Society at a Glance Asia/Pacific 2011

Contents Chapter 1. How do People in the Asia/Pacific Region Spend their Time? Chapter 2. Interpreting OECD Social Indicators Society at a Glance Chapter 3. Society at a Glance: An Overview Chapter 4. General Context Indicators Asia/Pacific 2011 Chapter 5. Self-sufficiency Indicators Chapter 6. Equity Indicators Chapter 7. Health Indicators Chapter 8. Social Cohesion Indicators www.oecd.org/els/social/indicators/asia

Society at a Society Glance Asia/Pacific 2011 Asia/Pacific

Please cite this publication as: OECD (2011), Society at a Glance: Asia/Pacific 2011, OECD Publishing. http://dx.doi.org/10.1787/9789264106154-en This work is published on the OECD iLibrary, which gathers all OECD books, periodicals and statistical databases. Visit www.oecd-ilibrary.org, and do not hesitate to contact us for more information.

ISBN 978-92-64-10569-0 81 2011 08 1 P www.oecd.org/publishing -:HSTCQE=VUZ[^U:

Society at a Glance Asia/Pacific 2011 This work is published on the responsibility of the Secretary-General of the OECD. The opinions expressed and arguments employed herein do not necessarily reflect the official views of the Organisation or of the governments of its member countries.

This document and any map included herein are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area.

Please cite this publication as: OECD (2011), Society at a Glance: Asia/Pacific 2011, OECD Publishing. http://dx.doi.org/10.1787/9789264106154-en

ISBN 978-92-64-10569-0 (print) ISBN 978-92-64-10615-4 (PDF)

Photo credits: Cover: © Monkey Business - Fotolia.com; © TheFinalMiracle - Fotolia.com; © Galyna Andrushko - Fotolia.com; © Wolszczak - Fotolia.com; © iStockphoto.com/Eternia; © iStockphoto.com/thinqkreations.

Corrigenda to OECD publications may be found on line at: www.oecd.org/publishing/corrigenda. © OECD 2011

You can copy, download or print OECD content for your own use, and you can include excerpts from OECD publications, databases and multimedia products in your own documents, presentations, blogs, websites and teaching materials, provided that suitable acknowledgement of OECD as source and copyright owner is given. All requests for public or commercial use and translation rights should be submitted to [email protected]. Requests for permission to photocopy portions of this material for public or commercial use shall be addressed directly to the Copyright Clearance Center (CCC) at [email protected] or the Centre français d’exploitation du droit de copie (CFC) at [email protected]. FOREWORD

Foreword

This is the second edition of Society at a Glance: Asia/Pacific, the biennial OECD overview of social indicators. This report addresses the growing demand for quantitative evidence on the state of and trends in social well-being. It updates some indicators included in the first edition and adds some new ones. Additionally, this edition makes a considerable effort to return to the roots of the publication with its “At a Glance” overview. It reduces the number of indicators to 25. There is a strong focus on simple language and a shorter text. Charts and boxes are direct and simple and reflect a higher degree of standardisation. We believe that such an approach is essential in this increasingly information-rich age. The 2011 report includes 36 countries and economies for which comparable data are available, compared to the 25 covered in the 2009 edition. This report includes a special focus on unpaid work (Chapter 1). It also provides a guide to help readers in understanding the structure of social indicators (Chapter 2), and a summary – Society at a Glance, at a glance (Chapter 3). More detailed information on all indicators, including those not in this edition, can be found on the OECD web pages (www.oecd.org/els/social/indicators). This report was prepared by Simon Chapple. As this report addresses a wide range of topics, it would have been impossible to complete without the contributions of many people inside and outside the OECD Social Policy Division. These include Alexandra Bytchkova, Michael De Looper, Angelica Del Pero, Pauline Fron, Maxime Ladaique, Luca Lorenzoni, Marlène Mohier, Andrew Reilly and Kim Robin. Monika Queisser, Head of the OECD Social Policy Division, supervised the report.

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 3

TABLE OF CONTENTS

Table of contents

Acronyms and Conventional Signs ...... 7

OECD/Korea Policy Centre ...... 9

Chapter 1. How do People in the Asia/Pacific Region Spend their Time?...... 11 Why time use is an important social indicator ...... 12 The challenge of time-use comparisons across Asia/Pacific countries and economies ...... 12 Country-specific issues from the surveys ...... 15 Working time...... 16 Free time ...... 18 Conclusion ...... 22 Notes...... 23 Further reading ...... 23

Chapter 2. Interpreting OECD Social Indicators ...... 25 The purpose of social indicators ...... 26 The framework of OECD social indicators ...... 26 The selection and description of indicators ...... 28 What can be found in this publication ...... 31

Chapter 3. Society at a Glance: An Overview ...... 33

Chapter 4. General Context Indicators...... 37 GE1. GDP per capita ...... 38 GE2. Fertility ...... 40 GE3. Marriage and divorce...... 42 GE4. Migration ...... 44 GE5. Support rate ...... 46

Chapter 5. Self-sufficiency Indicators ...... 49 SS1. Employment ...... 50 SS2. Unemployment...... 52 SS3. Education...... 54 SS4. Early childhood education ...... 56 SS5. Education spending ...... 58

Chapter 6. Equity Indicators ...... 61 EQ1. Poverty...... 62 EQ2. Income inequality ...... 64 EQ3. Gender ratio ...... 66 EQ4. Pensions ...... 68 EQ5. Social spending ...... 70

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 5 TABLE OF CONTENTS

Chapter 7. Health Indicators ...... 73 HE1. Life expectancy ...... 74 HE2. Infant mortality ...... 76 HE3. Water and sanitation ...... 78 HE4. Adult height ...... 80 HE5. Health expenditure ...... 82

Chapter 8. Social Cohesion Indicators...... 85 CO1. Life satisfaction ...... 86 CO2. Trust ...... 88 CO3. Social behaviour ...... 90 CO4. Suicide...... 92 CO5. Voting ...... 94

This book has... StatLinks2 A service that delivers Excel® files from the printed page!

Look for the StatLinks at the bottom right-hand corner of the tables or graphs in this book. To download the matching Excel® spreadsheet, just type the link into your Internet browser, starting with the http://dx.doi.org prefix. If you’re reading the PDF e-book edition, and your PC is connected to the Internet, simply click on the link. You’ll findStatLinks appearing in more OECD books.

6 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 ACRONYMS AND CONVENTIONAL SIGNS

Acronyms and Conventional Signs

Asia/Pacific countries and economies ISO codes

Armenia ARM Azerbaijan AZE Bangladesh BGD Bhutan BTN Brunei Darussalam BRN Cambodia KHM China CHN Fiji FJI Hong Kong (China) HKG India IND Indonesia IDN Kazakhstan KAZ Korea, Democratic People’s Republic of PRK Kyrgyzstan KGZ Lao People’s Democratic Republic LAO Macao (China) MAC Malaysia MYS Maldives MDV Mongolia MNG Myanmar MMR Nepal NPL Pakistan PAK Papua New Guinea PNG Philippines PHL Samoa WSM Singapore SGP Sri Lanka LKA Tajikistan TJK Thailand THA Timor-Leste TLS Tonga TON Viet Nam VNM

OECD Asia/Pacific countries ISO codes

Australia AUS Japan JPN Korea, Republic of KOR New Zealand NZL

Asia/Pacific refers to all countries including OECD members Australia, Japan, the Republic of Korea and New Zealand.

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 7 ACRONYMS AND CONVENTIONAL SIGNS

Conventional signs .. Not available. (➘) in the legend relates to the variable for which countries are ranked from left to right in decreasing order. (➚) in the legend relates to the variable for which countries are ranked from left to right in increasing order.

8 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 OECD/KOREA POLICY CENTRE

OECD/Korea Policy Centre

The Joint OECD/Korea Policy Centre (www.oecdkorea.org) is an international co-operation organisation established by a Memorandum of Understanding between the OECD and the Government of the Republic of Korea. The Centre – officially opened on 7 July 2008 – results from the integration of four pre-existing OECD/Korea Centres, one of which was the Regional Centre on Health and Social Policy (RCHSP), established in 2005. The major functions of the Centre are to research international standards and policies on international taxation, competition, public governance and social policy sectors in OECD member economies and to disseminate research outcomes to public officials and experts in the Asian region. In the area of health and social policy, the Centre promotes policy dialogue and information-sharing between OECD economies and non-OECD Asian/ Pacific countries and economies. There are three main areas of work: social protection statistics (jointly with the International Labour Organisation and the Asian Development Bank); health expenditure and financing statistics (jointly with the Asian Pacific National Health Account Network and the World Health Organisation) and pension policies (jointly with the World Bank). In pursuit of this vision, the Centre hosts various kinds of educational programmes, international meetings, seminars and workshops in each sector and provides policy forums presented by experts at home and abroad.

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 9

Society at a Glance Asia/Pacific 2011 © OECD 2011

Chapter 1

How do People in the Asia/Pacific Region Spend their Time?

11 1. HOW DO PEOPLE IN THE ASIA/PACIFIC REGION SPEND THEIR TIME?

Why time use is an important social indicator This special chapter considers time-use patterns across Asia/Pacific countries and economies and compares these to OECD patterns. It focuses particularly on unpaid work and free-time. Unpaid work within the home includes cooking, cleaning, shopping and raising children. Unpaid work contributes significantly to the well-being of families and others. Focussing solely on market income measures, rather than also considering unpaid work, may lead to incorrect inferences about levels and changes in well-being. Since women typically do much of the unpaid work, considering unpaid work acknowledges the importance of a major female contribution to well-being. In addition to unpaid work within the household, people carry out vital unremunerated work for relatives who live outside the household and for the wider community. Voluntary work, such as helping out neighbours, caring for older people and people with disabilities, aiding charities, assisting new immigrants, training sports teams and administering schools, also contributes to societal well-being but is not included in traditional economic measures. Traditional income-based measures of well-being also neglect the measurement of the amount of free time and leisure time that people have available to them. Leisure time is self-evidently an important component of people’s well-being.

The challenge of time-use comparisons across Asia/Pacific countries and economies Time-use statistics are quantitative summaries of how people allocate their time over a specified period – usually 24 hours. Time-use statistics include information on the type of activities people engage in and the duration of each of these activities. The respondent’s description of his or her activities are coded and systematically aggregated into a set of broad categories such as “time spent in work”, “time spent on household chores”, “time spent in leisure activities” or “time spent on personal care”. Ideally, they sample the population over a year in a manner that takes into consideration weekends and holidays, as well as seasonal effects. Much as elsewhere in the world, up-to-date, reliable and comparable data on time use are very limited for countries of the Asia/Pacific region. Time-use data were obtained for 12 out of 36 countries included in the Society at a Glance: Asia/Pacific edition. Table 1.1 lists the countries for which surveys were available, the dates of the survey, the number of respondents, the data source and information on other data features. Although data were obtained for only a minority of countries, the size of the total regional population covered is substantial. Three of the four most populous countries in the region – India, China and Pakistan – are covered. Additionally, the countries for which time-use information is available cover a broad spectrum. The subset covered includes countries at very different levels of economic development, with industrial developed economies, the large emerging economies of China and India, former Soviet transition economies, and some of the world’s poorest countries with substantial peasant sectors, such as Cambodia and Pakistan. The geographic coverage is also good.

12 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 1. HOW DO PEOPLE IN THE ASIA/PACIFIC REGION SPEND THEIR TIME?

Table 1.1. Available time-use surveys

Age range Years of the survey Sample size Type of survey Source Other data features (years)

Armenia 2008 Unknown Unknown 15-80 UNECE Statistical Division Database, Not representative of the year Gender Statistics Database or of the population Armenia 2004 235 people, Two diaries, One 15-80 National Statistical Service of the Not representative of the year 60 households for weekdays and Republic of Armenia or of the population one for weekends Armenia TU Pilot Survey 2004 Australia 2006 6 961 people, Two consecutive > 15 Australian Bureau of Statistics, 3 643 households days, diary Time Use Survey Cambodia November 2003 to 2 000 households Diary sheet > 5 National Institute of Statistics of January 2005 included in the Cambodia household survey www.nis.gov.kh/index.php/statistics/ surveys/cses China 2008 37 142 people, One weekday + one 15-74 National Bureau of Statistics of China, Not representative of the year 16 661 households week-end day Time Use Survey India July 1998-June 1999 18 591 households One week – > 6 Ministry of Statistics and Programme interviewing Implementation. Government of India method Japan 2007 18 291 individuals, Unknown > 10 Statistics Bureau and Statistical Not representative of the year 3 866 households Research Training Institute, Survey on Time Use and Leisure Activities (Questionnaire B) Kazakhstan 2006 Unknown Unknown 20-74 UNECE Statistical Division Database Gender Statistics Database Korea, Rep. 2004 12 750 households Two consecutive > 10 Korea National Statistical Office, Not representative of the year days, diary Time Use Survey Kyrgyzstan 2005 Unknown Unknown 20-74 UNECE Statistical Division Database Gender Statistics Database Mongolia 2000 2 753 people, 24-hour diary > 12 NSO, UNDP, Not representative of the year 1 086 households A Pilot Time Use survey, 2000. New Zealand July 1998-June 1999 8 532 people Two consecutive > 12 Statistics New Zealand, days, diary Time Use Survey Pakistan 2007 19 380 households 24-hour diary > 10 Government of Pakistan, Statistics Representative for the Division, Federal Bureau of Statistics, population and the year Time Use Survey 2007, April 2009

However, several strong caveats regarding the inter-country comparability of the time- use data used here need to be kept in mind. One relates to the data collection method. Data on time use can be collected in two ways: surveys and diaries. Surveys ask respondents how much time they spent on activity j in a defined reference period, such as the previous 24 hours. However, this method can miss activities that take short amounts of time, and can lead to misreporting if activities do not have a fixed schedule over time (Ilahi, 2000). Diaries, on the other hand, are filled by the respondent him/herself or with the help of fieldworkers. Usually based on a 24-hours recall, this method emphasises the chronology of events. Because it is more detailed, it is also more costly in money and in the respondent’s time. Contrary to the survey method, it captures non-routine activities well. However, there is a risk that respondents confuse multiple or simultaneous activities. For instance, childcare often tends to be over-reported, because it is the most common multiple activity (Ilahi, 2000). Differences in survey coverage also complicate the cross-country comparison. Given that time allocation is not constant over the year, the season of data collection may introduce a bias in the comparability of time-use data, especially in agricultural economies (Ilahi, 2000). For example, in China, Japan, the Republic of Korea, New Zealand, Armenia and Mongolia, the survey took place within a period of one month or less, so the sample

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 13 1. HOW DO PEOPLE IN THE ASIA/PACIFIC REGION SPEND THEIR TIME?

may not be representative for the year. Moreover, while some surveys were conducted in spring (China, Mongolia, New Zealand), others were conducted in autumn (the Republic of Korea, Japan, Armenia). In Pakistan, seasonal variations in the time-use pattern were captured by conducting the survey through the four quarters of the year. In Cambodia and Australia, the period covered by the survey is equal or longer than a year, also avoiding the issue of seasonality. Another type of bias concerns the days covered by the survey. In Australia, the Republic of Korea and New Zealand, diaries are filled for two consecutive days. However, this method does not distinguish between weekdays and week-end days. For Pakistan, respondents were asked about their activities only for 24 hours. However, the percentage distribution of diary days of the week shows that all days of the week are approximately equally represented except Sundays – reflecting the fact that Sunday is generally not regarded as a work day in Pakistan. Surveys that examine only the allocation of time during workdays are not entirely comparable with those that cover non-work days as well. The former tend to underestimate women’s work contribution, since unpaid/non-market work that is usually accomplished by women often takes place during weekends (United Nations, 1995). In the Chinese time-use survey, respondents filled the diary for one weekday and one week-end day. For the Mongolian survey, the seven days of a week were divided into five (Sunday, Monday, Tuesday/Wednesday, Thursday/Friday, Saturday). The same number of households was selected to answer for each group of days. In other words, two-thirds of the respondents were assigned to make two-day diaries and one-third of the respondents were assigned to make three-day diaries. In the Australian survey, to deal with the issue of holidays, four 13-day periods in 2006 containing a representative proportion of public and school holidays were defined. Geographic coverage is an important issue too. If the survey coverage is not national, the surveyed area may not be representative of the country as a whole. More generally speaking, averages may hide intra-country differences, such as urban and rural differences, which are usually wider in developing countries. In Pakistan, some areas located in difficult terrain could not be included in the survey due to constraints of time, access and cost. Nevertheless, the population excluded represents only 3% of the total population. Another related consideration concerns the population group actually covered. Indeed, average time spent on paid work or leisure will certainly vary according to the age of the population surveyed. For instance, the data for Kazakhstan and Kyrgyzstan covers the population aged 20 to 74. On the other hand, Cambodian data concerns the population aged 5 and over. In the former case, excluding those older than 75 means that paid work will tend to be higher than if they were included. In Cambodia’s case, the average time spent on paid work is expected to be lower than if only the adult population were covered, whereas in the former two countries the opposite holds. Surveys also differ in whether they record simultaneous activities and if so, how. Generally, the data are coded to show people engaged in one activity at a time. For some countries, however, surveys include separate questions designed to learn about simultaneous activities (i.e. watching television while cooking or supervising children while ironing clothes), which allows a distinction between “primary” and “secondary” activities. So one limitation of the general approach is that “primary” activities are meticulously tracked while “secondary” ones are usually overlooked. The comparability of estimates for secondary activities is also affected by whether activities that typically require only a few minutes of

14 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 1. HOW DO PEOPLE IN THE ASIA/PACIFIC REGION SPEND THEIR TIME?

one’s time, i.e. moving a load of laundry from the washer to the dryer, are reported consistently enough to produce comparable estimates of the time devoted to them. This in turn depends on the length of the time slots in which respondents can report their activities, which range from five minutes in Australia and New Zealand to 15 minutes in Japan. We were unable to adjust for differences in data collection methods, nor for differences in survey coverage or respondent age. These strong comparability limitations should be kept in mind when considering the results. Given that the categories of activities are not equally detailed in all the data sources used here, the activities were aggregated into five very broad categories for the purpose of harmonisation: 1) Market work and learning activities, 2) Non-market work, 3) Leisure, 4) Personal care, and 5) Other. Harmonisation was eased for several countries (Mongolia, Pakistan and India) for which the survey used the UN Trial International Classification of Activities for Time-Use Statistics (United Nations, 2009).1 In the “market work and learning activities” category, all jobs and work that encompass the creation of market value are included. Learning activities are also included, covering time spent at school, homework, personal studies and professional training. Following the ICATUS terminology, this first category covers employment for establishments, primary production activities (not for establishments), services for income and other production of goods (not for establishments) and all learning activities. “Non- market work” covers routine household chores and management (e.g. cooking, laundry, pet care, shopping and gardening), care for others (for household and non-household members) and volunteering activities. “Leisure” includes hobbies, sporting activities, participation in social or cultural events, use of mass media and socialising activities. “Personal care” includes sleep, eating and drinking as well as other household medical and personal services (hygiene, grooming, visiting doctors, etc.). “Other” contains religious activities, civic obligations and unspecified activities and travel when the travel could not be included in the other categories.2 For some countries, more detailed information on the type of leisure and personal care (including sleeping time) is available, so more detailed types of activities can be considered.

Country-specific issues from the surveys Although category harmonisation at the global level is desirable, when conducting a survey on time allocation there are cultural features peculiar to each society that deserve mention. Time may not have the same meaning in societies where large numbers of people are peasants, illiteracy is considerable, and society does not march according to the dictates of synchronised clocks, as it does in rich industrialised societies where people’s regular rhythms are co-ordinated like clockwork. What is more, the way time-use surveys are designed, or the way categories are organised and constructed from filled diaries, in itself gives some interesting insights into cultural differences across the region’s countries. Country-specific activities are illustrated by time-use survey categories. For instance, in the Indian 1998-99 time-use survey, a category was created for making cow dung cakes, which was recorded as a sub-category of animal husbandry in the primary production section. Another India-specific activity was created for “talking, gossiping and quarreling”, surprisingly categorised under “Personal care and Self-Maintenance”. Indeed, Indians do spend a consequential amount of time on this, especially in rural areas. Men are found to dedicate even more time than women to this activity, with on average almost eight hours

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 15 1. HOW DO PEOPLE IN THE ASIA/PACIFIC REGION SPEND THEIR TIME?

per week, compared to seven hours for women (United Nations, 2000). To maintain consistency in the cross-country comparison, talking, gossiping and quarreling were categorised here under “socialising”, a leisure category. The distinction between what constitutes leisure and what constitutes unpaid work is not always clear-cut between countries. The Japanese Statistics Bureau makes a distinction between what is done as part of housework and what is done as leisure for several activities, for example, “making sweets”, versus making sweets as a hobby. Additionally, in Japan, contrary to other countries, “pet care” and “walking the dog” were considered as leisure. To maintain consistency with all other countries, all these “productive hobbies” were categorised under “non-market work”. For Armenia, Kazakhstan and Kyrgyzstan, “volunteering” was categorised as “leisure”. Since detailed data on time spent on volunteering activities was not available for Kazakhstan, we could not extract the time spent on “volunteering” from the broader “leisure” category, so that leisure time for Kazakhstan still includes time dedicated to volunteering. The meaning of caring activities may vary according to the country. Contrary to most of the surveys, in the Korean, Chinese and Japanese time-use surveys, there is no division between care for household members and care for non-household members. Instead, they have one category for “family care” and another for “care for others”. It is interesting to note that these three countries are marked by the Confucian culture and philosophy, which encourages filial piety and is even given institutional legitimacy.

