National economic performance is often compared across countries, and such comparisons are frequently used to highlight countries whose national policies appear to promote growth and development more successfully. However, national averages can hide wide regional differences in economic conditions and performances. OECD Regions at a Glance therefore presents a set of regional indicators – mainly in the form of graphs and maps – in order to identify those regions that outperform their country as a whole or the OECD area and those that lag behind. The patterns of development may differ widely in urban and rural areas, for example, and some areas may lag behind even when the national economy is performing well.

Population is unevenly distributed among regions within OECD member countries. In 2003, approximately 40% of the OECD population was located in just 10% of regions (Figure 1.1). The concentration was greatest in Australia and Canada, where 10% of regions accounted for 64% and 61%, respectively, of the national population. Iceland (50%), the United States (49%) and Mexico (47%) followed with around half of their population living in 10% of regions. In contrast, the territorial distribution of the population was more balanced in the Slovak Republic (12%), the Czech Republic and Belgium (17%) and Denmark (18%).

Over the last 30 years the elderly population (those aged 65 years and over) has increased dramatically in all OECD countries. In 2003, the elderly population in OECD countries represented 14% on average (Figure 2.1).

Gross domestic product (GDP) is unevenly distributed among regions within countries. In 2003, 38% of total OECD GDP was generated by only 10% of regions (Figure 3.1).

Between 1998 and 2003, gross domestic product (GDP) in OECD countries grew at an average annual rate of 3.1% in real terms (Figure 4.1). International differences in growth rates were as large as 7.5 percentage points, ranging from 1% in Japan to 8.5% in Ireland. Although significant, international differences are rather small in comparison to differences among regions within the same country.

Industries are unevenly distributed across OECD countries. According to the geographic concentration index, in 2003 the industry in which employment was, on average, most concentrated in OECD countries (Figure 5.1) was financial intermediation (45), followed by real estate, renting and business activities (43) and transport, storage and communication (37). In contrast, agriculture, hunting, forestry and fishing (20) and construction (31) displayed the lowest average concentration.

Growth in employment varies significantly among OECD countries (Figure 6.1). During 1998-2003, international differences in employment growth rates were as large as 6.2 percentage points, ranging from –2.2% in Poland to 4% in Spain.

Patent statistics provide a measure of innovation, as they reflect the inventive performance of countries, regions and firms. The geographic distribution of patents therefore indicates the level of diffusion of technology and knowledge across regions.

GDP per capita varies significantly among OECD countries (Figure 8.1). In 2003, GDP per capita in Luxembourg (USD 53 390) was more than double the OECD average (USD 24 824) and more than seven times that of Turkey (USD 6 910).

Labour productivity, one of the main indicators of economic performance, varies significantly among OECD countries. In 2003, Luxembourg displayed the highest GDP per worker (measured at PPP in constant prices), about 47% higher than the OECD average. Turkey’s productivity in 2003 was the lowest, at about 39% (Figure 9.1).

Regional specialisation varies considerably among OECD countries. Specialisation is commonly measured by the Balassa-Hoover index: the ratio between an industry’s weight in a region and its weight in the country overall. A region is specialised in an industry when the index is above 1 and it is not specialised when the index is below 1. A region’s degree of specialisation, therefore, can be measured as the weighted average of its degrees of specialisation in each industry. The higher this value, the more specialised the region.

In today’s knowledge-based economy a region’s growth prospects depend to a large extent on its ability to generate and use innovation. This capability, in turn, depends, among other factors, on the skills level of the regional labour force. The proportion of the adult population with tertiary education is a common proxy for a region’s skills level. It includes university-level education, from courses of short and medium duration to advanced research qualifications.

Unemployment rates vary significantly among OECD countries. In 2003, international differences in unemployment rates were as large as 17 percentage points, ranging from 2.5% in Mexico to 19.6% in Poland (Figure 12.1).

Labour force participation rates vary significantly among OECD countries. In 2003, international differences in participation rates ranged from 51% in Turkey to 87% in Iceland (Figure 13.1).

Economic performance varies significantly among OECD regions. But why are some regions more competitive than others? Regional benchmarking makes it possible to identify the factors behind the success of certain regions and to perceive the existence of unused resources in others.

A region’s economic performance can be measured as the difference between its growth rate and that of all OECD regions. Competitive regions will grow faster than others and will increase their share of total OECD GDP. By the same token, GDP growth will be slower in less competitive regions and their share in total GDP will fall.

Regional performance is a result both of national factors – such as national policies and the business cycle – and regional factors – such as demographic trends and regional policies. If all regions in a country grow faster than the regions in other OECD countries, their faster growth can be ascribed to national factors. On the other hand, to the extent that a region exhibits faster growth than all other OECD regions, including those in the same country, that growth can be ascribed to the region’s good performance (regional factors).

Over 1998-2003, about one-third of OECD regions – 34% or 101 regions – increased their share in total OECD GDP owing to region-specific factors. The increase was due to a relative increase in population in 37% of these regions, a relative rise in GDP per capita in 22% and relative growth in both components in the other 41%.

High growth in GDP per capita for 1998-2003 was a result of a relative increase in GDP per worker in a large majority of OECD regions (77%).

Rapid growth in GDP per capita over 1998- 2003 was due to a strong rise in productivity in 77% of regions. However, in 23%, the relative boost in GDP per capita was driven by a relative increase in one or more of the following variables: employment rates, participation rates and working age population.

The well-being of a region’s inhabitants depends on their ability to access resources and services that are often available only in large urban centres. The travelling time necessary to reach the closest agglomeration gives a measure of a region’s ability to quickly access resources and services.

