicators included in it, the prefectural-level data is far richer and the analysis will, from time to time, use Statistics Japan’s MAs as a unit of analysis.

← 21. The pension eligibility age is now 65 for men (63 for women) for the basic pension and 61 for men (60 for women) for the Employees’ Pension Insurance.

← 22. See, inter alia, Joshi (1990); Joshi and Davies (1992); Gray and Chapman (2001); Joshi and Davies (2002); Davies and Pierre (2005).

← 23. Note that the cost of education is the same for women and men; lower earnings make it harder for women to recover the cost of their studies.

← 24. See OECD (2015h) for an overview. The Council on Science and Technology Policy, in particular, has been reorganised and renamed the Council on Science, Technology and Innovation.

← 25. Small, open economies must, of necessity, source more inputs from abroad in GVCs than larger countries such as Japan and the United States, where more of the value chain is likely to be domestic. The appropriate comparators for Japan are therefore the other G7 countries.

← 26. Japan maintains relatively low tariffs but non-tariff barriers are significant, such that the country ranks 49th out of 105 countries in the latest iteration of the World Bank’s Overall Trade Restrictiveness Index; only 4 OECD countries were found to be more restrictive.

← 27. Between JPY 120 million and JPY 260 million in subsidies are available each year for 5 years to 40 universities that commit to increase the number of Japanese students going overseas.

← 28. The reasons for this and its implications are explored in Chapter 3.