Working time In all countries, both males and females work more than 350 minutes per average day, i.e. approximately six hours, or one-quarter of the day. This amount of time represent one- third of people’s waking time (Figure 1.1). However, differences exist between countries in the time spent on work. People work the longest in Mongolia, where men and women work on average respectively 581 minutes (almost 10 hours) and 637 minutes (almost 11 hours) per day. In Cambodia, where people work the shortest time, males and females work on average four hours less than in Mongolia (384 minutes, or 6.4 hours, for males; and 414 minutes per day, 6.9 hours, for females). The low amounts of work in Cambodia almost certainly reflect the fact that the base population is five years and older, meaning that considerable numbers of children are included. The four rich OECD countries stand in the middle of the ranking. People in the region tend to work more in total than the OECD average. Generally women tend to do more total work than men, except in Pakistan and New Zealand, where men do marginally more work. Now consideration turns to the two major sub-categories of total work – market work and learning, and unpaid work, respectively. If only time allocated to market activities and learning is considered, Figure 1.2 shows that Mongolia is the most marketised society, followed by China, the Republic of Korea and Cambodia. At the lower end are Armenia, Australia, New Zealand and Kazakhstan. In all the region’s countries, men always do more market work and learning than women. This gender gap in market work is proportionately greatest in Pakistan and smallest in Mongolia. Figure 1.3 presents the time-use information with regard to unpaid work. Unpaid work is especially high in Kazakhstan, being the highest in the region for both women and men. Unpaid work is very low in Cambodia and, to a lesser extent, the Republic of Korea. Pakistani men also do very small amounts of unpaid work. Overall, women always do more unpaid work than men, but the

16 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 1. HOW DO PEOPLE IN THE ASIA/PACIFIC REGION SPEND THEIR TIME?

Figure 1.1. Total working time is the longest in Mongolia Total hours worked in an average day, by country and gender

Men Women () 12

11 10 10 10 9 9 9 9 8 8 8 8 8 8 8 8 8 8 8 7 8 7 7 7 7 7 7 6 6 6 6

4 n d n a n n ta n alia n ta dia ta s r s n s golia t Japa Chi I n ia 2008 ia 2004 n Paki n ea, Rep. us rgyz OECD28 Cambodia r A Mo rme rme New Zeala Ko Ky Kazakh A A

Source: See Table 1.1 on available Time Use Surveys. 1 2 http://dx.doi.org/10.1787/888932545996

gender gap varies greatly, being relatively high in Pakistan and low in New Zealand. The overall picture from Figures 1.2 and 1.3 is that Pakistan is very gender specialised in terms of the composition of work and New Zealand is relatively unspecialised. Contrary to expectation, women spend similar amounts of time on unpaid activities in OECD countries and in non-OECD developing countries. In industrialised countries, the diffusion of household appliances, such as washing machines and freezers, and good access to basic services, such as electricity and water, should reduce the time spent on

Figure 1.2. Mongolians spend the longest time in market work and learning Time spent on market and learning activities in hours per average day, by country and gender

Men Women () 8

8

7 7 77 6 7 6 6 6 6 5 5 55 5 5 4 5 4 4 3 4 4 3 3 3 2 2 2 2

0 n d n n n a ta alia n dia ta ta n s r n s s golia t I Japa Chi n ia 2004 ia 2008 n n Paki us rgyz ea, Rep. OECD28 A r Cambodia Mo rme rme New Zeala Kazakh Ky Ko A A

Source: See Table 1.1 on available Time-use Surveys. 1 2 http://dx.doi.org/10.1787/888932546015

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 17 1. HOW DO PEOPLE IN THE ASIA/PACIFIC REGION SPEND THEIR TIME?

Figure 1.3. Kazakhs spend the most time on non-market activities Time spent on non-market/unpaid work in hours per average day, by country and gender

Men Women () 8

6 6 6 5 5 5 5 5 5 5 5 4 4 4 4

2 4 2 3 3 2 2 2 2 1 1 1 1 1 0 0 0 a n d n n n n n ta alia ta dia ta golia s r s n s Chi Japa n t I ia 2008 us ia 2004 ea, Rep. Paki n rgyz n OECD28 Cambodia r Mo A Ko New Zeala rme Ky rme Kazakh A A

Source: See Table 1.1 on available Time-use Surveys. 1 2 http://dx.doi.org/10.1787/888932546034

household work considerably. For instance, in developing countries such as Pakistan and Cambodia, women spend a large amount of time collecting water and wood (Ilahi, 2000) – activity that does not exist in developed countries. However, the data shows that time spent by women on unpaid work does not vary much across our sample countries. The average time spent on unpaid work by women in Asian OECD countries was 268 minutes per day (4.5 hours), compared to the OECD average of 220 minutes per day (3.6 hours). Nevertheless, although time spent on unpaid work does not vary much between OECD and non-OECD countries, the physical intensity of work is unlikely to be the same, as in developing countries it may involve carrying water, collecting and carrying firewood, or processing foodstuffs by hand. Although it is not directly captured by time-use surveys, the intensity of effort is an important element to be kept in mind.

Free time Free time is all time not spent working. Free time can in turn be divided into personal care and leisure. The way time is allocated between all the possible activities people can do in their free time depends on their preferences and on various socio-economic incentives. Sleeping time is essential for individual’s well-being. It can also be assumed that sleep adds to an individual’s productivity and is necessary for efficiency and is required even if the goal is to maximise income. As underlined by Biddle and Hamermesh (1990), many labour economics models consider sleep as a fixed biological constant. However, theory and evidence tend to show that sleep time, like any other activity, responds to economic incentives. If this is the case, sleep becomes a leisure-like activity and depends on individual preferences. Several recent studies thus understand sleep as leisure and categorise all sleep-related activities as leisure (OECD, 2009, p. 22). As already mentioned, both leisure and sleeping time matter for an individual’s physical and mental health and well-being. As a consequence, looking at leisure time and sleep as social indicators should be considered as a fruitful initiative.

18 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 1. HOW DO PEOPLE IN THE ASIA/PACIFIC REGION SPEND THEIR TIME?

People in Cambodia (note the caveat of age – the sample is of those 5 years and older) and Pakistan have the largest amounts of free time (Figure 1.4). Cambodians have more than 16 hours of free time per day (1 000 minutes), whereas in Mongolia, which is the country where people work the longest, people have only 13 hours (800 minutes) of free time. Generally speaking, men have slightly more free time than women (Figure 1.5), reflecting longer total working time for women (Figure 1.1).

Figure 1.4. Cambodians have the most free time Total free time by gender in hours per average day

Men Women () 18

17 17 17 17 16 16 16 16 16 16 16 16 16 15 16 15 15 15 17 18 16 16 14 14 14 14 16 16 13 14

12 n n a n d n ta ta dia n alia n ta golia s s n r s n I Chi Japa t ia 2008 ia 2004 us n rgyz n ea, Rep. Paki OECD28 Mo r A Cambodia Kazakh rme Ky rme Ko New Zeala A A

Source: See Table 1.1 on available Time-use Surveys. 1 2 http://dx.doi.org/10.1787/888932546053

Figure 1.5. Men usually have more free time than women Gender gap in total free time in minutes per average day

100 95 90

80 75 70

60 56 51 53 50 44 40

30 29 26 24 20 11 10 4 0 -8-8 -2 -10 n d n a n n ta n alia n ta ta dia s r s s golia n t Japa Chi n I ia 2008 ia 2004 Paki n us ea, Rep. rgyz n OECD28 A Cambodia r Mo rme New Zeala Ko Kazakh Ky rme A A

Source: See Table 1.1 on available Time-use Surveys. 1 2 http://dx.doi.org/10.1787/888932546072

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 19 1. HOW DO PEOPLE IN THE ASIA/PACIFIC REGION SPEND THEIR TIME?

How do people actually allocate their free time? People spend most of it – well above half – on personal care, which means primarily sleeping time and secondarily eating time (Figure 1.6). The composition of free time varies greatly across the region. Personal care is an especially high share of free time in Armenia, at over 90%, and is especially low in Kazakhstan, at a bit above 50%.

Figure 1.6. Personal care represents more than three-fifths of people’s free time Leisure and personal care time and their share in total free time, in hours per average day

Personal care time Leisure time

Men Women 12 1 Armenia 2008 1 12 11 4 Mongolia 3 10 13 4 Pakistan 3 14 12 4 China 4 12 12 5 India 4 11 11 5 Japan 4 11 11 5 Armenia 2004 4 11 11 5 New Zealand 4 12 10 5 Kyrgyzstan 4 10 11 5 Korea, Rep. 4 11 11 5 Australia 4 11 13 5 Cambodia 5 13 8 6 Kazakhstan 6 8

11 5 OECD28 5 11 100 80 60 40 20 0020 40 60 80 100 % %

Note: Countries are ranked by women’s leisure time. Source: See Table 1.1 on available Time-use Surveys. 1 2 http://dx.doi.org/10.1787/888932546091

In all countries except the Republic of Korea, people dedicate at least half of their free time to sleeping (Figure 1.7). Except in Pakistan, where people spend the longest time on personal care, sleep represents more than 70% of the total time dedicated to personal care. Sleeping time tends to be similar in geographically proximate countries (Figure 1.7). The Republic of Korea and Japan are the two countries where people sleep the least, with less than 8 hours sleep per 24 hours. Sleeping times are the same in Australia and New Zealand and in Kazakhstan and Kyrgyzstan. In the former two countries men and women sleep on average, respectively, 511 and 514 minutes (8 hours and a half) per 24 hours whereas in Kazakhstan and Kyrgyzstan men sleep 475 minutes and women 472 minutes (approximately 8 hours). People sleep the most in China, with 9 hours of sleep (540 and 544 minutes of sleep for men and women, respectively). There are no strong gender differences in hours of sleep. Although on average women have less free time than men (since they work more), they do not sleep less. In Japan, where the gender gap in sleeping time is the largest, women sleep only 10 minutes less than men. Biddle and Hamermesh (1990) found that the fact of having young children reduces sleep times only of mothers; they also showed that paid work and higher educational attainment tend to reduce hours of sleep. Indeed, in their model, higher wage rates reduce the hours of sleep because individuals have a greater incentive to spend time in the labour market as income (and thus the price of time) increases. This effect is even higher for men than for women.

20 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 1. HOW DO PEOPLE IN THE ASIA/PACIFIC REGION SPEND THEIR TIME?

Figure 1.7. The Chinese sleep the most Time spent on sleep in hours per 24 hours

Men Women () 10

9 9 9 9 9 9 9 8

8 9 8 8 8 8 9 9 9 8

8 8 8 8

7 n n n d n a n dia n sta sta ralia n sta Japa t I Chi us rgyz rea, Rep. A Paki OECD28 Kazakh Ky Ko New Zeala

Note: Countries are ranked by women's sleeping time. Source: See Table 1.1 on available Time-use Surveys. 1 2 http://dx.doi.org/10.1787/888932546110

Turning from personal care time to the other major component of free time – leisure – people in Asia/Pacific countries and economies spend more than 10% of their time on leisure activities (Figure 1.8), though this is less than in OECD countries. Mongolian women spend the least time on leisure, with three hours per day (174 minutes) on average, whereas men in Kazakhstan spend the longest time on leisure activities, with more than 6 hours (378 minutes) per day, representing 27% of daily time. However, as already

Figure 1.8. People have the longest leisure time in Kazakhstan Leisure time in hours per average day, by country and gender

Men Women () 7

6 6 6

5 5 5 5 5 4 4 4 4 5 5 4 4 4 4 4 4 4 4 3 5 3 3

5 5 5 2 5 1 1 1 0 n a n d n n ta n dia n ta alia ta golia s n s r s n Chi I Japa t ia 2008 ia 2004 us n Paki n rgyz ea, Rep. OECD28 Mo r A Cambodia rme rme New Zeala Ky Ko Kazakh A A

Source: See Table 1.1 on available Time-use Surveys. 1 2 http://dx.doi.org/10.1787/888932546129

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 21 1. HOW DO PEOPLE IN THE ASIA/PACIFIC REGION SPEND THEIR TIME?

mentioned, the Kazakhstan data on leisure includes time spent on volunteering activities, while data for other countries classify volunteering in non-market work. This categorisation issue may bias the Kazakh leisure time upward. Nevertheless, the bias should not be large, given that on average the time dedicated to volunteering is low.3 Men have more leisure time than women (Figure 1.9), with on average 39 minutes more per day. The gender gap in leisure time is the highest in Armenia, with a difference of over one hour. The smallest gap, 16 minutes, is found in New Zealand.

Figure 1.9. Men have more leisure time than women Gender gap in leisure time in minutes per average day

90

80 77

70 62 60 52 49 50 45 40 40 37 37 38 30 30 27 23 20 16 10

0 -1 -10 d n n a n n n alia ta n ta ta dia r s golia s s n Japa t Chi n I ia 2008 us ia 2004 n ea, Rep. Paki rgyz n OECD28 A Cambodia r Mo rme New Zeala Kazakh Ko Ky rme A A

Source: See Table 1.1 on available Time-use Surveys. 1 2 http://dx.doi.org/10.1787/888932546148

Conclusion Although reliable and comparable data is limited, much can still be learned on time use in Asia/Pacific countries and economies. The accent in this special chapter was on total working time and free time. It showed that important disparities exist across countries and gender in terms of both the types and duration of work time. In the large majority of countries, women work more than men. This difference is driven by the time spent on non-market unpaid work. Time-use data tends to underline that women do contribute to wealth creation when performing housework and caring activities. Differences in the way people allocate their free time to various activities exist as well, although no strong gender/geographic pattern emerges. Nonetheless, the time spent on sporting activities, on inactive leisure activities (such as watching TV) or on sleep has implications for personal development and health. The type of activities people engage in is directly related to life satisfaction. People experience different feelings of satisfaction, anxiety, nervousness or reward depending on the activity being performed, and this affects their mental state. For instance, among a sample of 625 Germans, outdoor activities and watching TV were found to be the most pleasurable, with volunteering and working the most rewarding (White and Dolan, 2009).

22 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 1. HOW DO PEOPLE IN THE ASIA/PACIFIC REGION SPEND THEIR TIME?

The feelings associated with each daily activity may vary according to the society and the individual, but are definitely linked to subjective well-being. All these elements can be of great use to government agencies as they monitor and design public policies. Additionally, gender differences in time use, especially in work time, as explained in this chapter, should be taken into account when designing any labour policy, since policies in this field can have differentiated gender impact. More generally, learning about people’s time allocation ensures a better understanding of a society for policymakers who are concerned with the efficiency and equity of their policies. Thus, conducting surveys on time use on a regular basis is highly desirable for every country. Longitudinal data on time use – which is scarce today – would make possible comparisons in time use patterns across time. Furthermore, international comparisons of time use, which would be facilitated by a substantial effort to harmonise activity categories at the global level, would allow researchers to enquire more deeply into countries’ socio-cultural specificities and to think in a more advised way about suitable social policies.

Notes 1. http://unstats.un.org/unsd/methods/timeuse/icatus/icatus_2.htm. 2. When possible time spent on travel is classified in the category of the activity to which it is linked. When there is no information on the purpose of the travel, travel time is included in the “Other” category. 3. On average, the time dedicated to the care of non-household members and to community services is seven minutes per day (this average is calculated for nine countries and does not include Cambodia or Kazakhstan, since data for these countries are not available). However, if we do not include New Zealand where the time dedicated to care of non-household members and to community services is the highest (with 35.5 minutes per day for women and 28 minutes for men), the average falls to four minutes per day.

Further reading Biddle, J.E. and D.S. Hamermesh (1990), “Sleep and the Allocation of Time”, Journal of Political Economy, Vol. 98, No. 5, Part 1, October, pp. 922-943. Eun, Ki-Soo (2003), “Changing Roles of the Family and State for Elderly Care. A Confucian Perspective”, in V.L. Bengston. and A. Lowebstein (eds.), Global Aging and Challenges to Families, Walter de Gruyter, New York, pp. 253-271. Fan, R. (2006), “Confucian Filial Piety and Long Term Care for Aged Parents”, HealthCare Ethics Committee Forum, Vol. 18, No. 1, March. pp. 1-17. Fuess, S.M. Jr. (2006), “Leisure Time in Japan: How Much and for Whom?”, IZA Discussion Paper, No. 2002, Bonn, March. Ilahi, N. (2000), “The Intra-household Allocation of Time and Tasks: What Have We Learnt from the Empirical Literature?”, Policy Research Report on Gender and Development, World Bank Working Paper Series, No. 13, June. Ironmonger, D. (1999), “An Overview of Time Use Surveys”, International Seminar on Time Use Studies, 7-10 December, Ahmedabad, India. Kripke, D.F., L. Garfinkel, D.L. Wingard, M.R. Klauber and M.R. Marler (2002), “Mortality Associated With Sleep Duration and Insomnia”, Archives of General Psychiatry, Vol. 59, pp. 131-136. Morris, E. (2004), “Promoting Employment in Cambodia: Analysis and Options”, Cambodia Socio- Economic Survey, ILO Subregional Office for East Asia.

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 23 1. HOW DO PEOPLE IN THE ASIA/PACIFIC REGION SPEND THEIR TIME?

OECD (2009), Society at a Glance 2009, OECD Publishing, Paris. Patel, S.R., N.T. Ayas, M.R. Malhotra et al. (2004), “A Prospective Study of Sleep Duration and Mortality Risk in Women”, Sleep, Vol. 27, No. 3, pp. 440-444. Stiglitz, J., A. Sen and J.-P. Fitoussi (2007), “Report by the Commission on the Measurement of Economic Performance and Social Progress”, www.stiglitz-sen-fitoussi.fr. United Nations (1995), Human Development Report, Oxford University Press, New York. United Nations (2000), “Country Report: INDIA. Conducting the Time Use Survey – Indian Experience”, Gender Issues in the Measurement of Paid and Unpaid Work, Expert Group Meeting on Methods for Conducting Time-Use Surveys. 23-27 October, New York. White, M.P. and P. Dolan (2009), “Accounting for the Richness of Daily Activities”, Psychological Science, Vol. 20, No. 8, August, pp. 1000-1008.

24 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 Society at a Glance Asia/Pacific 2011 © OECD 2011

Chapter 2

Interpreting OECD Social Indicators

25 2. INTERPRETING OECD SOCIAL INDICATORS

The purpose of social indicators Society at a Glance 2011 provides information about two questions: ● Compared with their own past and with other OECD countries, what progress have countries made in their social development? ● How effective have been societies’ efforts to further their own development? Addressing the first question about societal progress requires indicators that cover a broad range of social outcomes across countries and over time. As social development requires improvements in health, education and economic resources, as well as a stable basis for social interactions, indicators have to be found for all these dimensions. The second question about societal effectiveness is even more challenging to answer. Societies try to influence social outcomes, often through government policy. Whether policies are effective in achieving their aims is a critical issue. Indicators help to make that assessment. A first step is to compare the resources intended to change outcomes across countries and contrast those resources with social outcomes. While this comparison is far from being a comprehensive evaluation of policy effectiveness, indicators can contribute to highlighting areas where more evaluative work may be needed.

The framework of OECD social indicators The structure applied here is not a full-scale framework of social indicators. But it is more than a simple list of indicators. This framework has been informed by experiences in other parts of the OECD on policy and outcome assessment in a variety of fields. It draws, in particular, on the OECD experience with environmental indicators. These indicators are organised in a framework known as “Pressure-State-Response” (PSR).* In this framework human activities exert pressures on the environment, which affect the state of natural resources and environmental conditions, and which prompt a societal response to these changes through various policies. The PSR framework highlights these sequential links, which in turn helps decision-makers and the public to interconnections that are often overlooked. A similar approach for social indicators is followed in this report. Indicators are grouped along two dimensions their nature and the policy fields that they cover. The first dimension is broken down into three areas: 1. Social context refers to variables that, while not usually direct policy targets, are crucial for understanding the social policy context. For example, the proportion of elderly people in the total population is not a policy target. However, it is relevant information about the social landscape in which, for example, health, taxation or pension policy

* The PSR framework is itself a variant of an approach that has also given rise to the “Driving force- State-Response” (DSR) model used by the United Nations Committee for Sustainable Development; and the “Driving force-Pressure-State-Impact-Response” (DPSIR) model used by the European Environment Agency.

26 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 2. INTERPRETING OECD SOCIAL INDICATORS

responses are made. Unlike other indicators, trends in social context indicators cannot be unambiguously interpreted as “good” or “bad”. 2. Social status indicators describe the social outcomes that policies try to influence. These indicators describe the general conditions of the population. Ideally, the indicators chosen are ones that can be easily and unambiguously interpreted – all countries would rather have low poverty rates than high ones, for example. 3. Societal response indicators provide information about what society is doing to affect social status indicators. Societal responses include indicators of government policy settings. Additionally, the activities of non-governmental organisations, families and the broader civil society also involve societal responses. Comparing societal response indicators with social status indicators provides an initial indication of policy effectiveness. While social indicators are allocated to one of the three groups above, the allocation between context and status categories is not always straightforward. For example, reducing fertility rates may be a policy objective in countries such as China, but in other countries, like Australia, low fertility rates are part of the context of social policy. Similarly, family breakdown can be regarded as a failure of public policies in some countries, whereas it may not be an explicit policy concern in others. An important limitation of the social context, social status and societal response indicators used here is that these are presented at a national level. For member countries with a significant degree of federalism, such as the United States, Canada and Australia, indicators may not be reflective of the different regions within the federation, which may have different contexts, outcomes and social responses. This limitation should be borne in mind in considering the indicators presented below. The second dimension of the OECD framework groups indicators according to the broad policy fields that they cover. Four broad objectives of social policy are used to classify indicators of social status and societal response: 1. Self-sufficiency is an underlying objective of social policy. Self-sufficiency is promoted by ensuring people’s active social and economic participation, and their autonomy in activities of daily life. 2. Equity is another longstanding objective of social policy. Equitable outcomes are measured mainly in terms of access by people and families to resources. 3. Health status is a fundamental objective of health care systems, but improving health status also requires a wider focus on its social determinants, making health a central objective of social policy. 4. Social cohesion is often identified as an over-arching objective of countries’ social policies. While little agreement exists on what it means, a range of symptoms are informative about a lack of social cohesion. Social cohesion is more positively evident in the extent to which people participate in their communities.