A highly educated labour force is a major factor in determining regional competitiveness. The enrolment ratio is a commonly used measure of the level of participation in tertiary level education.

Voter turnout provides an indication of the degree of public trust in government and of citizens’ involvement in the political process. Figure 22.1 shows the variation in voter turnout across regions in OECD countries in the last national election. In Australia, where voting is mandatory, Tasmania records the highest OECD-area turnout rate (96%). Belgium, Austria, Italy and Turkey also record very high turnout rates in some regions. Among these countries, Belgium has the smallest regional variation (87%-93%).

Safety is an important factor in the attractiveness of regions, but the lack of international standards for crime statistics makes international comparisons difficult. Statistics on reported crime are affected by how crime is defined in the national legislation and by the statistical criteria used in recording offences. In addition, public propensity to report offences varies greatly, not only among countries, but also among regions in the same country.

The number of murders per inhabitant is a main indicator of a region’s safety level. Unlike other safety indicators, such as reported crime against property, the number of reported murders is less affected by the public propensity to report an offence. It is therefore more suitable for international comparison.

In many OECD countries home ownership is an important dimension of well-being. It protects owners from fluctuations in rents and ensures families a stable and secure shelter. In addition, the value of a property represents a major source of wealth for households. Differences in the rate of home ownership across OECD countries depend significantly on several factors, including rental subsidies, the existence of high-quality social housing and the deductibility of interest payments on loans from taxable income.

Motor vehicles emit millions of tons of pollutants into the air. In many urban areas, motor vehicles are the single largest contributor to groundlevel ozone, a major component of smog. The reduction of motorised traffic is therefore a policy target in many OECD countries. The number of private vehicles per capita is the indicator most commonly used to set policy targets for the integration of environmental objectives with transport policies.

Waste has an economic impact because waste disposal represents a significant cost for local authorities. It also has an environmental impact because waste is usually buried in landfills or burned in incinerators, often resulting in groundwater pollution, poor air quality and other forms of environmental degradation. Waste also has a social impact related to the location of waste disposal facilities. Concerns include odours, increased traffic and potential health risks. Anecdotal evidence indicates that poor and minority communities may be burdened with more than their fair share of waste disposal facilities.

The age-adjusted mortality rate is a basic indicator of the population’s health status. At the national level, it is the death rate that would occur in a country if its population’s age profile was the same as the OECD average. Therefore, a value higher than the OECD average indicates that, after accounting for differences in age, that country’s mortality rate is higher than the OECD average.

Premature mortality, measured in terms of potential years of life lost (PYLL), is often interpreted as a measure of preventable deaths. This indicator places the emphasis on deaths among younger people, in particular infant mortality and deaths due to illnesses and accidents suffered by children and young adults. Advances in medical technology, together with prevention and control, can reduce such deaths. Many of the main causes of premature mortality in the developed world are non-medical or involve risk-taking behaviour (accidents, smoking, alcohol, drugs) but also diseases such as cancer.

Cancer is the second major cause of death in most OECD countries, after cardiovascular diseases. Incidence rates of cancer can therefore be used as a partial measure of regional disparities in terms of healthcare needs. The steady rise in the elderly population has brought an increase in the number of new cases of cancer. It will rise even more steeply if exposure to behavioural risk factors – such as smoking, alcohol and an unhealthy diet – persists.

Density of physicians is frequently used as an indicator of health-care provision. An adequate number of qualified practising physicians, located according to need, helps to ensure the delivery of safe, high-quality medical services. However, it is hard to estimate the minimum number of physicians required to guarantee adequate provision. As well as the number of physicians, the hours they work and the presence of complementary and substitute health professionals (nurses, for instance) also determine actual levels of provision. However, the density of physicians is seldom expressed in full-time equivalents. Furthermore, the density indicator does not specify whether the physicians actually practise, nor does it reflect features specific to the region. The mix of private/hospital practice may carry a risk of double counting, depending on how the data are collected (e.g. by professional organisations). Another area not covered by the indicator is cross-border health-care provision.

Nursing staff are involved in several ways in the provision of both primary health care and hospital care. They form the largest category of health-care providers in almost all OECD member countries.

The number of hospital beds usually provides a measure of the resources available for delivering health services in hospitals. It does not, however, provide a comprehensive measure of capacity since it does not capture the capacity of hospitals to furnish services for nonadmitted patients (e.g. outpatient consultations, day care and ambulatory surgery). Nor is it a measure of physical accessibility to hospital health services. In fact, a region may have a large number of hospital beds but accessibility may be low if the hospital is located far from the population.

The number of computerised tomography (CT) scanners and magnetic resonance imaging (MRI) units can be used to measure the diffusion of modern medical technology and, more specifically, diagnostic techniques based on medical imaging. Both of these technologies are used to diagnose a wide range of disorders.

Tobacco is considered by the World Health Organization (WHO) to be the second major cause of death worldwide. It is a major risk factor for at least two of the leading causes of premature mortality: circulatory diseases and a range of cancers. In addition, it is an important contributory factor for respiratory diseases and remains the largest avoidable risk to health in OECD countries.

Obesity is a known risk factor for several health problems, including diabetes, hypertension and cardiovascular diseases, as well as respiratory and musculoskeletal disorders. There has been a considerable increase in obesity-related problems over the past two decades, along with an associated rise in health-care costs. Adults with a Body Mass Index (BMI) of over 30 are defined as obese. However, some ethnic groups may have equivalent levels of risk at lower or higher BMIs. Survey definitions differ significantly among countries. As a consequence, results may be quite different depending on whether obesity is self-reported (e.g. Australia, the United States) or measured (see Sources and Methodologies).