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 27 2. INTERPRETING OECD SOCIAL INDICATORS

The selection and description of indicators Asia/Pacific countries and economies differ substantially in the ways that they collect and publish social indicators. In selecting indicators for this report, the following questions were considered. ● What is the minimum degree of indicator comparability across countries? This report strives to present the best comparative information for each of the areas covered. However, the indicators presented are not confined to those for which there is “absolute” comparability. Readers are, however, alerted as to the nature of the data used and the limits to comparability. ● What is the minimum number of countries for which the data must be available? This report includes only indicators that are available for a considerable majority of Asia/Pacific countries and economies. ● What breakdowns should be used at a country level? Social indicators can often be decomposed at a national level into outcomes by social sub-categories, such as people’s age, gender and family type. Pragmatism governs here: the breakdowns presented here vary according to the indicator considered. Chapters 4 to 8 describe the key evidence. Individual indicators can be relevant for multiple areas of social policy. That is to say, they could plausibly be included under more than one category. For example, the ability to undertake activities of daily living without assistance is potentially an indicator of social cohesion, self-sufficiency and health. Indicators are presented here under the category for which they are considered to be most relevant. Throughout this volume, the code associated with each indicator (e.g. GE1) is used to relate it to a policy field (as listed in the tables below), while a numbering of the indicators is used to simplify cross-references. While the name and coding of indicators used in this volume may differ from those in the previous issue of Society at a Glance: Asia/Pacific, an effort is made to assure the continuity of the areas covered.

General social context indicators (GE) When comparing social status and societal response indicators, it is easy to suggest that one country is doing badly relative to others, or that another is spending a lot of money in a particular area compared with others. It is important to put such statements into a broader context. For example, national income levels vary across OECD countries. If there is any link between income and health, richer countries may have better health conditions than poor ones, irrespectively of societal responses. If the demand for health care services increases with income (as appears to be the case), rich countries may spend more on health care (as a percentage of national income) than poorer countries. These observations do not mean that the indicators of health status and health spending are misleading. They do mean, however, that the general context behind the data should be borne in mind when considering policy implications. General context (GE) indicators, including fertility, marriage and divorce, migration and the support rate, provide the general background for the other indicators in this report. GDP per capita is a social outcome in its own right, giving an indication of the average material well-being of that society.

28 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 2. INTERPRETING OECD SOCIAL INDICATORS

Table 2.1. List of general context indicators (GE)

GE1. GDP per capita GE2. Fertility GE3. Marriage and divorce GE4. Migration GE5. Support rate

Self-sufficiency (SS) For many people, paid employment (SS1) provides income, identity and social interactions. In addition, social security systems are funded by contributions by paid working people. Hence promoting higher paid employment is a priority for most countries. Being unemployed (SS2) means that supporting oneself and one’s dependants is difficult, despite being available for work. Education (SS3) provides information on human capital accumulation. A better education enables longer term self-sufficiency now and in the future, including in paid employment. Early childhood education provides a foundation for future learning, as well as freeing up mothers to choose to work (SS4). Education spending provides information on the primary social response made by governments to help ensure self-sufficiency (SS5). Table 2.2 lists the indicators of social status and societal response that are most relevant for assessing whether OECD countries have been successful in meeting the goals of assuring the self-sufficiency of people and their families.

Table 2.2. List of self-sufficiency indicators (SS)

Social status Societal responses

SS1. Employment SS2. Unemployment SS3. Education SS4. Early childhood education SS5. Education spending

EQ1. Poverty EQ5. Social spending EQ2. Income inequality

Note: Indicators in italics are those that, while presented in another sub-section, are also relevant for an assessment of self-sufficiency.

Equity (EQ) Equity has many dimensions. It concerns the ability to access social services and economic opportunities, as well as equity in outcomes. Opinions vary widely as to what exactly entails a fair or a just distribution of opportunities. Additionally, as it is hard to obtain information on all dimensions of equity, the social status equity indicators are focussed on inequality in financial resources. Poverty (EQ1) is a natural starting point for considering equity at the bottom of society. Absolute measures of poverty are used here, since many of the region’s countries are very poor. In addition to an absolute poverty measure, an indicator of relative inequality across the distribution is also considered (EQ2). Differences in the life chances of boys and girls are an important feature of several larger Asia/Pacific countries and economies, and these are addressed by consideration of the sex ratio (EQ3). Pension coverage and the old-age

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 29 2. INTERPRETING OECD SOCIAL INDICATORS

Table 2.3. List of equity indicators (EQ)

Social status Societal responses

EQ1. Poverty EQ5. Social spending EQ2. Income inequality EQ3. Gender ratio EQ4. Pensions

SS1. Employment HE5. Health expenditure SS2. Unemployment SS3. Education

Note: Indicators in italics are those that, while presented in another sub-section, are also relevant for an assessment of equity outcomes.

replacement rate are important indicators of the extent to which society treats its older people in an equitable fashion (EQ4). Many Asia/Pacific countries and economies have social protection systems that redistribute resources and insure people against various contingencies. These interventions are summarised by public social spending (EQ5).

Health (HE) The links between social and health conditions are strong. Indeed, educational gains, accompanied by public health measures, better access to health care and continuing progress in medical technology, have contributed to significant improvements in health status, as measured by life expectancy (HE1). To a significant extent, improvements in life expectancy reflect lower infant mortality (HE2). Water and sanitation (HE3) are important public health issues in many Asia/Pacific countries and economies, and are linked to both life expectancy and infant mortality via the transmission of disease. Adult height gives an indication of cumulated net nutrition and the burden of disease of adults over their entire period of childhood (HE4). Health spending (HE5) is a general and key part of the policy response of health care systems to concerns about health conditions. Nevertheless, sometimes health problems are rooted in interrelated social conditions – such as unemployment, poverty and inadequate housing – that are beyond the reach of health policies. Moreover, the effectiveness of health interventions often depends less on spending levels per se than on other characteristics of the health care system, such as low coverage of medical insurance or the role of co-payments, which may act as barriers to seeking medical help.

Table 2.4. List of health indicators (HE)

Social status Societal responses

HE1. Life expectancy HE5. Health expenditure HE2. Infant mortality HE3. Water and sanitation HE4. Adult height EQ5. Social spending

Note: Indicators in italics are those that, while presented in another sub-section, are also relevant for an assessment of health outcomes.

Social cohesion (CO) Promoting social cohesion is an important social policy goal in many countries. However, because there is no commonly-accepted definition, identifying suitable

30 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 2. INTERPRETING OECD SOCIAL INDICATORS

indicators is especially difficult. In Society at a Glance: Asia/Pacific, considerable effort has gone into finding new and better indicators of social cohesion. The approach taken in Society at a Glance is to assess social cohesion through indicators that describe the extent to which citizens participate in societal life and derive satisfaction from their daily activities, indicators that inform about various pathologies and conditions that put affected people at risk of social exclusion, or indicators that reveal the extent of social strife. Life satisfaction is strongly associated with good family relationships and confidence in the broader society (CO1). A general measure of trust in other people (CO2) may indicate the degree to which economic and social exchange is facilitated, enhancing well-being and facilitating socially productive collective action. Pro-social behaviour is behaviour that contributes to the positive functioning of society, such as giving one’s money or time or helping strangers. Anti-social behaviour, such as corruption, is the contrary (CO3). The intentional killing of oneself is not just evidence of a personal breakdown, but also of a deterioration of the social context in which a person lives (CO4). High voter turnout indicates that a country’s political system enjoys a high level of participation, increasing its effectiveness and reflecting a broad public consensus about its legitimacy (CO5).

Table 2.5. List of social cohesion indicators (CO)

CO1. Life satisfaction CO2. Trust CO3. Social behaviour CO4. Suicide CO5. Voting

What can be found in this publication In each of the five domains covered in Chapters 4 to 8 of this report, there is a page of text and a page of charts for each of the five indicators. Both charts and text are, to a degree, standardised. The text typically commences with a very brief justification for inclusion of the indicator. Both the text and charts then address the most recent headline indicator data, with country performances ranked from best to worst. Changes in the indicator over time are then considered on a chart to the right. Providing a standardised introduction and opening charts for each of the 25 indicators is intended to facilitate interpretation by the reader. The choice of the time period considered for changes is partly determined by data constraints. However, ideally changes are examined either 1) over the last generation, to compare how society is evolving in the longer term, or 2) over the period of the current economic crisis, so the extent to which recent adverse economic events are influencing social indicators can be studied. Having addressed the indicator and the changes, the text and charts then typically consider interesting alternative breakdowns of the indicator, or relationships with other social outcomes or policies. For each indicator, an opening boxed section on “Definition and measurement” provides the definitions of the data used and a discussion of potential measurement issues. Finally, suggestions for further reading are sometimes given.

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 31

Society at a Glance Asia/Pacific 2011 © OECD 2011

Chapter 3

Society at a Glance: An Overview

33 3. SOCIETY AT A GLANCE: AN OVERVIEW

Society at a Glance: Asia/Pacific 2011 uses the 25 social indicators presented below. There is obvious value in providing a short summary – hence Society at a Glance: An Overview. In Society at a Glance 2009, a summary was provided through a table that selected two indicators per domain, chosen on the basis of their a priori importance and through consultation with member countries, and assigned “green circle” for performance in the top two deciles, “yellow triangle” for performance in the middle six deciles and “red diamond” for the bottom two deciles. A very similar “traffic light” approach is taken below, but including all relevant indicators. Some of the 25 indicators are excluded because they are not capable of unambiguous interpretation (a higher value being desirable and a lower value being undesirable, for example). For the general context (GE) indicators, only household income is capable of an unambiguous interpretation, so the others are excluded. Additionally, the social response indicators (education, social and health spending) are not included, as these are policy inputs, not social outcomes. A similar traffic lights table could also be constructed for changes in indicators. However this was not done, for data reasons. First, change data are missing for many more countries than levels data, and hence such a table is much less informative as a summary. Second, changes cannot be considered over consistent periods. Indeed, it is not clear over what period changes should be compared. For some indicators, which change only slowly, longer periods may be more desirable than for indicators that can change fairly rapidly over short periods. These exclusions leave 18 out of the 25 social indicators to be summarised in Table 3.1. Circles are the top two deciles, triangles are the middle six deciles and diamonds are the bottom two deciles. Blanks are placed where there is no country indicator information. It is necessary to make numerous caveats about the meaning and interpretation of Table 3.1 in terms of national comparisons. Different governments and different countries will have different policy priorities. Their priorities may be economic outcomes (inflation, GDP, or fiscal balance, for example) rather than social outcomes such as those considered here. In such a case, having all red social outcomes in Table 3.1 may be the price they are willing to pay for success elsewhere. Or, given a focus on social outcomes, they may be willing to trade-off many “diamonds” for the one “circle” social outcome they deem most desirable. Alternatively, there may be other social outcomes, not considered here, which are stronger priorities at a national level than the ones that have been chosen for international comparisons here. Observed patterns of diamonds, triangles and circles may simply reflect national differences in preferences for outcomes. Equally, observed patterns may reflect that social outcomes change relatively slowly, rather than current or even recent policy settings. Lastly, the trade-offs between social outcomes may vary between countries because of societal or cultural differences that are unrelated to policy choices, making it easier for some countries to generate outcomes for a given policy effort. For all these reasons, we caution against ranking country performance by an aggregate social index, such as summing the numbers of green or red lights across indicators.

34 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 3. SOCIETY AT A GLANCE: AN OVERVIEW ...... ◆ ● ● ◆ ◆ ● ◆ ◆ ◆ ● ● ● ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ (%) CO5 Voter turnout available Latest year ● ● ● ◆ ● ◆ ◆ ◆ ◆ ◆ ◆ ● ◆ ● ● ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ CO4 Suicide Suicide 100 000 Mid-2000s (deaths per (deaths population) ...... ● ◆ ◆ ◆ ● ● ● ● ◆ ● ◆ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ CO3 2009 Index score World Giving Giving World (scale 0(scale to 1) ...... ● ● ◆ ◆ ● ◆ ◆ ● % ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ http://dx.doi.org/10.1787/888932546642 CO2 trust trust Share Share available of people in others, Latest year year Latest expressing expressing ...... ◆ ● ◆ ◆ ● ● ◆ ● ● ◆ ● ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ CO1 score score available on a scale on a Well-being Well-being Latest year year Latest of 10 ladders ) those in the six intermediate deciles intermediate ) those in the six

1 2 ▲ ...... ◆ ◆ ◆ ◆ ◆ ◆ ● ● ● ● ● ● ● ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ HE3 of total available improved improved population population Proportion Proportion Latest year year Latest served with sanitation (%) sanitation .. ◆ ◆ ◆ ◆ ◆ ◆ ● ● ● ● ● ● ● ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ HE2 2008 Infant births) both sexes mortality rate rate mortality (per 1(per 000 live ● ◆ ● ◆ ● ◆ ● ● ◆ ◆ ● ◆ ● ◆ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ Life HE1 total 2008 years at birth, expectancy expectancy population, population, ...... ● ◆ ◆ ● ◆ ● ◆ ● ◆ ● ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ EQ4 force system, available Latest year Latest the pension % of labour Coverage of Coverage .. ◆ ◆ ● ● ◆ ◆ ◆ ● ◆ ● ● ● ◆ ● ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ EQ3 females Mid-2000s Gender ratio ratio Gender ) those in the bottom two deciles and “triangles” ( deciles and “triangles” two the bottom ) those in males per 100 males per ratio as number of as number ratio (population 0-4(population yr) ◆ Ratio defined as sex as sex Ratio defined ...... ● ● ● ● ◆ ◆ ◆ ◆ ● ● ◆ ● ◆ ◆ % ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ Gini Gini EQ2 coefficient, Mid-2000s ...... ● ◆ ● ◆ ● ◆ ● ◆ ● ◆ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ (%) EQ1 Share Share less than living with living Mid-2000s of population USD 2 per day ...... ● ◆ ● ● ◆ ● ◆ ● ◆ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ (%) SS4 education education childhood childhood Mid-2000s Participation Participation rates in early Overview of the social situation in Asia/Pacific countrieseconomies and Asia/Pacific in situation the social of Overview ...... ● ● ● ◆ ● ◆ ◆ ◆ ◆ ◆ ● ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ SS3 2010 years of total of total Average schooling ...... ◆ ◆ ◆ ◆ ◆ ● ● ● ● ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ (%) SS2 2009 Table 3.1. Table ment rates rates ment Unemploy- ...... ◆ ● ● ● ◆ ● ◆ ◆ ● ◆ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ (%) SS1 rates rates 2009 Employment Employment ) denotes countries are in the top two deciles, “diamonds” ( in the top two are ) denotes countries .. ● ◆ ● ◆ ● ● ● ◆ ◆ ◆ ● ◆ ◆ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ GE1 GDP 2008 per capita, (2009 USD) “Circles” ( ● “Circles” Armenia Australia Azerbaijan Bangladesh Bhutan Darussalam Brunei Cambodia China Fiji Hong Kong (China) India Indonesia Japan Kazakhstan DPR Korea, Rep. Korea, Kyrgyzstan Lao PDR (China) Macao Malaysia Maldives Mongolia Myanmar Nepal Zealand New Pakistan Guinea New Papua Philippines Samoa Singapore Sri Lanka Tajikistan Thailand Timor-Leste Tonga Viet Nam

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 35

Society at a Glance Asia/Pacific 2011 © OECD 2011

Chapter 4

General Context Indicators

GE1. GDP per capita GE2. Fertility GE3. Marriage and divorce GE4. Migration GE5. Support rate

37 4. GENERAL CONTEXT INDICATORS

GE1. GDP per capita

Gross Domestic Product per person (GDP per Poorer countries in the Asia/Pacific region capita) is the most widely used comparative indi- are tending to grow at a faster rate than richer cator of economic performance. It measures the ones (GE1.2).There is a negative correlation sum of marketed goods and services produced between the pace of growth in GDP per capita over within the national boundary, averaged across the period 1995-2008 and the initial level of GDP everyone who lives within the national boundary. per capita in 1995. Thus poorer countries are catching up to richer ones in the region, providing There are vast differences in GDP per capita evidence for economic theories of GDP conver- across the region (Panel A, GE1.1). The region gence. Azerbaijan, Armenia and China are catch- includes some of the richest as well as some of the ing up more rapidly than one would predict given poorest countries in the world. Australia’s GDP per their level of GDP, while Nepal and Papua New capita is 180 times higher than that of Myanmar. Guinea are performing at lower rates than would The Asia/Pacific OECD countries have relatively be expected (Papua is actually diverging). high GDP per capita, as do Brunei, Singapore and the Hong Kong (China) economy. Other countries with very low GDP per capita include Timor, Nepal, Definition and measurement Bangladesh and the Lao People’s Democratic Among the different measures available in the Republic. Differences in GDP per capita within the System of National Accounts (SNA), Gross Asia/Pacific region are much greater than within Domestic Product (GDP) per capita is the one most the OECD. commonly used for comparing the sizes of econo- mies across countries. GDP per capita is calcu- Equally, there are massive differences in per lated using a country’s GDP in 2009 USD and capita GDP growth rates between 1995 and 2008 dividing that by the country’s total population. (Panel B, GE1.1). Annual average growth rates span The 2009 USD value is used to convert national from a negative nearly 1% in Brunei to rates in currencies so that cross-national comparisons can be made. Annual average growth rates in GDP excess of a positive 8% in Myanmar, China, Armenia per capita are calculated using GDP per capita and Azerbaijan. The large Indian emerging economy expressed in constant national currency. The data was also a strong performer. It is interesting to note comes from the 2009 International Monetary Fund that the six lowest performers on GDP growth were World Economic Outlook online database all island nations – Brunei, Papua New Guinea, Fiji, (www.imf.org/external/pubs/ft/weo/2009/02/ Japan, Tonga and New Zealand – albeit with very dif- weodata/weoselco.aspx?g=2001&sg=All+countries). ferent levels of economic development.

38 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 4. GENERAL CONTEXT INDICATORS

GE1.1. GDP per capita and recent trends

Panel A. Current GDP per capita (), 2008 Panel B. Real annual average growth rate or latest year available, 2009 USD of GDP per capita, 1995-2008 46 824 Australia 2.1 40 909 OECD 3.1 36 384 Singapore 3.2 35 633 Japan 1.0 31 901 Brunei Darussalam -0.8 30 726 Hong Kong (China) 2.9 30 453 New Zealand 1.8 21 653 Korea, Rep. 3.8 50 000 40 000 30 000 20 000 10 000 0 8 719 Kazakhstan 6.7 8 537 Asia/Pacific 4.0 8 118 Malaysia 2.7 4 116 Thailand 2.2 3 674 Fiji 0.5 3 654 Maldives 4.6 3 259 China 8.8 2 736 Samoa 2.9 2 464 Azerbaijan 11.0 2 307 Tonga 1.3 2 114 Bhutan 7.1 1 915 Indonesia 2.2 1 845 Philippines 2.3 1 623 Sri Lanka 3.8 1 503 Mongolia 4.3 1 113 Armenia 10.3 1 022 Pakistan 2.1 1 017 India 5.1 835 Viet Nam 5.8 725 Kyrgyzstan 3.9 686 Papua New Guinea -0.3 649 Cambodia 7.1 578 Tajikistan 5.6 467 Lao PDR 4.2 464 Bangladesh 3.6 444 Nepal 2.2 382 Timor-Leste 257 Myanmar 8.1 10 000 8 000 6 000 4 000 2 000 0 -2 024681012

GE1.2. GDP per capita is converging across the Asia/Pacific region

% change 1995-2008 13

11 AZE ARM

9 CHN MMR KHM BTN 7 KAZ TJK VNM 5 IND MDV LAO MNG Asia/Pacific BGD KGZ LKA KOR OECD 3 WSM SGP PHL THA MYS HKG NPL PAK IDN NZL AUS TON 1 JPN FJI PNG BRN -1 1.8 2.3 2.8 3.3 3.8 4.3 4.8 Level 1995 log of GDP per capita

Source: Data from the 2009 International Monetary Fund World Economic Outlook online database (www.imf.org/external/pubs/ft/weo/2009/02/ weodata/weoselco.aspx?g=2001&sg=All+countries). 1 2 http://dx.doi.org/10.1787/888932546167

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 39 4. GENERAL CONTEXT INDICATORS

GE2. Fertility

The total fertility rate indicates the number have provided a context in which many couples of children an average woman will have in her desire a small family, while fertility control is more life time. The population is replaced at a total fer- widely considered as culturally acceptable. A tility rate of a little over two, allowing for some fertility drop has sometimes been encouraged by mortality during infancy and childhood. restrictive policies, as in China with the one-child Fertility rates in Asia/Pacific countries and policy launched in the early 1970s. economies display huge disparities (GE2.1). In the Birth rates are lower in richer countries Macao (China) economy, women have less than (GE2.2). There is a negative correlation between one child on average, whilst in Timor-Leste, GDP per capita and births per woman. Improve- women average 6.5 children. Fertility rates tend to ments in human development, such as higher be much lower in richer countries and economies health care expenditure (see HE2), improved life (Hong Kong [China], the Republic of Korea, Japan, expectancy (see HE1) reduced mortality (see HE4), Australia and New Zealand), emerging economies and increased female education and labour force excepting India (China, Thailand, Viet Nam and participation rates (see EQ2, SS1 and SS3) play an Indonesia) and in countries of the former Soviet essential role in reducing fertility rates. As a Union (Armenia, Kazakhstan and Kyrgyzstan). consequence, since decline in fertility is associ- High rates are found in Pakistan and the Pacific ated with socio-economic improvements in living Islands – Timor-Leste, Papua New Guinea, Tonga standards, today poor countries are likely to expe- and Samoa. Unlike the OECD, the majority of rience future decreases in fertility as their level of countries and economies in the Asia/Pacific group income increases, as has happened for the OECD have fertility rates in excess of replacement. countries. There have been rapid declines in fertility rates over the last generation (GE2.1). Other OECD countries that already went through their demo- Definition and measurement graphic transition have experienced a much The total fertility rate in a specific year corre- slower drop in fertility, at an average of 1% sponds to the number of children that would be decrease annually. The various paces of fertility born to each woman if she were to live to the declines reflect the important variations in the end of her childbearing years and if the likeli- stages of demographic transition across Asia/ hood of her giving birth to children at each life Pacific countries and economies. Countries such stage followed the currently prevailing age- as the Republic of Korea, Mongolia, Viet Nam, specific fertility rates. Maldives, Thailand and China, which experienced The data presented here are extracted from the very rapid declines in fertility, are now at the end World Bank’s World Development Indicators online of their demographic transition, with 2008 fertility database (2009, http://devdata.worldbank.org/data- rates almost comparable to the OECD’s. The socio- query). economic conditions of many countries in Asia

40 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 4. GENERAL CONTEXT INDICATORS

GE2.1. Fertility rates and changes

Panel A. Fertility rates in Asia/Pacific higher than in OECD Panel B. Rapid declines in Asia/Pacific fertility Fertility rate, total (births per woman) 2008 ()Annual average rate of decline, 1970-2008 6.5 Timor-Leste 0 4.1 Papua New Guinea 1 4.0 Tonga 1 4.0 Pakistan 1 4.0 Samoa 1 3.5 Lao PDR 1 3.4 Tajikistan 2 3.1 Philippines 2 2.9 Cambodia 2 2.9 Nepal 2 2.7 India 2 2.7 Fiji 1 2.7 Kyrgyzstan 2 2.6 Bhutan 2 2.6 Kazakhstan 1 2.6 Malaysia 2 2.5 Asia/Pacific 2 2.3 Bangladesh 3 2.3 Sri Lanka 2 2.3 Myanmar 3 2.3 Azerbaijan 2 2.2 New Zealand 1 2.2 Indonesia 2 2.1 Brunei Darussalam 3 2.1 Viet Nam 3 2.0 Maldives 3 2.0 Mongolia 3 2.0 Australia 1 1.9 Korea, DPR 2 1.8 Thailand 3 1.8 China 3 1.7 OECD 1 1.7 Armenia 2 1.3 Japan 1 1.3 Singapore 2 1.2 Korea, Rep. 3 1.0 Hong Kong (China) 3 0.9 Macao (China) 3 567 43 210 -1 0 1 2 34

GE2.2. Rich countries have higher fertility rates

GDP per capita, 2008, USD (log) 5.0 OECD AUS SGP JPN BRN 4.5 HKG NZL KOR

4.0 Asia/Pacific KAZ MYS THA AZE ARM FJI 3.5 MDV CHN IDN WSM TON MNG PHL BTN LKA PNG 3.0 VNM IND KHMTJK LAO PAK BGD MMR NPL TLS 2.5

2.0 021 3 4 5 67 Fertility rate (births per woman)

Source: The data presented here are extracted from the World Bank’s World Development Indicators online database (2009, http://devdata.worldbank.org/data-query). 1 2 http://dx.doi.org/10.1787/888932546186

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 41 4. GENERAL CONTEXT INDICATORS

GE3. Marriage and divorce

The formation of adult partnerships through housing market may also play a role. In almost all marriage and their dissolution through divorce countries (Thailand and Viet Nam are exceptions), provides important information on a fundamental adult sons are more likely to live at home than the social institution in all Asia/Pacific countries and daughters, possibly reflecting traditions of patrilo- economies: families. Cultural attitudes and cality. Sex differences are especially apparent in approaches to marriage as a form of adult partner- Bangladesh, India and Pakistan. ing, as well as to divorce, differ substantially High marriage rates in China reflect a very within the region. high married proportion of the adult Chinese There is important diversity in both marriage population (GE3.3). Indian and Indonesian pat- and divorce rates across Asia/Pacific countries and terns are similar. In a number of countries like economies (Panel A, GE3.1). Crude marriage rates New Zealand and Australia, cohabiting relation- are highest in the Maldives, Tajikistan and ships are common. These are almost non-existent Bangladesh. The lowest marriage rates are found in elsewhere. Separated or divorced adults are Samoa, Macao (China), New Zealand, Australia and uncommon in China, India and Indonesia. Thailand. Turning to divorce rates, the crude divorce rate in the Maldives is three times higher than the average of the Asia/Pacific countries and economies. Definition and measurement Divorce rates are low relative to high marriage rates The crude marriage rate conveys the number of in Viet Nam, Indonesia, Tajikistan, Mongolia, marriages formed each year as a ratio to the total Azerbaijan and Armenia. adult population; similarly, the crude divorce rate Over the past two decades, the changes were is the number of marriages dissolved in a given more pronounced in marriage rates than in divorce year, also expressed with respect to the total adult rates (Panel B, GE3.1). The marriage rate fell the population. Data on crude marriage rates and crude divorce rates are expressed per population sharpest in the Maldives, with a 4 percentage point of 1 000. The share of adults living with their par- decrease. At the same time, the drop in the Maldives’ ents refers to the number of people aged over divorce rate was also the fastest, at 5 percentage 24 answering “yes” to the question “do you live points. Big rises in marriage rates were found in with your parents?” divided by the total adult pop- China, Tajikistan and Mongolia. Marriage declined in ulation aged over 24. the wealthiest countries, such as the Republic of The data used here comes from the 1997 Korea, Japan, Australia, New Zealand and Singapore. and 2008 Demographic Yearbook of the UN Depart- Divorce rates are also increasing slowly but surely in ment of Economic and Social Affairs’ Statistics the Republic of Korea, Hong Kong (China) and Japan, Division (http://unstats.un.org/unsd/demographic/ reflecting an upward trend since 1980. For instance, products/dyb/2000_round.htm) and of the World the Republic of Korea’s crude divorce rate nearly Marriage data 2008 of the UN Department of Eco- doubled between 1980 and 2004. nomic and Social Affairs’ Population Division (www.un.org/esa/population/publications/WMD2008/ Inter-generational patterns of family living WP_WMD_2008/Data.html). The data on adults also differ considerably across the Asia/Pacific living with their parents comes from the World region (GE3.2). Very high proportions of adult Indi- Values Survey Database (www.wvsevsdb.com/wvs/ ans, Thais and Bangladeshis live with their parents, WVSAnalize.jsp). The most recent available data whereas this pattern is uncommon in Australia and was used for each country so that the time spread New Zealand. The differences may be cultural and goes from 1997 to 2008. Weights provided in the may also be influenced by the scale of the welfare database were used when available. Partner status state, which substitutes for some family relation- data come from the Gallup World Poll. ships. The nature of the housing stock and the

42 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 4. GENERAL CONTEXT INDICATORS

GE3.1. Marriage and divorce and recent trends

Crude marriage rates Crude divorce rates Panel A. Crude marriage () and divorce rates, Panel B. Percentage points change of marriages per 1 000 persons, 2008 or latest available year and divorces, 1995 to most recent year

Maldives Tajikistan Bangladesh Mongolia Azerbaijan Kyrgyzstan Kazakhstan Fiji Asia/Pacific Indonesia China Hong Kong (China) Singapore Tonga Korea, Rep. Philippines Japan Armenia Viet Nam Brunei Darussalam Australia Thailand OECD New Zealand Macao (China) Samoa 20 15 10 5 0 -6 -4 -2 0 2 4 6 4

GE3.2. Adults living with their parents GE3.3. Partner status in selected Asia/Pacific countries and economies Total () Men Women Separated % of adults leaving with their parents % of adult Married Domestic 70 or Widowed Single Total population (➘) partner divorced 60 60 55 China 79 0 1 2 18 100 50 50 India 75 0 1 3 21 100

40 Indonesia 72 0 1 6 21 100 40 38 36 38 35 35 32 32 Japan 68 0 4 5 23 100 31 31 30 28 28 27 Korea, Rep. 63 0 1 3 34 100 25 26 21 25 24 20 20 21 21 20 19 Australia 54 8 6 4 27 100 19 13 14 13 14 New Zealand 53 12 6 4 25 100 7 10 9 7 4 0 2 d a n ia ia ia n e n h d n a) n n s s r s n dia ralia n e sta n t OECD Chi Japa me n baija I us r do r gapo glade A rea, Rep.g (ChiViet Nam A n Malay in Paki n Thaila n I Aze S New Zeala Ko Asia/Pacific Ba g Ho n Ho

Note: Dates refering to marriage data and those for divorce rates may differ. Source: 1997 and 2008 Demographic Yearbook of the UN Department of Economic and Social Affairs’ Statistics Division (http://unstats.un.org/ unsd/demographic/products/dyb/2000_round.htm) and of the World Marriage Data 2008 of the UN Department of Economic and Social Affairs’ Population Division (www.un.org/esa/population/publications/WMD2008/WP_WMD_2008/Data.html). The data on adults living with their parents comes for the World Values Survey Database. 1 2 http://dx.doi.org/10.1787/888932546205

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 43 4. GENERAL CONTEXT INDICATORS

GE4. Migration

Immigrants can provide social and economic region, except Japan and the Republic of Korea dynamism and bring considerable positive diver- where laws regarding the foreign-born are restric- sity to their new countries. For many countries tive, host the highest shares of international and economies of the Asia/Pacific region, skilled migrants. emigration is a pressing policy issue. Other coun- Some countries experience high migration tries in the region, such as the Philippines and rates combined with increases in the share of Samoa, rely considerably on emigrant remittances foreign-born people in the total population for their incomes. (GE4.3). Samoa, Fiji, Tonga and Tajikistan wit- There is an impressive variation in the share nessed large flows of out-migration between 2000 of international migrants in the total population and 2005 along with a growth in the share of inter- across Asia (Panel A, GE4.1). In the three econo- national migrants, ranging from 4 percentage mies of Singapore, Macao (China) and Hong Kong points for Fiji to 75 percentage points for Armenia (China) well over one-third of the population has over the same period. been born off-shore. Australia and New Zealand also have high migrant shares, to a significant extent reflecting migration from elsewhere in the Definition and measurement region. In China, migrants represented less than Place of birth is used to estimate the immi- 0.05% of the total population in 2005 – a negligible grant population. The international migrant share. The foreign-born population is also negligi- stock is the number of people born in a country ble – less than two in every 100 people – in 17 of other than that in which they live, expressed as the region’s other countries. a share of the resident country’s total popula- tion. The net migration rate is the number of In a small majority of the countries consid- immigrants minus the number of emigrants ered, net migration is negative – people are flowing over a period of time divided by the population out (Panel B, GE4.1). Between 2000 and 2005, emigra- of the receiving country over that same period, tion was largest in poorer island countries (Samoa, expressed as the net number per 1 000 popula- Tonga, Fiji, Sri Lanka) and countries of the former tion. Although the inflow and outflow data are Soviet Union (Tajikistan, Armenia, Kyrgyzstan, generally not comparable, the net migration Kazakhstan, Azerbaijan). The most important statistics “net out” the main source of non- in-migration flows were in three country types. The comparability in the flow data, namely short- first were the economically dynamic economies term movements. However, it tends to hide – Macao (China), Singapore and Hong Kong (China). migration of countries that have both large The second were relatively wealthier countries such immigration and emigration. as Australia, New Zealand and Thailand. The third The data on net migration comes from the were countries such as Bhutan and Timor that had United Nations’ World Population Prospects large emigration flows in previous periods for politi- Database (2008, http://esa.un.org/unpp/) and are cal reasons. Net migration rates hide migration based on “medium variant” population projec- flows of countries that have both large in and out- tions. Data on the migrant stock and on total population are from the World Bank Database migration flows, such as the Philippines, where (http://data.worldbank.org/indicator/SM.POP.TOTL). thousands of increasingly skilled labourers arrive (Figure GE4.) There are vast differences in the and leave the country every year. share of international migrants in the total The share of international migrants in the population across Asia/Pacific countries and total population is higher in richer countries economies. (GE4.2). The wealthiest countries of the Asia/Pacific

44 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 4. GENERAL CONTEXT INDICATORS

GE4.1. Migration

Panel A. Share of migrants in total population, Panel B. Net migration rate 2000-05, percentage, 2005 () per 1 000 in the population 57.0 Macao (China) 17.2 39.9 Hong Kong (China) 3.3 35.0 Singapore 6.7 33.6 Brunei Darussalam 2.0 21.3 Australia 6.5 20.7 New Zealand 5.1 19.6 Kazakhstan -2.7 16.1 Armenia -6.5 10.6 OECD 3.3 60 50 40 30 20 10 8.2 Asia/Pacific -0.7 7.9 Malaysia 1.2 5.7 Bhutan 11.6 5.6 Kyrgyzstan -2.9 4.7 Tajikistan -10.9 4.0 Samoa 20.8 3.0 Azerbaijan -2.4 3.0 Nepal -0.8 2.3 Pakistan -1.6 2.2 Cambodia 0.2 2.1 Fiji -10.3 1.9 Sri Lanka -4.6 1.6 Japan 0.1 1.5 Thailand 4.4 1.2 Timor-Leste 9.1 1.1 Korea, Rep. -0.3 1.1 Tonga -15.9 1.1 Maldives 0.0 0.7 Bangladesh -1.0 0.5 India -0.3 0.4 Philippines -2.2 0.4 Papua New Guinea 0.0 0.4 Mongolia 1.4 0.3 Lao PDR -4.1 0.2 Myanmar -4.2 0.2 Korea, DPR 0.0 0.1 Viet Nam -0.5 0.1 Indonesia -0.9 0.0 China -0.3 45678 312 0 -25 -20-15 -10 -5 0 5 10 15 20

GE4.2. Richer countries have a higher share GE4.3. High migration rates combined of foreign-born population with increases in foreign-born people

2005 GDP per capita (2009 current USD) (log) Variation in the share of migrants (in percentage points), 2000-05 5.0 3.0 NZL MAC BRN SGP 4.5 JPN AUS 2.0 NZL HKG KOR BRN SGP MYS 4.0 1.0 WSM MYS KHM THA FJI KAZ JPN AUS KAZ FJI MMR TLS 3.5 THA MDV 0 NPL HKG BTN WSM TON LKA LAO PHL CHN TON MDV IDN ARM TJK PNG AZE PNG PAK KORVNM 3.0 BTN -1.0 PRK IND IDN BGD LKA AZE MNGCHN PAK IND LAO KHM KGZ PHL TJK MNG KGZ 2.5 NPL -2.0 VNM MMR BGD TLS ARM 2.0 -3.0 0 5 10 15 20 25 30 35 40 45 -25 -15 -5 5 15 Share of international migrants in total population (%) Net migration rate 2000-2005, per 1 000 in the population

Source: The data on net migration comes from the United Nations’ World Population Prospects Database (2008, http://esa.un.org/unpp/) and are based on “medium variant” population projections. Data on the migrant stock and on total population are from the World Bank Database (http://data.worldbank.org/indicator/SM.POP.TOTL). Figure GE4.2: There are vast differences in the share of international migrants in total population across Asia/Pacific countries and economies. 1 2 http://dx.doi.org/10.1787/888932546224

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 45 4. GENERAL CONTEXT INDICATORS

GE5. Support rate

The support rate is the ratio of the popula- keep on increasing in the future at an increasingly tion who are more likely to be economically rapid pace, especially for developing countries. active to children and older people more likely to Declines in the dependency ratio were associ- be economically inactive. The support rate thus ated with higher growth rates in GDP per capita provides a rough indicator of the number of active during the past 15 years. As shown in Figure GE5.3, people who potentially are economically and countries with a higher annual average decrease socially supporting inactive people. It also gives a in the dependency ratio grew faster. In other words, broad indication of the age structure of the popu- rapidly growing countries like China, Viet Nam, lation. Changes in the support rate depend on Mongolia and India may have benefitted from an mortality and fertility rates and, to a much lesser increase in the working-age population, which degree, on net migration. boosted economic activity. However, given that the There are substantial variations in support trend in the dependency ratio is predicted to rates across Asia/Pacific countries and economies reverse in years to come, the heavier burden of (GE5.1). Whereas the number of dependent people dependants is likely to affect economic pros- in economies like Macao (China), Hong Kong pects, especially in China and the Republic of (China) and Singapore is not higher than 35 per Korea, where the dependency ratio is predicted 100 persons, in Nepal, the Lao People’s Democratic to be multiplied by 1.6 and 2.3, respectively, Republic and Pakistan, this number is doubled, between 2010 and 2050. with almost 70 dependants for 100 persons. In terms of the total dependency ratio, OECD and Asian countries and economies do not differ much, Definition and measurement with a ratio of one dependent out of two people. However, the age structure of the dependent Support ratios measure the age structure of a population by calculating the ratio of the num- persons is very different in developing countries ber of individuals who are likely to support compared to OECD countries. In developing coun- themselves for their daily living to the number tries, the total dependency ratio is driven by the of those who are dependent on such support, i.e. youth-dependency ratio, whereas in OECD coun- the children and the elderly. The key indicator tries, the share of old people in the total working of the support ratio relates the combined num- population is higher. This is due to both longer life ber of individuals aged 15 to 64 to the popula- spans (see HE1) and lower birth rates (see GE3) in tion aged less than 15 years old and those the OECD countries, which are producing a popula- aged 65 and over. The combined share of the tion distribution that is slanted towards the elderly. population aged 15-64 as a percentage of the population aged less than 15 and more than 64 At the same time, in the majority of coun- can be read as the ratio between the economi- tries the dependency ratio is likely to increase in cally active population and the economically the future. Whereas the ratio of dependants on non-active population. All ratios are presented the working population decreased in all Asia/ as the number of working age (15-64) people per Pacific countries and economies but Japan during one dependant person. the past 20 years (Figure GE5.2), the trend is going Data come from the United Nations’ World to reverse in the future, with an increase in the Population Prospects online database (2009, http:// dependency ratio for most countries in the region. esa.un.org/unpp/index.asp?panel=2). The projec- According to projections, changes will happen at a tions for age-dependency ratios used in this faster rate in the 40 years to come. Changes in the section are based on the “medium variant” dependency ratio will be driven mostly by the population projections. increase in old-age dependency rates that will

46 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 4. GENERAL CONTEXT INDICATORS

GE5.1. Support rate levels and trends Panel B. Changes in total support ratio, 1990-2010 Panel A. Total support ratio, 2010 () Percentage points

4.0 Macao (China) 1.9 3.1 Hong Kong (China) 0.8 2.9 Singapore 0.2 2.7 Korea, Rep. 0.4 2.6 China 0.6 2.4 Thailand 0.5 2.4 Mongolia 1.2 2.4 Brunei Darussalam 0.7 2.3 Azerbaijan 0.7 2.2 Kazakhstan 0.6 2.2 Korea, DPR 0.0 2.2 Armenia 0.4 2.2 Viet Nam 0.9 2.2 Maldives 1.2 2.1 Sri Lanka 0.5 2.1 Myanmar 0.7 2.1 OECD 0.1 2.1 Indonesia 0.5 2.1 Australia 0.0 2.0 Asia/Pacific 0.5 2.0 New Zealand 0.1 1.9 Malaysia 0.5 1.9 Kyrgyzstan 0.6 1.9 Bhutan 0.7 1.9 Bangladesh 0.7 1.8 India 0.4 1.8 Japan -0.5 1.8 Fiji 0.3 1.8 Cambodia 0.7 1.6 Philippines 0.4 1.5 Tajikistan 0.4 1.5 Nepal 0.3 1.5 Lao PDR 0.4 1.5 Pakistan 0.3 1.4 Papua New Guinea 0.1 1.3 Tonga 0.0 1.3 Samoa 0.1 1.1 Timor-Leste -0.3 5 4 32 10 -1.0 0.5 0 0.5 1.0 1.5 2.0

GE5.2. Point change in support ratio, 2010-50 GE5.3. GDP growth and changes in total Percentage points support ratio Annual average growth rate 1995-2008, GDP per capita 1.0 11 ARM 0.5 9 CHN MMR 0 KHM BTN 7 KAZ VNM -0.5 TJK 5 Asia/Pacific IND MDV -1.0 LAO KGZ MNG OECD SGP LKA BGD 3 KOR MYS WSM NPLPHL HKG -1.5 AUS IDN NZL PAK THA 1 JPN TON -2.0 FJI e a r s s a) a) r n d n ia ka d n ia n ia n n sh e n ga te ea n n n n alia n n n s ta s ta ta Fiji n dia ta n s n ta r golia alam ma e s s u In s i s PNG Chi JapaOECDt me n baija n n To Nepal-Le u amoa BRN gapoea, Rep. us r i La r ea, DPR do gyz Bh glade r S -1 g (Chiin r Thaila A A Viet Namruss r n Maldive Malayr n Paki Lao PDR n S Sr Mo Aze MyaI Cambodia Tajiki Ko Ko Kazakh Ky Ba Philippi Timo New Zeala Asia/Pacific a New G -1.0 -0.5 0 0.5 1.0 1.5 Macaog Ko (Chi ei Da u n run Ho B Pap Changes in total support ratio in % points (1990-2010)

Note: The support ratio is defined as the ratio of the population aged 15-64 per one person of the population aged 0-14 or 65+. Source: UN World Population Prospects, 2008 Revision. 1 2 http://dx.doi.org/10.1787/888932546243

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 47

Society at a Glance Asia/Pacific 2011 © OECD 2011

Chapter 5

Self-sufficiency Indicators

SS1. Employment SS2. Unemployment SS3. Education SS4. Early childhood education SS5. Education spending

49 5. SELF-SUFFICIENCY INDICATORS

SS1. Employment

Employment provides people with income the total employment rate observed in Panel A with which to support themselves and their fam- of SS1.1. Gender gaps in labour force participation ilies. For many people, paid employment is also a are notably small – less than 10 percentage source of meaning, social interactions and life points – in the former Soviet Republics and satisfaction. New Zealand and Australia. Male participation The highest employment rates in the Asia/ rates are much higher than female employment Pacific region are found in Thailand and China, rates in Japan and the Republic of Korea. Especially where more than seven in every ten adults are large gender gaps in participation rates are found employed (Panel A, SS1.1). On the other hand, in countries with low aggregate employment rates fewer than half the adult population are employed like Pakistan and Sri Lanka. in Pakistan and Sri Lanka. OECD Asia/Pacific coun- Participation rates are also lower for younger tries and economies have a wide range of employ- people and older people than for people of prime ment rates, largely reflecting variations in female working age (SS1.3). However, these age gaps employment. differ greatly by country. There are large gaps The employment rate consequences of the between young workers and prime age worker crisis of 2008 have been relatively muted in the participation in Azerbaijan, the Republic of Korea, Asia/Pacific region (Panel B, SS1.1). The larger Singapore and Japan. Participation rates of older falls in employment rates have been concentrated people are very similar to prime age workers in in the richer countries or economies of the region Pakistan, the Philippines and Indonesia. – Australia; the Republic of Korea, Hong Kong (China), Japan and New Zealand. In contrast, a number of poorer countries have managed to Definition and measurement increase their employment rates significantly over the crisis period, particularly Mongolia, Indonesia The employment rate is the ratio of employed and Viet Nam. people over age 15 to the total population over age 15. The data on the headline employment There are large differences between Asia/ rate and the employment rate change come Pacific countries and economies in terms of from Key Indicators for Asia and the Pacific 2010, labour force participation rates for men and Asian Development Bank (August) and were women, which are a key driver of employment derived from the unemployment rate and rate gaps (SS1.2). Low relative female employment labour force participation rate figures. contributes markedly to the observed variation in

50 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 5. SELF-SUFFICIENCY INDICATORS

SS1.1. Employment rates and recent changes

Panel A. Levels, 2009 () Panel B. Percentage point change, 2007/09

71.7 Thailand 0.3 70.5 China -0.5 67.3 OECD -1.6 66.6 Azerbaijan -1.2 66.0 Kazakhstan 0.8 65.8 Bhutan 1.0 65.3 Brunei Darussalam -0.2 64.4 Mongolia 2.0 64.1 New Zealand -2.0 62.7 Singapore -0.4 62.1 Indonesia 2.1 61.6 Australia -0.7 61.1 Kyrgyzstan -0.1 60.6 Malaysia -0.6 60.0 Asia/Pacific 0.0 59.2 Philippines -0.1 58.6 Korea, Rep. -1.2 57.4 Hong Kong (China) -1.3 56.8 Japan -1.2 56.8 Samoa -0.5 54.2 Viet Nam 4.0 53.7 Fiji -0.8 50.4 Armenia -0.3 46.4 Sri Lanka -0.4 43.2 Pakistan 0.4 75 70 65 60 55 50 45 40 35 -2 -1 0 1 2 3 4

SS1.2. Gender gap in labour force SS1.3. Age gap in labour force participation rate (men minus women) participation rate Percentage points Percentage points

Age difference (25-54)/(15-24) () Age difference (25-54)/(55-64) 40 39.3 50 35.3 35 32.4 40 30

25 23.8 30 20.3 20.4 19.1 19.6 20 17.8

15 13.0 20 10.5 10 9.0 9.4 9.5 10 4.4 5 3.3

0 0 n ia d n a) a) es n re ia n ka n d s n e n n n ta alia n n n s ta n n ia e a) a) r s r e s alia ta es n n n baija me t Japa OECD n tr s n r r us gapo i La OECD Japa baija A in ea, Rep.do Paki us do gapo r A g (Chi r In Sr A Paki n g (Chi in rea, Rep. Aze n Philippi S I n S Aze New ZealaKazakh Asia/Pacific Ko Philippi Ko Macao (Chi New Zeala Asia/Pacific g Ho g Ho Macao (Chi n n Ho Ho

Note: Data refer to the population aged 15 years and over. Source: ADB, Asian Development Bank (www.adb.org); Key Indicators for Asia and the Pacific 2010, ILO LABORSTA online database, International Labour Organisation ILO, OECD Labour Force Statistics Database. 1 2 http://dx.doi.org/10.1787/888932546262

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 51 5. SELF-SUFFICIENCY INDICATORS

SS2. Unemployment

Unemployment reduces people’s ability to while in the Maldives, female unemployment is support themselves and their families and more than three times the male rate. makes them reliant on others, especially the Youth unemployment rates are always state benefit system or others in the family. higher than those for prime age workers, but Unemployment also has substantial psychological sometimes older people have lower unemploy- costs for people, leading to lower life satisfaction ment rates (SS2.3). In Sri Lanka, the Maldives and in a permanent way. Indonesia, more than one in every four economi- Unemployment rates are relatively low cally active young people is unemployed. In the amongst Asia/Pacific countries and economies cases of Sri Lanka and Indonesia, the gap in unem- (Panel A, SS2.1). Especially low unemployment ployment rates between young people and prime rates are found on the Pacific Island of Samoa, at age people is large – about 15 percentage points. In under 2%, while the highest rates are found on some cases the unemployment rates of older peo- another Pacific Island, Fiji. Other countries with ple are lower than those of prime age workers. In low unemployment rates include Thailand and fact, in Indonesia the unemployment rate of older Tajikistan. Higher unemployment rates can also workers is over 5 percentage points lower than be found in Kyrgyzstan, and, among the larger that of prime age workers. countries, in Indonesia and the Philippines. As in the OECD, the trends following the 2008 economic crisis have pushed unemployment Definition and measurement rates up in most countries (Panel B, SS2.1). As with The unemployment rate is the ratio of people employment changes in the opposite direction, the out of work and actively seeking it to the work- larger rises in unemployment rates have been ing-age population either in work or actively concentrated in the richer countries or economies seeking it (15-64 years old). of the region – Australia, Hong Kong (China), Japan, The data are gathered through labour force New Zealand and Singapore. In contrast, Indonesia surveys of member countries. According to the has reduced its unemployment rate over the crisis standardised ILO definition used in these sur- period, as have five other countries, but by much veys, the unemployed are those who did not lesser amounts. work for at least one hour in the reference week of the survey but who are currently available for Women generally have higher unemploy- work and who have taken specific steps to seek ment rates than men (SS2.2). However, in many employment in the four weeks preceding the countries and economies, such as Malaysia, this survey. Thus, for example, people who cannot gender gap is negligible or even reversed, as in work because of physical impairment, or who Hong Kong (China). Generally speaking, the higher are not actively seeking a job because they have the unemployment rate, the larger the gender gap little hope of finding work, are not considered to the disadvantage of women. In Pakistan, the as unemployed. female unemployment rate is double that of men,

52 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 5. SELF-SUFFICIENCY INDICATORS

SS2.1. Unemployment levels and trends

Panel A. Levels, 2009 () Percentages Panel B. Percentage point change, 2007/09

1.3 Samoa 0.0 1.5 Thailand 0.1 2.1 Tajikistan -0.3 2.4 Viet Nam 0.4 3.6 Korea, Rep. 0.4 3.6 Mongolia 0.8 3.7 Brunei Darussalam 0.3 3.7 Malaysia 0.5 4.0 Bhutan 0.3 4.0 Myanmar 0.0 4.1 Singapore 1.2 4.3 China 0.3 4.9 Asia/Pacific 0.3 5.1 Japan 1.2 5.4 Hong Kong (China) 1.4 5.5 Pakistan 0.2 5.6 Australia 1.2 5.6 Sri Lanka -0.4 6.0 Azerbaijan -0.5 6.1 New Zealand 2.5 6.6 Kazakhstan -0.7 6.9 Armenia -0.1 7.5 Philippines 0.2 8.1 Indonesia -1.7 8.1 OECD 2.5 8.2 Kyrgyzstan 0.1 8.6 Fiji 0.9 9 87 65432101-2 -1 0 2 3

SS2.2. Unemployment rates by gender SS2.3. Unemployment rates by age Percentages Percentages

Total () Male (%) Female (%) Total (15-24) () (55-64) 25 30 24

25 25 20 22 21 20 15 19

13 15 15 12 14 11 11 10 12 12 10 11 9 9 9 10 8 8 8 10 9 7 8 9 9 9 9 7 6 6 7 8 8 6 6 6 6 7 5 5 6 6 7 5 5 5 5 5 6 6 4 4 4 4 4 5 5 4 4 4 4 5 5 3 3 3 4 3 3 4 3 4 4 3 3 3 3 3 1 3 3 3 3 2 1 2 2 1 1 1 2 0 0 d n n n e h d n s s d n d n e u h n s n n s n a) r a) s n ia e ka ia ga n a) ea ia n a) r s e ka ia ia n ta ta n alia s n n es n n ta r s n alia ta n n ta Fijies es u s tr n amoa u golia tr s ribati is n n Japa gapo OECDrbaija To Ko n Japa Pala OECD baijan Bh ea, Rep.n glade us i La do S Bh gapo us glade Ki i La r do Thaila Paki r i g (Chin A Malay Sr Maldiven Thaila Malay g (Chiin A n Paki n Maldive S n Aze I Mo n S Sr Aze I Ko Ba Philippi Ba Philippi Afgha New Zeala Asia/Pacific New Zeala Asia/Pacific ch Poly Macao (Chi g Ho Macao (Chi g Ho en n n r F Ho Ho

Note: Data refer to the population aged 15 years and over. Source: ADB, Asian Development Bank (www.adb.org); Key Indicators for Asia and the Pacific 2010, International Labour Organisation ILO LABORSTA online database, OECD Labour Force Statistics Database. 1 2 http://dx.doi.org/10.1787/888932546281

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 53 5. SELF-SUFFICIENCY INDICATORS

SS3. Education

The level of education of the population Australia. It is likely that convergence in human gives an indication of its stock of human capital. capital levels is one important – but probably not A higher stock of human capital means higher the only – factor behind the convergence in per labour productivity and hence higher income- capita GDP observed above. generating capacity. Better education is also asso- Women tend to have fewer years of educa- ciated with better health of both adults and chil- tion than the population as a whole (SS3.3). There dren. The most readily comparable measure has been considerable catch-up in schooling out- available across the wide range of countries cov- comes by gender. However, the stock of human ered is the total years of education of the working- capital represents investments over a period of age population. half a century, and thus includes periods when New Zealanders have three times the num- inequality was much greater than today. The ber of years of schooling of Nepalese (Panel A, Philippines is an exception, as women have more SS3.1). Other OECD Asia/Pacific countries and education than the population average, while the economies have similarly high levels of schooling, years are equal in New Zealand. Gender differ- well in excess of ten years. India, the Lao People’s ences are rather more pronounced to the disad- Democratic Republic, Papua New Guinea and vantage of women in India, Pakistan and China. Myanmar all have around five years of schooling or less. Although Tonga and Fiji are relatively low GDP countries (GE1), they stand out for their Definition and measurement comparatively high levels of human capital. Given the wide variety of countries at very dif- There has been a tendency in the last ferent stages of development that are consid- 30 years across all Asia/Pacific countries and ered here, the core measure of education likely economies for average years of schooling to rise to have the greatest comparability is years of (Panel B, SS3.1). The average rise has been well over education. The data on years of education, and two years per person across the region. These rises years of women’s education, were taken from have been less than one year in New Zealand, the Barro-Lee data set (Barro and Lee, 2010). The data set did not directly include years of educa- Australia and Cambodia, but four years in Fiji and tion for men. Malaysia. As with GDP per capita (GE1), there is evi- dence of a convergence in years of education across countries (SS3.2). Countries with low years Further reading of schooling in 1980, such as Pakistan and India, Barro, R. and J.W. Lee (2010), “A New Data Set of have tended to experience more rapid growth in Educational Attainment in the World: 1950-2010”, years of education than countries that had high NBER Working Paper, No. 15902, Cambridge, MA, years of schooling in 1980, like New Zealand and www.nber.org/papers/w15902.

54 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 5. SELF-SUFFICIENCY INDICATORS

SS3.1. Education levels and changes

Panel B. Change in average years of total schooling, Panel A. Average years of total schooling, 2010 () 1980-2010 (in years)

12.7 OECD 12.7 New Zealand 0.9 12.1 Australia 0.6 11.8 Korea, Rep. 3.6 11.6 Japan 2.3 11.3 Fiji 4.3 10.5 Tonga 2.3 10.4 Armenia 1.0 10.4 Kazakhstan 3.8 10.4 Hong Kong (China) 2.4 10.1 Malaysia 4.5 9.3 Tajikistan 1.2 9.1 Singapore 3.9 9.0 Philippines 2.3 8.7 Kyrgyzstan 1.7 8.5 Brunei Darussalam 1.9 8.4 Sri Lanka 1.8 8.4 Mongolia 2.0 8.3 Asia/Pacific 2.5 8.2 China 3.4 8.2 Macao (China) 3.3 7.5 Thailand 3.1 6.4 Viet Nam 1.1 6.2 Indonesia 2.3 6.1 Maldives 1.6 6.0 Cambodia 0.7 5.8 Bangladesh 3.5 5.6 Pakistan 3.4 5.1 India 2.8 5.1 Lao PDR 2.3 4.8 Papua New Guinea 2.6 4.8 Myanmar 2.6 4.0 Nepal 3.0 15 10 5 0 0 1 2 3 45

SS3.2. Convergence in years of schooling SS3.3. Gender difference in years of schooling

Change in average years of schooling, 1980-2010 (in years) Average years of total schooling, 2010 5 Average years of total schooling for women, 2010 () MYS FJI 14 12.5 12.712.7 4 SGP 12.1 KAZ 12 BGD CHN KOR PAK MAC NPL 10 3 9.2 IND THA 9.0 PNG Asia/Pacific 8.2 8.3 MMR HKG 8.1 PHL JPN 8 7.6 IDN MNG 2 LAO TON BRN LKA 6 5.6 MDV KGZ 5.1 4.1 4.3 VNM TJK 4 1 ARM NZL KHM AUS 2 n a s d dia n e n 0 n sta n ralia I Chi t 0 5 10 15 us Paki A Philippi Average years of schooling in 1980 Asia/Pacific New Zeala

Note: SS3.1: Data for the OECD refers to the number of years at which over 90% of the population are enrolled. Change in average years of total schooling 1980-2010 (years) is not available for the OECD. Source: Barro and Lee (2010); OECD (2010), Education at a Glance: OECD Indicators. 1 2 http://dx.doi.org/10.1787/888932546300

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 55 5. SELF-SUFFICIENCY INDICATORS

SS4. Early childhood education

Public investment in early childhood educa- with a slightly higher attendance of boys. A more tion indicates the commitment within a society noticeable difference is in Kazakhstan where the to equip its children with the necessary skills to attendance difference is 3.7% in favour of boys. support their future lives. Early childhood educa- Attendance rates between urban and rural areas tion also allows women to balance motherhood also differ. Except for Bangladesh, more children with paid employment, should they so chose, attend early childhood education programmes in and thus gives an indication of women’s self- urban areas. sufficiency. Higher rates of early childhood education The percentage of pre-school children and care are associated with lower rates of infant attending an early education programme differs mortality (SS4.2). It is likely that this relationship significantly across countries (Panel A, SS4.1). is observed because richer countries invest more From the available data, the lowest attendance is publicly and privately in young children, and observed in the Lao People’s Democratic Republic this investment shows up both in lower infant (total 3%) and the highest in New Zealand (95%). In mortality and in higher early childhood education general, the higher-income countries of the region participation. have a higher attendance of children in ECE programmes. However, Maldives, Thailand and Viet Nam, all relatively low-income countries, are Definition and measurement not far behind Australia, a higher-income country The data on early childhood education (ECE) in terms of participation. participation come from UNESCO (2006) and the OECD Family Database (for the four OECD mem- Attendance in early childhood education has ber countries). There are a number of caveats been rising in most countries in the recent past attached to the data for the non-OECD countries (Panel B, SS4.1). The rises were largest in India, of the region, for which the UNESCO publication Nepal and Kazakhstan. Attendance rates actually should be consulted. The data come from a fell in Bangladesh and Tonga. variety of sources for a variety of years, in many Gender differences in pre-school attendance cases cover slightly different age groups, and in most countries are insignificant (SS4.2). For the consequently are only broadly comparable majority of countries, there are very minor differ- between countries. ences in attendance by gender, and when such differences do exist, there is a higher attendance of girls. The exceptions are Maldives, Azerbaijan, Further reading Tajikistan and the Lao People’s Democratic Republic UNESCO (2006), Strong Foundations, UNESCO, Paris.

56 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 5. SELF-SUFFICIENCY INDICATORS

SS4.1. Early childhood education levels and trends Early childhood education participation between 3 and 6 years, mid-2000s

Panel A. Most recent participation rates, % () Panel B. Percentage points change, 1999 to mid-2000s

94 New Zealand 92 Macao (China) 3 90 Thailand 2 88 Japan 80 Korea, Rep. 77 OECD 59 Australia 52 Brunei Darussalam 1 49 Samoa -2 48 Maldives 2 47 Viet Nam 6 41 Asia/Pacific 36 China -2 36 India 16 36 Nepal 25 35 Mongolia 10 31 Armenia 5 31 Kazakhstan 16 28 Azerbaijan 6 23 Tonga -7 22 Indonesia 4 16 Fiji -1 12 Bangladesh -14 12 Kyrgyzstan 2 11 Timor-Leste 9 Cambodia 3 9 Tajikistan 1 8 Lao PDR 0 100 80 60 40 20 0 -20 -10 0 10 20 30

SS4.2. Early childhood education participation and infant mortality

Early childhood education participation (%) 100 NZL MAC 90 JPN THA 80 KOR OECD 70

60 AUS BRN 50 PHL VNM MDV Asia/Pacific 40 NPL CHN IND MNG AMR KAZ 30 AZE TON 20 IDN FJI KGZ BDG TMP 10 LAO KHM TJK 0 0 20 40 60 80 Infant mortality rate (per 1 000 live births)

Source: UNESCO (2006), Strong Foundations;OECD Family Database; OECD Education Database. 1 2 http://dx.doi.org/10.1787/888932546319

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 57 5. SELF-SUFFICIENCY INDICATORS

SS5. Education spending

Education spending represents society’s education. While some poorer countries seem to response to providing its children with the social have chosen a strategy of higher public spending and economic skills needed to be self-sufficient on education, like the Maldives, other poorer in life. Countries spend different amounts on edu- countries, like Myanmar, spend almost nothing on cation. They also vary the resources committed in public education. These are very different develop- different ways across the child’s educational life ment strategies. cycle, a phenomenon that is not examined here. There is only a weak positive relationship Public education spending as a share of GDP between public spending on education and aver- differs greatly across the Asia/Pacific region age years of schooling (SS5.3). Cambodia and the (Panel A, SS5.1). The Maldives, Timor and Kyrgyz- Maldives have very similar average years of stan spend a large amount of their GDP on public schooling, but the Maldives’ share of GDP invested education, while Azerbaijan, Cambodia and Myan- in education is about four times larger. The lack of mar spend less than one-quarter of this amount. a stronger relationship could be because educa- Public education spending as a share of GDP in tion spending is changing the stock of average the region can change considerably over a years of education at the margin. relatively short period. There is no dominant overall pattern across the region of increases or decreases in the GDP share spent on education. Larges rises are observed in Kyrgyzstan, of 3 percentage points of Definition and measurement GDP, and large falls in adjacent Azerbaijan, of a Data on public education spending as a per- similar amount. centage of GDP was extracted from UNESCO There is no link between GDP per capita and Institute for Statistics, http://stats.uis.unesco.org/ public expenditure on education (SS5.2). Richer unesco/TableViewer/tableView.aspx. countries do not tend to spend higher shares on

58 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 5. SELF-SUFFICIENCY INDICATORS

SS5.1. Education spending levels and trends

Panel A. Public expenditures on education as % of GDP, Panel B. Change in the GDP share of spending on education, latest year available () (early to late 2000s)

8.1 Maldives -0.1 7.1 Timor-Leste 6.6 Kyrgyzstan 2.5 6.2 Fiji 1.0 6.2 New Zealand -0.7 5.4 Samoa 0.9 5.3 Viet Nam 5.2 OECD 0.1 5.1 Bhutan -0.7 5.1 Mongolia -1.0 4.9 Thailand -0.1 4.7 Tonga -1.9 4.7 Australia -0.4 4.5 Malaysia -1.2 4.2 Korea, Rep. 0.4 4.0 Asia/Pacific -0.1 3.8 Nepal 0.9 3.7 Brunei Darussalam -1.2 3.5 Indonesia 1.1 3.5 Tajikistan 1.4 3.4 Japan -0.1 3.3 Hong Kong (China) -0.6 3.2 Singapore 0.1 3.2 India -1.3 3.0 Armenia 0.8 2.9 Pakistan 0.3 2.8 Kazakhstan -1.1 2.6 Philippines -0.9 2.4 Bangladesh 0.0 2.3 Lao PDR 1.3 2.2 Macao (China) -1.5 1.9 China 1.9 Azerbaijan -2.3 1.6 Cambodia 0.6 1.3 Myanmar 0.6 8 6 4 2 0 -3 -2 -1 0 1 2 3

SS5.2. Rich countries do not spend relatively SS5.3. Spending on education and average more on education years of schooling correlate only weakly

Log of GDP per capita Average number of years of schooling (15+ years old), 2010 4.8 15 AUS 4.6 SGP JPN BRN HKG NZL 13 NZL 4.4 KOR KOR AUS JPN FJI 4.2 11 KAZ ARM HKG TON 4.0 Asia/Pacific SGP MYS KAZ TJK MYS 9 PHL BRN KGZ 3.8 MAC MNG CHN Asia/Pacific THA THA 3.6 FJI MDV 7 CHN BGD IDN VNM WSM KHM MDV 3.4 AZE TON PAK BTN IND PHL IDN 5 LAO 3.2 MNG MMR ARM NPL PAK IND 3.0 3 082 4 6 10 082 4 6 10 Expenditure on education as % of GDP Expenditure on education as % of GDP

Source: UNESCO Institute for Statistics, http://stats.uis.unesco.org/unesco/TableViewer/tableView.aspx; OECD (2010), Education at a Glance. 1 2 http://dx.doi.org/10.1787/888932546338

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 59

Society at a Glance Asia/Pacific 2011 © OECD 2011

Chapter 6

Equity Indicators

EQ1. Poverty EQ2. Income inequality EQ3. Gender ratio EQ4. Pensions EQ5. Social spending

61 6. EQUITY INDICATORS

EQ1. Poverty

Poverty measures the relative numbers of peo- poverty reduction underlines the variety of poverty ple at the bottom end of the income distribution. elasticities to growth. Thus in Viet Nam the poverty- The equity concerns of a society tend towards the reducing effect of growth was higher than in China. disadvantaged. The focus here is on absolute pov- erty, using a common poverty line across countries. Almost half of the population living in Asia/ Definition and measurement Pacific countries and economies is poor, and one- Poverty is commonly measured by using fifth is extremely poor (Panel A, EQ1.1). More than income or consumption levels. A person is con- 80% of the region’s poor live in Bangladesh, India sidered poor if his or her consumption or income and China. The poorest country is Bangladesh, level falls below a predetermined poverty line, where more than 80% of the population live with which corresponds to a minimum level necessary less than USD 2 a day. Poverty levels are the lowest to meet basic needs. At the international level, in Azerbaijan and Kazakhstan (the four OECD Asia/ two reference poverty lines are set at USD 1.25 Pacific countries and economies are not included and USD 2 per person per day (2005 purchasing here, as their rate is zero. Singapore, Macao (China) power parity). The USD 1.25 poverty line corres- and Hong Kong (China) economies are excluded for ponds to the mean of national poverty lines for the 10-20 poorest countries of the world, while the same reason). the USD 2 line is the median poverty line found Despite important disparities in the pace of among developing countries as a whole. People poverty reduction, poverty rates decreased in all living with less than USD 1.25 a day are said to be the Asia/Pacific countries and economies over in extreme, or absolute, poverty. Data on under- the last decade (Panel B, EQ1.1). Kyrgyzstan, the nourishment is also shown here. Undernourish- Lao People’s Democratic Republic and Malaysia ment refers to the condition where people’s witnessed the slowest poverty decline. Declines dietary energy consumption is continuously were highest for Azerbaijan, Viet Nam, Mongolia below the minimum dietary energy requirement and China. for maintaining a healthy life and carrying out Absolute poverty is a measure of inability to light physical activity with an acceptable mini- satisfy subsistence needs, including the need for mum body-weight for an attained height. nourishment. However, the share of undernour- The data presented here are the percentage of ishment is not strongly correlated with the share people living with a level of income or consump- of the population living under the USD 2 poverty tion that falls under the poverty lines. Data on the line (EQ1.2). Thus poverty is not synonymous with population share living below national and/or hunger. This implies that there may be space for universal poverty lines are available only for the social policies focused on food security. Some least developed economies. Data is from the countries, such as Azerbaijan, Thailand, Armenia World Bank’s World Development Indicators online and Mongolia, have comparatively low levels of database (2010, http://data.worldbank.org/indicator) and from the Key Indicators for Asia and the poverty but worryingly high levels of undernour- Pacific 2010 of the Asian Development Bank. Data ishment – even higher than the poverty rate. This are based on household surveys or obtained from means that some people who are not considered government statistical agencies and World Bank poor are undernourished. country departments. Given that poverty is mea- Declining poverty was more rapid in countries sured based on either income or expenditure, with higher GDP growth (EQ1.3). The pace of both headcount ratios are not strictly comparable growth and poverty reduction was fastest in Azer- across countries. Data on undernourishment is baijan, where GDP per capita multiplied by four from the Food and Agriculture Organization of the between 1995 and 2008 and the poverty rates United Nations (www.fao.org/economic/ess/food- declined by more than 35 percentage points. How- security-statistics/en/). ever, the imperfect correlation between growth and

62 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 6. EQUITY INDICATORS

EQ1.1. Poverty rates and trends

Panel A. Almost half of the population living Panel B. Over the past decade, poverty declined in Asia/Pacific countries and economies is poor in all Asia/Pacific countries and economies Headcount ratio at the USD 1.25 and USD 2 () poverty line Percentage point decline in poverty rate (latest year available), % (USD 2 poverty line) USD 2 a day () USD 1.25 a day 81 50 Bangladesh -6.1 78 55 Nepal -10.6 77 53 Timor-Leste -4.6 77 44 Lao PDR -3.1 76 42 India -6.1 68 40 Cambodia -20.0 60 23 Pakistan -3.5 57 36 Papua New Guinea 55 29 Indonesia -23.4 52 22 Kyrgyzstan -2.6 51 21 Tajikistan -27.7 49 26 Bhutan 49 25 Asia/Pacific -15.0 49 22 Mongolia -29.9 -37.3 48 21 Viet Nam 45 23 Philippines -7.6 43 11 Armenia -17.9 40 14 Sri Lanka -6.9 36 16 China -28.8 17 3 Kazakhstan -16.8 12 Thailand 2 -5.9 8 Malaysia 2 -3.2 -37.4 2 Azerbaijan 2 100 80 60 40 20 0 0 -10 -20 -30 -40

EQ1.2. In some countries, the prevalence EQ1.3. Higher growth rates are linked of undernourishment persists despite with faster poverty reduction low levels of poverty The size of the bubble is proportional to the number The size of the bubble is proportional to the number of poor people of poor people Share of population living with less than USD 2 per day % points decline in poverty rate (1995-latest year available) 90 0 MYSKGZ BGD PAK LAO 80 -5 NPL IND LAO THA BGD IND TLS LKA 70 PHL -10 NPL PAK 60 KHM IDN -15 TJK KAZ 50 VNM ARM PHL -20 KHM 40 LKA CHN IDN -25 30 TJK KGZ CHN -30 MNG 20 ARM MNG 10 THA -35 VNM AZE KAZ AZE 0 -40 5 10 15 20 25 30 35 0 50 100 150 200 250 300 Share of undernurished people (%) Change in GDP per capita (%) 1995-2008

Note: EQ1.1, Panel A: Timor-Leste: 2001 data; the Lao People’s Democratic Republic and Sri Lanka: 2002 data; Armenia, Bhutan, Kazakhstan and Uzbekistan: 2003 data; Cambodia, Kyrgyzstan, Malaysia, Nepal, Tajikistan and Thailand: 2004 data; Azerbaijan, Bangladesh, China, India, Mongolia and Pakistan: 2005 data; and Philippines and Viet Nam: 2006 data. Definition: Population below USD 1.25 a day is the percentage of the population living on less than USD 1.25 a day at 2005 international prices. As a result of revisions in PPP exchange rates, poverty rates for individual countries cannot be compared with poverty rates reported in earlier editions. Data showing as 2.0 signifies a poverty rate of less than 2.0%. Source: EQ1.1: World Bank (2009), World Development Indicators online database, http://ddp-ext.worldbank.org/ext/DDPQQ/ member.do?method=getMembers&userid=1&queryId=6. World Bank, Development Research Group. Data are based on primary household survey data obtained from government statistical agencies and World Bank country departments. Data for high-income economies are from the Luxembourg Income Study Database. EQ1.2 and EQ1.3: Food and Agriculture Organization (FAO), www.fao.org/economic/ess/food- security-statistics/en/. 1 2 http://dx.doi.org/10.1787/888932546357

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 63 6. EQUITY INDICATORS

EQ2. Income inequality

Income inequality indicates how material resources are distributed across society. Some con- Definition and measurement sider high levels of income inequality to be morally OECD measures of inequality are based on undesirable. Others believe that income inequality is income. For many developing countries, where bad because it causes conflict, limits co-operation or most people are self-employed in agriculture or creates psychological and ultimately physical casual labourers, income data is often not stresses. Often the policy concern is more for the relevant or non-existent. For most countries, direction of changes in inequality, rather than for its inequality measures are expenditure-based. Thus country comparisons should be made level. with caution, as expenditure-based measures Income inequality is high in Asia/Pacific typically show lower inequality than do countries and economies compared to the OECD income-based measures. (Panel A, EQ2.1). The Gini coefficient is the most The main distributive indicator used is the unequal in Indonesia, India and Hong Kong (China). Gini coefficient. Values of the Gini coefficient In India the Gini is as high as 54, which is 50% higher range between 0 in the case of “perfect equality” than the OECD average. The Gini is most equal in the (each person gets the same income) and 1 in the countries of the former Soviet Union – Azerbaijan, case of “perfect inequality” (all income goes to Armenia and Kazakhstan. Expenditure-based the share of the population with the highest inequality measures are much lower than income- income). Also used is the inter-decile ratio, based measures, as expected. Inter-decile ratios i.e. the ratio between the upper limit of the measure income differences as a ratio between the 9th decile and that of the 1st decile. Data on poorest 10% and the richest 10% of the population. In inter-decile income/expenditure ratios comes from the World Bank Development Research countries and economies like Hong Kong (China), Group (http://data.worldbank.org/indicator). Gini Singapore, Papua New Guinea or Bhutan, society coefficients for OECD countries come from OECD tends to be extremely polarised between rich and (2008). Data for India, China, Hong Kong (China), poor. In countries with the fewest inequalities, such Macao (China), Fiji, Singapore and Indonesia as Armenia, Kazakhstan, Bangladesh, Pakistan, were derived from government reports or pub- Timor-Leste and the Lao People’s Democratic lished studies. For other countries, data comes Republic, the richest 10% have around six times from the 2010 Asian Development Bank’s report higher income than the poorest 10%. In Azerbaijan, Key Indicators for Asia and the Pacific 2010. Data they gain only three times more than the comparability across countries may be affected poorest 10%. by the variation in sample sizes, the choice of an income/expenditure measure, the choice of Over the last decade, income inequality income pre/post transfers and taxes, inequalities increased in most countries (Panel B, EQ2.1). among individuals/households, and the choice Inequality increased in some of the poorest of the equivalence scale used to attribute each countries in the region – like Nepal, Cambodia person of a household an income when the and Sri Lanka – as well as two rich economies – measure of inequalities is among individuals. Singapore and Hong Kong (China). Big declines in inequality took place in the three former Soviet Republics of Kyrgyzstan, Azerbaijan and Armenia. There is no clear country-correlation between economic growth and changes in inequalities (EQ2.2). However, when one allows for population Further reading sizes and discards some small country outliers, OECD (2008), Growing Unequal? Income Distribution there is some suggestion of a positive relationship and Poverty in OECD Countries, OECD Publishing, between GDP growth and rises in inequality. Paris, www.oecd.org/els/social/inequality.

64 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 6. EQUITY INDICATORS

EQ2.1. Inequality levels and trends

Panel B. In most countries, the level of inequalities Panel A. Inequalities are higher in Asia/Pacific countries increased but changes were the greatest in countries and economies where inequalities declined Gini coefficient and ratio D9/D1, latest year available Change in Gini (mid-1990s-latest) Gini coefficient () D9/D1 Percentage point change in inter-decile ratio

57.0 -5.0 9.6 Indonesia 53.6 8.6 India 52.5 4.9 17.8 Hong Kong (China) 50.9 21.5 Papua New Guinea 50.0 Fiji 47.8 5.8 17.7 Singapore 47.3 9.6 14.8 Nepal 46.7 17.2 Bhutan 46.2 5.1 13.2 China 46.0 Tonga 45.0 Macao (China) 44.2 5.9 13.7 Cambodia 44.0 1.2 14.1 Philippines 43.0 Samoa 42.5 -0.9 13.1 Thailand 41.1 5.7 11.7 Sri Lanka 39.6 -1.3 9.7 Asia/Pacific 37.9 -10.6 11.0 Malaysia 37.8 2.1 9.7 Viet Nam 37.4 10.8 Maldives 36.6 3.4 9.6 Mongolia 33.6 2.1 8.3 Tajikistan 33.5 9.1 New Zealand 0.0 33.5 -20.2 7.1 Kyrgyzstan 32.6 7.3 Lao PDR -2.3 32.1 10.1 Japan -0.2 31.9 6.9 Timor-Leste 31.2 10.1 Korea, Rep. 31.2 0.9 6.7 Pakistan 31.0 7.9 OECD 0.0 31.0 0.4 6.2 Bangladesh 30.9 -4.4 6.9 Kazakhstan 30.2 -14.2 6.8 Armenia 30.1 -0.8 7.2 Australia 16.8 -18.1 2.9 Azerbaijan 60 40 20 0 -50 -40 -30 -20 -10 0 10

EQ2.2. Growth and inequality are unrelated The size of the bubble is proportional to the population

Expenditure based Income based Point change in Gini 14

9 NPL SGP LKA KHM HKG CHN 4 MNG BGD TJK VNM FJI PAK PHL -1 THA LAO KAZ -6 IDN

-11 MYS ARM -16 AZE KGZ -21 024681012 Average annual in GDP per capita 1995-2008

Source: OECD (2008), Growing Unequal? – Income Distribution and Poverty in OECD Countries and OECD (2011), The Causes of Growing Inequalities; World Bank Development indicators, http://data.worldbank.org/. 1 2 http://dx.doi.org/10.1787/888932546376

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 65 6. EQUITY INDICATORS

EQ3. Gender ratio

Missing females are measured by the The picture through time in the Republic of shortfall in the number of females relative to Korea follows a pattern that is broadly similar to the number that would be expected from nor- China but less accentuated until the sharp rise in mal birth ratios. The gender shortfall is believed the late 1980s (EQ3.3). From 1990 to 2000 the sex to take place through sex-selective abortion, female ratio falls as rapidly as it has risen. However, from infanticide or the unequal allocation of health that point, the ratio appears to have started to drift care and food towards males. The natural male- up again. to-female ratio at birth is between 105 and 107 males per 100 females. This ratio, however, tends to even out and then skew in favour of females as population cohorts age due to the Definition and measurement greater biological resilience of females. Data from the United Nations Department of The most acute demographic imbalance Economic and Social Affairs was utilised for the amongst the youngest age cohort of 0-4 years is population demography values and to compute in China (Panel A, EQ3.1). China has 122 boys the sex ratios. The population demography data is aged 0-4 for every 100 girls. High ratios of young presented annually with the number of females boys to girls are also observed in Armenia and and males broken down into 5-year age groups up Azerbaijan. Ratios in excess of the upper range of to 80+. Since the age group “80+” is already a good birth norms are also observed in the Republic of representation of the older cohort, data beyond Korea, Hong Kong (China) and India. this benchmark was not disaggregated further. The sex ratio is computed mainly for the 0-4 age The last 20 years have witnessed a general group since that presents a better reflection of the tendency across virtually all countries for the sex gender bias. Even though typically more boys are ratio to rise (Panel B, EQ3.1). This rise may be due born than girls, the boys have a higher childhood to the fact that large cross-country reductions in mortality rate, which virtually cancels out the infant mortality over that period make it more numerical advantage at birth by the age of five likely that more biologically fragile male children (Hudson and Boer, 2004). The sex ratios calculated survive. The chart also shows that the unequal sex on an aggregate level present an overall picture of ratios in China, Armenia and Azerbaijan have unbalanced gender proportions. arisen within the last generation. There is a sharp fall in the sex ratio in China in the early 1950s, followed by a gradual drift upwards until the mid-1980s (EQ3.2). From the Further reading late 1980s there is an extremely rapid rise in Hudson, V. and A.M. Den Boer (2004), Bare Branches: missing females for the youngest age cohort. The The Security Implications of Asia’s Surplus Male relationship seems to have peaked and may again Population, MIT Press, Cambridge, MA. be in decline.

66 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 6. EQUITY INDICATORS

EQ3.1. Gender ratios and changes

Panel A. Gender ratio for the population aged 0-4 years Panel B. Changes in gender ratios of the population Number of males per 100 females aged 0-4 years between 1985 and 2005 122 China 117 Armenia 11.7 15.4 116 Azerbaijan 10.9 109 India 1.3 108 Hong Kong (China) 0.7 108 Korea, Rep. 0.0 108 Samoa 1.8 108 Papua New Guinea 1.0 107 Singapore 1.5 107 Tonga 0.8 107 Brunei Darussalam 2.0 106 Asia/Pacific 1.7 106 Fiji 0.6 106 Malaysia -0.2 106 Pakistan -0.1 106 Japan 0.7 106 New Zealand -0.1 105 Australia 0.4 105 OECD 0.2 105 Nepal -0.5 105 Korea, DPR 0.6 105 Kazakhstan 1.4 105 Macao (China) -0.3 105 Philippines 0.6 105 Thailand 0.3 105 Timor-Leste -0.3 104 Cambodia 1.9 104 Viet Nam 1.2 104 Mongolia 0.0 104 Tajikistan 1.1 104 Lao PDR 0.9 104 Indonesia 0.8 104 Sri Lanka 0.6 104 Bangladesh 0.5 103 Maldives 0.2 103 Bhutan 0.6 101 Myanmar 0.1 124 120 116 112 108 104 100 -1 01 2 3 4 5 6 7 8 910 11 12 13 14 1516

EQ3.2. Gender ratio of the youngest age EQ3.3. Gender ratio of the youngest age cohort in China, 1950-2009 cohort in the Republic of Korea, 1950-2009

Gender ratio, 0-4 age cohort Gender ratio, 0-4 age cohort 125 114

112 120 110

115 108

106 110

104 105 102

100 100 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005

Source: United Nations, Department of Economic and Social Affairs, OECD Population Statistics. 1 2 http://dx.doi.org/10.1787/888932546395

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 67 6. EQUITY INDICATORS

EQ4. Pensions

Old-age pensions are a means through replicated in pension inequality – include Singapore which societies materially provide for older peo- (where replacement rates are very low, only about ple during their retirement. They are a partial 10% of earnings), Brunei, Sri Lanka, Thailand and substitute for individual and family provision in Viet Nam. Lastly, China, Viet Nam and Pakistan old age. There are two important dimensions of noticeably combine very low pension coverage pensions here. The first is the proportion of the with relatively high replacement rates. population who are covered by a pension scheme. Countries where more adult males live with The second is the extent to which the pension their parents have lower pension coverage paid replaces previous earnings. (EQ4.2). If adult males living with their parents can Pension coverage varies from 90% of the be taken as an indicator of inter-generational labour force in Japan, Australia and New Zealand adult family bonds, it would seem that family pro- to well under 10% in Pakistan, Nepal and vision of material support for the older generation Bangladesh (Panel A, EQ4.1). The latter are all may well be greater in the countries that have countries that typically rely heavily on family pro- lower coverage. vision for the material needs of older people. Countries with a higher proportion of people Pension provision is always significantly lower for living on under USD 2 per day have lower pen- the working-age population than for the labour sion coverage (EQ4.3). It is likely that countries force. Pension coverage in the Republic of Korea is with high amounts of absolute poverty have low only about one in every two people active in the coverage for a number of reasons, including cost labour market, much lower than its fellow OECD and low amounts of paid employment and hence member countries. This places considerable stress of marketisation of income available to be taxed on those, especially women, who find themselves for pension purposes. Additionally, it may be that working, providing for elderly relatives and under low pension coverage causes high old-age poverty pressure also to have children. China, which is rates. rapidly ageing, has very low pension coverage and will therefore rely heavily on family provision for older people in the coming years. Definition and measurement There is much less variability in replacement Data on pension coverage is sourced from the rates than in pension coverage (Panel B, EQ4.1). World Bank Database. Replacement rates come Typically replacement rates are higher for lower from OECD Pensions at a Glance Asia/Pacific 2009. earners in most countries, which will cause reduc- Data on adult males living with parents is tions in inequality amongst older people. The described in GE3 and poverty data in EQ1. exceptions – where earnings inequality is simply

68 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 6. EQUITY INDICATORS

EQ4.1. Pension coverage and replacement of earnings

Panel A. Coverage of the pension system (), latest year available

% of labour force Panel B. Gross replacement rates, 2006 % of working age population Average earners Low earners 95 34 75 Japan 47 93 39 72 New Zealand 78 91 42 70 Australia 69 82 58 64 OECD 72 78 38 56 Hong Kong (China) 35 63 13 45 Singapore 13 49 45 34 Korea, Rep. 67 48 34 Brunei Darussalam 47 32 33 Malaysia 32 42 29 Kyrgyzstan 41 29 Azerbaijan 40 28 Armenia 39 47 28 Asia/Pacific 60 35 48 22 Sri Lanka 48 34 27 Kazakhstan 34 21 Mongolia 27 68 14 Philippines 95 26 50 21 Thailand 50 24 16 Maldives 21 68 17 China 88 14 68 11 Viet Nam 68 13 9 Bhutan 9 40 6 India 67 7 75 4 Pakistan 80 4 3 Nepal 2 2 Bangladesh 100 80 60 40 20 0 0 20406080100

EQ4.2. Lower pensions coverage associated EQ4.3. Poorer countries have lower pension with more adult males living with parents coverage

Percentage of male adult children living with parents Pension coverage, % of working-age population 70 35 IND MYS 60 SGP 30 KGZ PAK AZE ARM ARM KAZ 50 BGD MYS 25 AZE LKA THA MNG 40 THA 20 JPN CHN HKG 30 CHN KOR 15 VNM PHL

20 10 VNM BTN

10 AUS 5 IND NZL PAK NPL BGD 0 0 0 20 40 60 80 0 20 40 60 80 Percentage of the working-age population covered by pensions % of population living on less than USD 2 per day

Source: Pension coverage from the World Bank Database. Replacement rates come from OECD Pensions at a Glance Asia/Pacific 2009. 1 2 http://dx.doi.org/10.1787/888932546414

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 69 6. EQUITY INDICATORS

EQ5. Social spending

Public social spending measures the amount Asia/Pacific region (EQ5.3). Any sign of a minimal of resources committed by the government in the relationship is driven by three outlying richer areas of pensions, benefits (social support) and OECD countries – Australia, New Zealand and health. Much of the traditional argument for social Japan. Within the other countries, there is no rela- spending is to prevent disadvantage and thus tionship whatsoever. There is no reason to suggest enhance equity that as the countries of the Asia/Pacific region fur- The ratio of social spending to GDP varies ther develop, that there will be any automatic significantly across Asian countries (EQ5.1). Aver- development of the sort of welfare states that exist age social spending in the Asia/Pacific region is within the richer OECD countries. one-quarter of the average across the OECD as a whole. All four Asia/Pacific OECD countries fell below average OECD social spending. Social spending in Japan, New Zealand and Australia, Definition and measurement which have the highest social spending shares, is Social support to those in need is provided by each around 18% of GDP. On the other hand, nine families, community institutions and govern- Asia/Pacific countries and economies spend ments through a variety of means. In OECD less than 2% of GDP on social spending, with countries much of this support takes the form Papua New Guinea the lowest, at only 3/10th of a of social expenditure, including cash benefits percentage point. and tax advantages and the in-kind provision of The distribution of social spending across var- goods and services. To be included in social spending, benefits have to address one or more ious sorts of programmes also differs markedly by contingencies, such as low income, old age, country (EQ3.2). The biggest spending item in Japan, unemployment or disability. Programmes that China and Malaysia is social insurance, while the regulate the provision of social benefits involve biggest items in Bangladesh and Cambodia are the redistribution of resources across house- micro- and area-based spending policies. In many holds and/or compulsory participation. Social Asia/Pacific countries and economies, social expenditure is classified as public when general insurance focuses on the public and formal sec- government controls the relevant financial tors, which inevitably excludes the great majority flows. of the population, as well as most of the poor. The data presented here are extracted from Countries with higher social spending tend the Asian Development Bank’s Social Protection to be those with lower amounts of absolute pov- Index for Committed Poverty Reduction – Vol. 2: Asia erty (EQ5.2). It is likely that countries with high (2008, www.adb.org/Documents/Books/Social- Protection/Volume2), and from the OECD’s Social amounts of absolute poverty have low social Expenditure Database (SOCX, www.oecd.org/els/ spending because of cost, and low amounts of paid social/expenditure). It is worth noting that the employment and hence of marketisation of ADB divides social protection into labour mar- income available to be taxed for social purposes. ket programmes, social insurance, social assis- Additionally, low social spending may also con- tance, micro- and area-wide programmes tribute to higher poverty rates. (including microcredit) and child protection. There is no tendency for a higher level of These definitions vary significantly from the economic development to be associated with OECD’s, so data may not be fully comparable. higher spending on social protection across the

70 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 6. EQUITY INDICATORS

EQ5.1. Social spending levels and composition

Panel A. Social protection expenditure Panel B. Social expenditure by programme category, as % of GDP () 2004/05 Labour market programmes Social insurance Social assistance Child protection

19.8 OECD 18.6 Japan 18.1 New Zealand 16.5 Australia 11.0 Kyrgyzstan 9.8 Mongolia 6.4 Korea, Rep. 5.7 Sri Lanka 5.3 Azerbaijan 5.3 Bangladesh 5.2 Asia/Pacific 4.6 China 4.6 Kazakhstan 4.5 Armenia 4.1 Viet Nam 4.0 India 3.9 Malaysia 2.9 Fiji 2.3 Nepal 2.2 Philippines 1.9 Indonesia 1.6 Pakistan 1.5 Maldives 1.4 Bhutan 1.4 Cambodia 1.3 Lao PDR 1.3 Tonga 1.0 Tajikistan 0.3 Papua New Guinea 30 25 20 15 10 5 0 0 20406080100 %

EQ5.2. Social spending and poverty EQ5.3. Social spending and GDP per capita

Share of people living with less than USD 2 a day Log of GDP per capita, USD at current prices, 2009 90 5.0

80 BGD AUS JPN LAO NPL OECD IND 4.5 70 KOR NZL PAK Asia/Pacific 60 PNG 4.0 KHM MYS KAZ IDN 50 TJK VNM MDV FJI BTN PHL Asia/Pacific 3.5 CHN BTN 40 LKA PHL AZE IDN TON CHN IND LKA MNG 30 3.0 PAK ARM KGZ PNG KHM VNM KGZ 20 ARM TJK BGD LAO MNG 2.5 NPL 10 MYS KAZ AZE 0 2.0 0 5 10 15 0 5 10 15 20 25 Social protection expenditure as % of GDP Social protection expenditure as % of GDP

Source: ADB (2008), Social Protection Index for Committed Poverty Reduction, Vol. 2, Asia/Pacific Edition; OECD Social Expenditure Database (SOCX) www.oecd.org/els/social/expenditure. 1 2 http://dx.doi.org/10.1787/888932546433

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 71

Society at a Glance Asia/Pacific 2011 © OECD 2011

Chapter 7

Health Indicators

HE1. Life expectancy HE2. Infant mortality HE3. Water and sanitation HE4. Adult height HE5. Health expenditure

73 7. HEALTH INDICATORS

HE1. Life expectancy

Life expectancy at birth is a core social indi- country (Figures HE1.2 and HE1.3). The gender gap cator of the quantity of life. It is perhaps the most in life expectancy stood at 4.4 years on average general and best known measure of a population’s across Asian countries in 2008, less than the OECD health status. average of 5.6 years. Life expectancy at birth for the whole popu- Higher national income (as measured by GNI lation across 22 Asian countries and economies per capita) is generally associated with higher life reached 71.6 years on average in 2008, a gain of expectancy at birth (HE1.2). There are some more than 14 years since 1970. By comparison, notable differences in life expectancy between OECD countries gained nine years during the countries with similar income per capita. same period (Figure HE1.1). However, a large Viet Nam and Japan have higher life expectancies regional divide persists in life expectancy at birth. and Brunei and Thailand lower ones than would The country with the longest life expectancy be predicted by their GNI per capita alone. in 2008 was Japan, with a combined value for men and women of 82.6 years. Hong Kong (China), Australia, Macao (China), Singapore and New Zealand all exceeded 80 years for total life expec- Definition and measurement tancy. In contrast, a number of countries in the Life expectancy at birth is the best known Asia/Pacific region have combined life expec- measure of a population’s health status, and is tancies of less than 70 years, and in Cambodia, often used to gauge the development of a coun- Papua New Guinea and Myanmar, a child born try’s health. It measures how long, on average, a in 2008 can expect to live an average of less than newborn infant would live if the prevailing pat- 62 years. Generally, East Asian countries (China, terns of mortality at the time of birth were to Japan and the Republic of Korea) had higher stay the same throughout their lifetime. Since the factors that affect life expectancy do not life expectancies at birth than Southeast Asian change overnight, variations are best assessed countries (Cambodia, Indonesia, Malaysia, the over long periods of time. Philippines, Thailand, Viet Nam) and South Asian Age-specific mortality rates are required to countries (India, Pakistan, Bangladesh). construct life tables from which life expectan- Life expectancy at birth continues to cies are derived. Countries calculate life expec- increase remarkably in Asia/Pacific countries and tancy according to methodologies that can vary economies, reflecting sharp reductions in somewhat, and these can lead to differences of mortality at all ages. Gains in longevity can be fractions of a year. Some countries base their attributed to a number of factors, including rising life expectancies on estimates derived from living standards and better nutrition, water and censuses and surveys, and not on the accurate sanitation. Improved lifestyles, increased educa- registration of deaths. tion and greater access to quality health services also play an important role. Despite health improvements, there are still disparities in life expectancy between men and Further reading women and within countries. Women live longer WHO (2008), Health in Asia and the Pacific, World than men, and have greater rates of survival to Health Organization, Regional Office for South- age 65, regardless of the economic status of the east Asia, New Delhi.

74 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 7. HEALTH INDICATORS

HE1.1. Life expectancy levels and trends

Panel A. Life expectancy at birth, total population, 1980 and 2008 Panel B. Life expectancy at birth by sex, 2008 2008 1980, total Female Male

82.6 Japan 79.3 86.11.9 82.3 Hong Kong (China) 79.3 85.5 81.4 Australia 79.2 83.7 80.7 Singapore 78.4 83.2 80.7 Macao (China) 78.6 82.8 80.2 New Zealand 78.2 82.2 79.8 Korea, Rep. 76.5 83.3 79.2 OECD 76.4 82.0 77.4 Brunei Darussalam 75.0 79.8 74.4 Malaysia 72.1 76.8 74.4 Viet Nam 72.5 76.4 74.1 Sri Lanka 70.4 78.0 73.5 Armenia 70.4 76.9 73.1 China 71.4 74.9 71.9 Tonga 69.1 74.8 71.8 Philippines 69.7 74.1 71.8 Samoa 68.7 75.0 71.6 Maldives 70.1 73.2 70.9 Asia/Pacific 68.6 73.4 70.8 Indonesia 68.8 72.8 70.2 Azerbaijan 67.9 72.6 68.9 Fiji 66.7 71.2 68.9 Thailand 65.8 72.1 67.4 Kyrgyzstan 63.0 72.0 67.7  67.2 Korea, DPR 65.1 69.3 66.7 Tajikistan 64.1 69.5 66.7 Nepal 66.0 67.4 66.6 Mongolia 63.4 69.9 66.5 Pakistan 66.2 66.9 66.6  66.4 Kazakhstan 60.7 72.5 66.1 Bangladesh 65.1 67.2 66.1 Bhutan 64.3 68.0 65.0 Lao PDR 63.6 66.4 63.7 India 62.3 65.2 61.6 Myanmar 59.4 63.8 61.1 Timor-Leste 60.3 62.0 61.1 Papua New Guinea 59.0 63.3 61.0 Cambodia 59.2 62.8 90 80 70 60 50 40 40 50 60 70 80 90 Years Years

HE1.2. Life expectancy at birth and GNI HE1.3. Survival rate to age 65, per capita, 2008 2008

Life expectancy in years Males Females 85 JPN % of cohort NZL KOR AUS 100 80 SGP 90 BRN 80 75 VNMLKA MYS 2 CHN R = 0.79 70 PHL IDN 70 FJI 60 MNG THA PAK SLB 50 LAO IND 65 NPL 40 BGD PNG 30 60 KHM 20 10 55 0 0 10 000 20 000 30 000 40 000 50 000 r n a) re a) d ka ia a es ia d n h s ea n alia n n n s n n s n Fiji ta s d dia n r alam e s golia n n ma i GNI per capita (USD PPPs) Japa t OECD Chi n n Nepal la I n u us gapo i La s A rea, Rep.in Malay do Thailarea, DPRPaki glade Lao PDR g (Chi S russSr Viet Nam In Mo n n I MyaCambodia n Ko Philippi Ko Ba New Zeala Asia/Pacific a New G g Ko Macao (Chi ei Da u n olomo run S Ho B Pap

Source: OECD Health Data 2010; World Bank, World Development Indicators online. 1 2 http://dx.doi.org/10.1787/888932546452

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 75 7. HEALTH INDICATORS

HE2. Infant mortality

Infant mortality is a central indicator of Inequalities in infant mortality rates also infant health. Infant mortality has a significant exist within countries (HE2.3), with the richest influence on overall life expectancy, especially for population quintile gaining access to key health those countries where infant mortality is higher. interventions more quickly than the poorest. Falling infant mortality is a major influence on Reducing both types of inequity – between and rises in life expectancy, especially for those coun- within countries – is crucial for achieving lasting tries starting at a higher infant mortality level. reductions in infant mortality across the Asia/ Infant mortality reflects the effect of economic Pacific region. and social conditions on the health of mothers and new-borns, as well as the effectiveness of health systems. Around two-thirds of the deaths that occur during the first year of life in the region Definition and measurement are neonatal deaths (i.e. during the first four weeks The infant mortality rate is one of the most of life). Factors such as the health of mothers, important statistics for measuring the health of maternal care and birth weight are important a population. It is defined as the number of chil- determinants of infant mortality. Diarrhoea, pneu- dren who die before reaching their first birthday monia and malnutrition in both mothers and in a given year, expressed per 1 000 live births. Some countries base their infant mortality rates babies are the causes of many deaths. on estimates derived from censuses and sur- Countries with higher levels of economic veys, and not on the accurate registration of development generally have lower infant mortal- births and deaths. Differences among countries ity rates. In 2008, OECD countries averaged in registering premature infants may also add five infant deaths per 1 000 live births; among slightly to international variation. 19 Asian countries, the average was 30 deaths The data for OECD countries come from OECD (HE2.1). Geographically, infant mortality is lower in Health Data 2010. The data for Hong Kong (China) East Asia, and higher in South and Southeast Asia. come from the 1985 and 2008 United Nations Singapore, Japan, the Republic of Korea and Aus- Demographic Yearbook. For other countries, the tralia had under five deaths per 1 000 live births data were extracted from the UNICEF Childinfo in 2008, whereas the rates in Pakistan, Myanmar, Database (www.childinfo.org/mortality.html). Data Cambodia, Papua New Guinea and India were on infant mortality by wealth quintile are based on Demographic and Health Surveys over 50. (DHS 2006-2009) and Gwatkin et al. (2007). Infant mortality rates have fallen dramati- cally in the Asia/Pacific region over the last 30 years. Many countries, including China, India and Indonesia, saw declines of between 50 and Further reading 70% (HE2.1 and HE2.2). In Singapore, Malaysia, the Gwatkin, D.R., S. Rutstein, K. Johnson, E. Suliman, Republic of Korea, Viet Nam and Thailand, rates A. Amouzou and A. Wagstaff (2007), “Socioeco- have fallen by three-quarters. Falls in Myanmar, nomic Differences in Health, Nutrition and the Solomon Islands, Cambodia, Papua New Poverty”, HNP/Poverty Thematic Group of the Guinea and Pakistan have been less pronounced, World Bank, World Bank, Washington, DC. even though these countries had high levels of infant mortality in 1980. This has led to growing UNICEF (2008), Tracking Progress in Maternal, New- gaps between these countries and others in the born and Child Survival: The 2008 Report, UNICEF, region. New York.

76 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 7. HEALTH INDICATORS

HE2.1. Infant mortality levels and trends

Panel A. Levels, 2008 () Panel B. Changes, 1980-2008 Infant mortality rate (per 1 000 live births), both sexes % change in infant mortality rate 1.8 Hong Kong (China) -83.9 2.3 Singapore -80.0 2.6 Japan -65.3 4.1 Australia -61.7 4.1 Korea, Rep. -75.9 4.6 OECD -74.7 4.9 New Zealand -62.3 5.5 Brunei Darussalam -66.5 5.9 Malaysia -77.0 12.5 Thailand -73.0 12.7 Maldives -89.0 13.2 Sri Lanka -61.3 15.5 Fiji -53.0 16.7 Tonga -34.1 17.8 China -61.5 19.9 Viet Nam -55.3 20.7 Armenia -65.9 22.0 Samoa -60.4 26.3 Mongolia -74.1 26.4 Korea, DPR 26.5 Philippines -50.4 26.9 Kazakhstan -55.8 27.9 Asia/Pacific -59.3 30.7 Indonesia -60.7 31.9 Azerbaijan -62.1 33.3 Kyrgyzstan -62.7 40.8 Nepal -68.5 43.4 Bangladesh -68.2 47.5 Lao PDR -62.7 51.1 Timor-Leste -70.9 51.9 India -49.7 52.6 Papua New Guinea -34.8 53.9 Bhutan -56.0 54.1 Tajikistan -44.9 54.7 Myanmar -42.0 69.3 Cambodia -33.3 71.9 Pakistan -38.5 80 60 40 20 0 0 -20 -40 -60 -80 -100

HE2.2. Infant mortality rates, HE2.3. Infant mortality rate, selected countries, 1980-2008 ratios by wealth quintiles, selected countries

Death per 1 000 live births High wealth = 1.0

Australia India Indonesia Low wealth 2nd Middle 4th China OECD Rate ratio 4.0 120

3.5 100 3.0 80 2.5

60 2.0

40 1.5

20 1.0

0.5 0 h ia n s s dia s e 1980 1985 1990 1995 2000 2005 n e sta n I n Nepal glade do Paki n Cambodia In Viet Nam Ba Philippi

Note: HE2.1: 2006 data for the Republic of Korea. Change is for 1981-2006. Source: For OECD countries: OECD Health Data 2010. For non-OECD countries and economies: United Nations Demographic Yearbook 2009-2010, UNICEF 2008 report; Demographic and Health Surveys; Gwatkin et al. (2007). 1 2 http://dx.doi.org/10.1787/888932546471

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 77 7. HEALTH INDICATORS

HE3. Water and sanitation

Having access to a high quality local envi- of infant mortality (HE3.2 and HE3.3). Improving ronment, measured here in terms of the propor- sanitation and water quality are public health tion of the population with access to improved measures that can reduce mortality, particularly of water and sanitation, is important for healthy infants, who are especially vulnerable to diarrheal living. Poor air and water quality can adversely diseases and consequently a higher risk of mortal- influence both physical and psychological health. ity. Additionally, the relationship may reflect some Improved water and sanitation coverage is common third factor. The most obvious other fac- highest amongst the OECD countries of the tor is the level of economic development, which Asia/Pacific region (Panel A, HE3.1). At the other constrains the affordability of high quality water end of the scale, one-third or fewer of the popula- and sanitation, as well as restricting tried and true tions of India, Nepal and Cambodia are covered by medical and other public health innovations that improved sanitation. Generally speaking, water can lower infant mortality. coverage is better than sanitation coverage in most countries. Additionally, those countries with better sanitation coverage also tend to have better Definition and measurement improved water coverage. Data on the percentage of the population with improved water (public taps, protected wells, Water and sanitation quality have been rainwater) and sanitation (flush toilet, septic improving for most Asia/Pacific countries and tank, pit latrine, composting toilet, etc.) come economies between 1990 and 2008. In Mongolia, from the WHO JMP data set (www.wssinfo.org/ for example, the coverage of improved sanitation datamining/introduction.html). Definitions are dis- rose by 50 percentage points, while coverage of cussed in detail in WHO/UNICEF (2010). Sources improved water rose by 18 percentage points. Sim- of infant mortality data are described in HE2. ilar gains were registered in Azerbaijan, with 45 percentage points and 10 percentage points, respectively, and in Viet Nam, with 40 percentage Further reading points and 36 percentage points, respectively. WHO/UNICEF (2010), “Joint Monitoring Programme Poor rates of improved sanitation and for Water Supply and Sanitation. Progress on improved water are associated with higher rates Sanitation and Drinking Water”, 2010 Update.

78 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 7. HEALTH INDICATORS

HE3.1. Water and sanitation

Panel A. Proportion of total population served Panel B. Proportion of total population served with improved sanitation (%) () with improved water (%) 100 Australia 100 100 Japan 100 100 Korea, Rep. 98 100 Samoa 98 Maldives 91 97 Tonga 100 97 Kazakhstan 95 96 Malaysia 100 96 Thailand 98 94 Tajikistan 70 93 Kyrgyzstan 90 91 Sri Lanka 90 90 Armenia 96 81 Myanmar 71 76 Philippines 91 75 Viet Nam 94 72 Asia/Pacific 85 65 Bhutan 92 55 China 89 53 Bangladesh 80 53 Lao PDR 57 52 Indonesia 80 50 Mongolia 76 50 Timor-Leste 69 45 Pakistan 90 45 Azerbaijan 80 45 Papua New Guinea 40 31 India 88 31 Nepal 88 29 Cambodia 61 n.a. New Zealand 100 100 80 60 40 20 0 0 20 40 60 80 100

HE3.2. Poor sanitation coverage HE3.3. Poor improved water coverage and high infant mortality go together and high infant mortality go together

Infant mortality rate, per 1 000 births, 2008 Infant mortality rate, per 1 000 births, 2008 80 80 PAK PAK 70 KHM 70 KHM

60 60 PNG MMR TJK TJK MMR IND BTN PNG IND BTN 50 TLS 50 TLS LAO LAO BGD BGD NPL 40 NPL KGZ 40 Asia/Pacific Asia/Pacific KAZ AZE KGZ 30 AZE IDN 30 IDN KAZ MNG PHL WSM MNG PHL TON VNM ARM 20 VNM ARM 20 CHN MDV CHN TON LKA LKA THA 10 THA 10 MDV MYS MYS KOR NZL JPN AUS JPN AUS KOR 0 0 20 40 60 80 100 40 50 60 70 80 90 100 % of the population with improved sanitation, 2008 % of the population with improved water, 2008

Note: HE3.1: Panel A data are not available for New Zealand. Source: World Health Organisation JMP dataset (www.wssinfo.org/datamining/introduction.html). Definitions are discussed in detail in WHO/ UNICEF (2010). 1 2 http://dx.doi.org/10.1787/888932546490

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 79 7. HEALTH INDICATORS

HE4. Adult height

Adult height is determined both by genetic potential and by net nutrition during the entire period Definition and measurement of childhood. Poor net nutrition – net nutrition being Ideally, adult height data is recorded for people the balance between food intake and losses due to aged between 20 and 49 years. People reach their full physical activity and diseases – may affect adult height adult height in their early 20s while above by up to 10-15 centimeters (Steckel, 2008). For adults to age 50 people start physically shrinking (Deaton, reach their height potential, six generations of optimal 2007). Measured height is preferred over self- growth conditions are necessary (Cole, 2003). In richer reported height, as people tend to overestimate their countries, about one-fifth of within-population height own stature (Gorber et al., 2007; Lim et al., 2009). Self- variation is due to family and neighbourhood circum- reported height data is used only for Australia. stances. In poorer countries, this proportion will be When possible, data was obtained directly from larger because of greater within-population environ- specialised health surveys. For India, Bangladesh, mental variation and a shorter history of optimal Nepal, Cambodia, Kazakhstan, Armenia and growth conditions. Azerbaijan, the data were taken from Demographic Adult height varies considerably across Asia/ and Health Surveys. Weights provided in the data- Pacific countries and economies. Men and women are bases were applied. For China, data is from the the tallest in island countries of the South Pacific: China Health and Nutrition Survey, which uses a Australia, Tonga, New Zealand and Fiji. They are multistage, random cluster process to draw the shortest in the larger Asian islands of Indonesia and sample surveyed in each of the nine provinces the Philippines and on the Indian sub-continent included. No additional weights were applied. For (HE4.1 and HE4.2). In Australia, men are as tall as Indonesia, data comes from the Indonesian Family 1.78 m, whereas in Indonesia men have an average Life Survey conducted in 2007/08 by RAND, the height of 1.62 m. Tongan women are 1.66 m tall, while Center for Population and Policy Studies of the women in the Maldives are 1.50 m tall. University of Gadjah Mada and Survey METRE. The In all countries except North Korea and Timor sample is representative of about 83% of the both men and women are getting taller. However, the Indonesian population. For the Democratic People’s height difference between cohorts aged 45-49 and Republic of Korea, data is from Korean escapees who arrived in the Republic of Korea between 1999 0-24 varies greatly by country (HE4.3). In the Republic of and 2003. Consequently there may be selection- Korea, young men are almost 6 cm taller than their bias issues. For the rest of the countries, data comes father’s generation. In the Democratic People’s Republic either from national official reports or from pub- of Korea young men are the same height as their lished health papers. Details on height sources for fathers. Except in Mongolia and Malaysia, height the OECD countries are in OECD (2009). gains are much greater for men than for women. Contrary to developments in the OECD, countries with shorter populations are not catching up to those with taller populations. Gorber, S.C. et al. (2007), “A Comparison of Direct vs. Countries with more rapid economic growth Self-report Measures for Assessing Height, Weight, when young women aged 20-24 were children are and BMI: A Systematic Review”, Obesity Reviews, getting taller faster (HE4.4). Economic growth allows Vol. 8, No. 4, pp. 307-326. height growth through better nutrition and the provi- sion of better water, sanitation and health care. At the Lim, L.Y., Sam-ang Seubsman and Sam-ang Sleigh same time, the enhanced ability of taller and stronger (2009), “Validity of Self-reported Weight, Height, and people to work more productively may also stimulate Body Mass Index Among University Students in growth (Deaton, 2007). Thailand: Implications for Population Studies of Obesity in Developing Countries”, Population Health Further reading Metrics, Vol. 7, No. 15. Cole, T.J. (2003), “The Secular Trend in Human Physical OECD (2009), Society at a Glance: OECD Social Indicators, Growth: A Biological View”, Economics and Human OECD Publishing, Paris. Biology, Vol. 1, pp. 161-168. Steckel, R.H. (2008), “Heights and Human Welfare: Deaton, A. (2007), “Height, Health, and Development”, Recent Developments and New Directions”, Working Proceedings of the National Academy of Sciences, Vol. 104, Paper, No. 14536, National Bureau of Economic pp. 13232-13237. Research.

80 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 7. HEALTH INDICATORS

HE4.1. Men and women heights Panel A. Women Panel B. Men Mean height for women aged 20 to 49 (in cm ) Mean height for men aged 20 to 49 (in cm) 166 Tonga 176 164 Australia 178 164 New Zealand 177 164 OECD 177 163 Fiji 175 160 Kazakhstan 159 Azerbaijan 172 159 Armenia 158 Japan 172 158 Korea, Rep. 171 157 Mongolia 167 156 Asia/Pacific 170 155 China 167 155 Korea, DPR 165 154 Malaysia 166 152 Cambodia 152 India 164 151 Philippines 163 151 Nepal 151 Timor-Leste 151 Indonesia 162 150 Bangladesh 150 Maldives 180 170 160 150 140 140 150 160 170 180 Centimeters Centimeters HE4.2. Height difference between those HE4.3. Height difference and economic growth aged 20-24 and 45-49 years go together

Men Women () Height difference between the female cohorts aged 45-49 years and 20-24 years (%) In centimeters 3.0 6 KOR 5 2.5 4 2.0 3

2 1.5 MNG OECD CHN 1 Asia/Pacific 1.0 0 AUS MYS NPL JPN IDN -1 0.5 NZL PHL IND BGD -2 0 -3 042 6 8 10 12 S S S TL PRK AZEBGD PHL IND NZL IDN JPN NPL AU KAZ CHN KOR Average economic growth rate 1982-1992 (%) KHM ARM MY OECD MNGMDV

Asia/Pacific Note: For Tonga, the mean heights are for people aged 15 and over. For Mongolia, mean heights are for people aged 25 to 54. For Maldives and the DPR of Korea, those aged 20-24 are compared to those aged 40-49. For Malaysia, those aged 20-29 are compared to those aged 40-49. For Philippines, those aged 20-39 are compared to those aged 40-59 and for Mongolia, those aged 25-34 are compared to those aged 45-54. For Maldives, Malaysia and the DPR of Korea, the oldest cohort includes those aged 40-49. For Philippines, the oldest cohort includes those aged 40-59 and for Mongolia those aged 45-54. Data on economic growth comes from the World Bank. Source: Craig, P., S. Colagiuri, Z. Hussain and T. Palu (2007), “Identifying Cut-points in Anthropometric Indexes for Predicting Previously Undiagnosed Diabetes and Cardiovascular Risk Factors in the Tongan Population”, Obesity Research and Clinical Practice, Vol. 1, pp. 17-25; Food and Nutrition Research Institute (2003), Philippine Facts and Figures 2003, Part II “Anthropometric Facts and Figures”; Ministry of Health, Republic of Maldives (2001), “Multiple Indicator Cluster Survey (MISC 2)”; Ministry of Health and National Statistics Office, Timor-Leste, and University of Newcastle, the Australian National University, ACIL Australia Pty Ltd., Australia (2004), “Timor-Leste 2003 Demographic and Health Survey”, University of Newcastle, Australia; My, A.R., R. Junidah, A. Siti Mariam et al. (2009), “Body Mass Index (BMI) of Adults: Findings of the Malaysian Adult Nutrition Survey (MANS)”, Malaysian Journal of Nutrition, Vol. 15, No. 2, pp. 97-119; National Food and Nutrition Centre (2007), “2004 Fiji National Nutrition Survey”, Main Report; Pak, S. (2004), “The Biological Standard of Living in the Two Koreas”, Economics and Human Biology, Vol. 2, pp. 511-521; World Health Organization (2007), “Mongolian STEPS Survey on the Prevalence of Noncommunicable Disease Risk Factors 2006”; OECD Health Data. 1 2 http://dx.doi.org/10.1787/888932546509

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 81 7. HEALTH INDICATORS

HE5. Health expenditure

Private and public health spending provides Additionally, the countries that spend an important indication of the extent to which considerably on health are those with higher governments and people devote resources to GDP (HE5.3).However, in terms of GDP shares, improving health. Health spending is thus an Brunei and Singapore spend much less on health important societal response to improve health as a proportion of GDP than one would anticipate outcomes that is comparable across countries. given their per capita income, and Viet Nam and There is much variation between the Asia/ Cambodia spend an unexpectedly high proportion Pacific countries and economies in the share of on health than would be expected given their per GDP spent on health (Panel A, HE5.1). New Zealand capita income. and Australia spend around 9% of their GDP. By way of contrast, Indonesia and Myanmar spend about Definition and measurement 2%. While the general pattern is for richer countries Total health expenditure is given by the sum to spend a higher share of GDP on health, Singapore of expenditure on all the core health care func- – a country with relatively high per capita GDP – tions – that is, total health care services, medi- spends a relatively low proportion. cal goods dispensed to outpatients, prevention Health spending in the Asia/Pacific region, and public health services, and health adminis- tration and health insurance – plus capital for- unlike in the OECD, has barely increased as a mation in the health care provider industry. share of GDP in the last ten years (Panel B, Expenditure on these functions is included as HE5.1).Rises have been most pronounced, but still long as it is borne for the final use of resident well below the OECD average, in Viet Nam, the units, i.e. as long as it is final consumption by Republic of Korea and New Zealand. Falls in health nationals in the country or abroad. For this rea- spending shares were observed in a number of son, imports for final use are included and countries, most notably Brunei and Nepal. exports for final use are excluded. Countries that spend more of their income on health tend to have better health outcomes, as Further reading measured by life expectancy (HE5.2). But this rela- tionship seems to be diminishing. That is to say, OECD (2000), A System of Health Accounts, version 1.0, there are stronger gains from increasing health OECD Publishing, Paris. spending for low life expectancy countries than WHO (2010), National Health Accounts country data, for high life expectancy countries. World Health Organisation, Geneva.

82 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 7. HEALTH INDICATORS

HE5.1. Health spending, 2008

Panel A. Total real per capita health spending, Panel B. Annual average growth in real per capita USD PPP, 2008 () health spending, 1998-2008

3 448 Australia 0.06 2 751 Japan 0.05 2 465 New Zealand 0.05 1 820 Korea, Rep. 0.11 1 757 Singapore 0.05 1 194 Brunei Darussalam -0.04 733 Asia/Pacific 0.06 620 Malaysia 0.09 323 Thailand 0.07 259 China 0.11 201 Viet Nam 0.12 184 Sri Lanka 0.07 166 Fiji 0.02 138 Cambodia 0.11 136 Mongolia 0.05 135 Philippines 0.06 116 India 0.07 85 Lao PDR 0.05 82 Indonesia 0.05 71 Pakistan 0.03 70 Papua New Guinea 0.00 54 Nepal 0.01 47 Bangladesh 0.09 4000 3 000 2 000 1 000 0 -0.05 0 0.05 0.10 0.15

HE5.2. Health spending and life expectancy, HE5.3. Health spending and GDP per capita, 2008 2008

Life expectancy, 2008 Health spending per capita, USD PPP (log scale) 85 10 000 JPN SGP AUS JPN AUS 80 NZL NZL KOR KOR SGP BRN BRN VNM 1 000 75 LKA MYS CHN PHL MYS THA IDN VNM FJI SLB CHN FJI R² = 0.8124 PHL LKA 70 THA KHM NPL MNG 100 IND MNG LAO PAK PAK PNG IDN BGD NPL BGD 65 LAO IND MMR MMR PNG KHM 60 10 0 500 1 000 1 500 2 000 2 500 3 000 3 500 1 000 10 000 100 000 Total health spending, USD PPP, 2008 GDP per capita, USD PPP (log scale)

Source: OECD Health Data, Paris; WHO (2010), National Health Accounts Country Cata, World Health Organisation, Geneva; WHO NHA Database (2010). 1 2 http://dx.doi.org/10.1787/888932546528

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 83

Society at a Glance Asia/Pacific 2011 © OECD 2011

Chapter 8

Social Cohesion Indicators

CO1. Life satisfaction CO2. Trust CO3. Social behaviour CO4. Suicide CO5. Voting

85 8. SOCIAL COHESION INDICATORS

CO1. Life satisfaction

Life satisfaction represents people’s subjec- In richer countries, people tend to have a tive evaluation of their satisfaction with life as a longer satisfied life expectancy (CO1.3). The cor- whole. It is strongly associated with good family relation is quite strong between GDP per capita relationships, health and confidence in gover- and satisfied life expectancy, with only a few nance in the broader society. For low-income outliers. countries, life satisfaction at a country level is Life satisfaction is higher in richer countries related to country-level income. (CO1.3). Nevertheless, there are interesting outli- Life satisfaction in the Asia/Pacific countries ers from the regression line, such as Thailand, and economies is lower than in the OECD coun- which generates higher life expectancy than pre- tries (Panel A, CO1.1). The OECD average stands dicted by its GDP per capita. On the other hand, at 6.6, whereas the average for the Asia/Pacific Hong Kong (China) has a lower well-being score countries and economies is only 5.3. Furthermore, than expected given its per capita level of wealth. this average is pushed upward by Australia and There was no relationship between income New Zealand, which rank at the top with, respec- inequalities and life satisfaction. tively, 7.4 and 7.3. There are considerable differences across Asia/Pacific countries and economies regarding the degree to which people are satisfied with Data and measurement their lives. New Zealand, Australia and Thailand, the three countries with the highest life satisfac- Subjective measures of life satisfaction assess tion, are three average steps higher up the ladder the extent to which individuals favourably eval- uate the overall quality of their lives. Data on compared with the bottom three countries (Sri life satisfaction were gathered by the Charities Lanka, Armenia and Cambodia). Aid Foundation and come from Gallup’s World Countries where well-being scores are high View World Poll (worldview.gallup.com). Intervie- also have a longer satisfied life expectancy wees were asked to assess the way they feel (Panel B, CO1.2). Furthermore, disparities in satis- using a ten-step ladder, with zero representing fied life expectancy are higher than for life satis- the worst possible and ten the best possible. faction. In New Zealand, the satisfied life Samples are representative of a population aged expectancy is 2.4 times higher than in Cambodia, 15 and older. The measure of satisfied life which stands at the bottom of the ranking. In expectancy corresponds to the standard life Hong Kong (China), despite a relatively high life expectancy multiplied by the score of well- being standardised on a scale ranging from 0 to expectancy, life satisfaction is rather low. People 1, as in Veenhooven (1996). Data on life expec- are, however, expected to live 42 years satisfied tancy come from the World Bank. with their lives, which is four years more than the average for Asia/Pacific countries and economies. Actually, Hong Kong (China) stands as an outlier in terms of the correlation between well-being and life expectancy, which is a more objective measure Further reading of health status. In Thailand however, life satisfac- tion is relatively high compared to life expectancy Veenhooven, R. (1996), “Happy Life Expectancy”, at birth (CO1.3). Social Indicators Research, Vol. 39, No. 1, pp. 2-58.

86 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 8. SOCIAL COHESION INDICATORS

CO1.1. Life satisfaction and satisfied life expectancy

Panel A. Well-being score on a ten-step scale () Panel B. Satisfied life expectancy 7.4 New Zealand 59.3 7.3 Australia 59.4 6.9 Thailand 47.5 6.6 OECD 52.2 6.1 Singapore 49.3 5.8 Japan 47.9 5.6 Korea, Rep. 44.7 5.5 Indonesia 38.9 5.4 Kazakhstan 35.9 5.4 Malaysia 40.2 5.3 Myanmar 32.6 5.3 Viet Nam 39.4 5.3 Asia/Pacific 38.1 5.2 Pakistan 39.5 5.1 Bangladesh 33.7 5.1 Hong Kong (China) 42.0 5.1 India 32.5 5.1 Kyrgyzstan 34.4 5.0 Lao PDR 32.5 4.9 Nepal 32.7 4.9 Philippines 35.2 4.6 Azerbaijan 32.3 4.6 Tajikistan 30.7 4.5 China 32.9 4.5 Mongolia 30.0 4.2 Armenia 30.9 4.2 Sri Lanka 31.1 4.1 Cambodia 25.0 8 64 2 0 20 30 40 50 60

CO1.2. In richer countries people are more CO1.3. Richer countries have a higher satisfied satisfied life expectancy

Level of life satisfaction Satisfied life expectancy (years) 7.5 65 NZL AUS 60 7.0 THA NZL AUS 55 6.5 OECD OECD 50 JPN SGP THA SGP 6.0 45 KOR JPN HKG 5.5 IDN KOR 40 KGZ PAK IDN MYS VNM MYS KAZ Asia/Pacific MMR BGD IND VNM TJK KAZ 35 BGD PHL Asia/Pacific HKG IND 5.0 LAO KGZ PAK MMR CHN PHL LAO AZE NPL 30 ARM LKA TJK MNG AZE NPL 4.5 CHN MNG ARM 25 KHM KHM LKA 4.0 20 2.0 2.5 3.0 3.5 4.0 4.5 5.0 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Log of GDP per capita Log of GDP per capita

Note: CO1.1. Satisfied life expectancy=standard life expectancy x 0-1 happiness. CO1.2: The GDP per capita is expressed in current USD per person. Data are derived by first converting the GDP in national currency to USD and then dividing it by the total population. 2008 data except Greece, the Republic of Korea, New Zealand, Portugal, Spain, Bangladesh, Brunei, Cambodia, Indonesia, Mongolia, Samoa, Singapore, Sri Lanka, Tajikistan, Timor-Leste, and Viet Nam 2007; Switzerland, Azerbaijan, Fiji, Myanmar and Tonga 2006; Japan and the Lao People’s Democratic Republic 2005; Armenia 2004; and Papua New Guinea 2000. Source: The World Giving Index 2010, calculations made by Charities Aid Foundation based on Gallup World Poll Data. Source for life expectancy: World Bank indicators, IMF (2009), World Economic Outlook online database, www.imf.org/external/ns/cs.aspx?id=28. 1 2 http://dx.doi.org/10.1787/888932546547

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 87 8. SOCIAL COHESION INDICATORS

CO2. Trust

Trust reflects people’s perception of other pronounced in Thailand and Indonesia, but this people’s reliability. Trust may affect economic trend is reversed in Kyrgyzstan and Armenia, and social development by facilitating market where there is greater trust amongst the female exchanges, promoting the better functioning of population. public institutions and increasing the capacity for collective action (Morrone et al., 2009). Levels of trust vary considerably across the region (Panel A, CO2.1). Trust is highest in China, Definition and measurement Viet Nam and New Zealand, where just over half Trust data is based on the question: Generally the population agree that people can generally be speaking would you say that most people can be trusted. At the opposite end of the scale, fewer trusted or that you need to be very careful in than one in ten people in Malaysia and the dealing with people? Data come from the World Philippines agree that people can generally be Values Survey for various years. trusted. There is no general pattern behind recent changes in trust (Panel B, CO2.1). Large rises can be observed in Pakistan and the Philippines, and Further reading large falls in Indonesia and India. Morrone, A., N. Tontoranelli and G. Ranuzzi (2009), In most countries across the region there is a “How Good is Trust? Measuring Trust and its persistent tendency for men to be more trusting Role for the Progress of Societies”, OECD Statis- than women (CO2.2). The difference is most tics Working Paper, OECD Publishing, Paris.

88 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 8. SOCIAL COHESION INDICATORS

CO2.1. Trust levels and trends

Panel A. Share of people expressing trust in others, Panel B. Annual average growth rate, mid-1990s latest year available () to latest year available, %

52 China 0.0

52 Viet Nam 4.9

51 New Zealand 0.7

46 Australia 1.5

43 Indonesia -3.8

42 Thailand

41 Hong Kong (China)

39 Japan -1.6

32 Asia/Pacific 1.6

31 Pakistan 10.6

28 Korea, Rep. -0.8

25 Armenia

24 Bangladesh 2.0

23 India -4.3

21 Azerbaijan

17 Singapore

17 Kyrgyzstan

9 Malaysia

8 Philippines 8.8

60 40 20 0 -5 0 5 10 15 %

CO2.2. Men tend to trust more than women Percentage points difference in the share of men expressing trust in others and the share of women %

7.7 7 6.5

5.0 5 4.4 3.5 3.6 3 2.2 1.9 1.5 1.7 1 0.7 0.9 0.0 0.0 0.2 0.2 -0.1 -1 -2.8 -1.5

-3 n ia d n n s h ia a n e d ia n n e dia s s a) n r n s sta n n n ralia sta e I t Chi n rme Japa rbaija gapo gyz A ea, Rep. glade us n do r Viet Nam r n Malay g (Chi A Paki i Thaila In Ky Aze Philippi n S New Zeala Ko Ba Asia/Pacific g Ko n Ho

Source: World Values Survey Databank, Online data analysis: www.wvsevsdb.com/wvs/WVSAnalizeStudy.jsp. 1 2 http://dx.doi.org/10.1787/888932546566

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 89 8. SOCIAL COHESION INDICATORS

CO3. Social behaviour

Pro-social behaviour is behaviour that than average pro-social behaviour, but a high per- contributes to the positive functioning of society formance on the low corruption index. and is the antithesis of anti-social behaviour. Richer countries do not do better on the giv- Pro-social behaviour measures an important posi- ing index (CO3.2). Contrary to what might be tive dimension of societal cohesion. The indicator expected, richer countries are not systematically here picks up three pro-social actions: volunteer- more generous with time, money or help for a ing time, giving money and helping a stranger. stranger. Japan and Australia are both rich coun- Anti-social behaviour is picked up by an indicator tries, but the former is low on the giving index and of perceived corruption. the latter high. Bangladesh and the Lao People’s Australia and New Zealand have the highest Democratic Republic are both poorer countries, amounts of pro-social behaviour in the Asia/ but the former is low on the giving index and the Pacific region and indeed in the world (Panel A, latter high. CO3.1). Pro-social behaviour is also notably high However, richer countries do have a better in Sri Lanka and the Lao People’s Democratic corruption perception index (CO3.3). The relation- Republic. Pro-social behaviour is notably low in ship between GDP per capita and corruption is China, Bangladesh and Pakistan, amongst the very strong. It may be that low corruption is a lux- lowest in the world, not simply in the region. ury good that only richer countries can afford, or it Australia and New Zealand also have the may be that lowering corruption may be one of the lowest amounts of anti-social behaviour in the conditions necessary to reach a certain level of Asia/Pacific region and indeed the world, as mea- GDP per capita. sured by perceived corruption (Panel B, CO3.1). Perceived corruption is notably high (low index) in Cambodia, Kyrgyzstan and Myanmar. Anti-social behaviour, in the form of corrup- Definition and measurement tion, is also high in those three countries. Data on pro-social behaviour come from the Hong Kong (China) is quite high on pro-social World Giving Index, and are collected by Gallup. behaviour suggesting that it is not simply culture Corruption data come from the Transparency that is driving low pro-social behaviour in China. International Corruption Perceptions Index. Singapore is another interesting case, with lower

90 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 8. SOCIAL COHESION INDICATORS

CO3.1. Pro-social behaviour and low corruption perception index

Panel A. Giving time, money and help: pro-social behaviour Panel B. Low corruption perception World Giving Index score (scale of 0 to 1) Low corruption perception index, 2009 (scale of 0 to 10) 0.6 Australia 8.7 0.6 New Zealand 9.4 0.5 Sri Lanka 3.1 0.5 Lao PDR 2.0 0.4 Hong Kong (China) 8.2 0.4 Myanmar 1.4 0.4 Thailand 3.4 0.4 OECD 7.0 0.4 Indonesia 2.8 0.4 Philippines 2.4 0.4 Tajikistan 2.0 0.3 Azerbaijan 2.3 0.3 Mongolia 2.7 0.3 Asia/Pacific 3.9 0.3 Malaysia 4.5 0.3 Korea, Rep. 5.5 0.3 Singapore 9.2 0.3 Nepal 2.3 0.2 Armenia 2.7 0.2 Japan 7.7 0.2 Kazakhstan 2.7 0.2 Kyrgyzstan 1.9 0.2 India 3.4 0.2 Viet Nam 2.7 0.2 Cambodia 2.0 0.2 Pakistan 2.4 0.2 Bangladesh 2.4 0.1 China 3.6 0.6 0.5 0.4 0.3 0.2 0.1 030 162 495 7 8 10

CO3.2. There is no correlation between giving CO3.3. Richer countries have a better and country wealth corruption perception index

GDP per capita, 2005 (log), 2009 current USD GDP per capita, 2005 (log), 2009 current USD 5.0 5.0

JPN AUS 4.5 JPN AUS SGP SGP HKG NZL 4.5 KOR KOR HKG NZL 4.0 4.0 MYS KAZ MYS 3.5 Asia/Pacific THA KAZ CHN ARM AZE 3.5 IDN LKA ARM THA 3.0 IND MNG PHL AZE IDN Asia/Pacific PAK VNM PHL CHN KGZ PAK LKA BGD LAO 3.0 2.5 KHM TJK VNM IND NPL KHM MMR KGZ MNG LAO BGD 2.0 2.5 TJK NPL MMR

1.5 2.0 0 0.1 0.2 0.3 0.4 0.5 0.6 0612345 78910 World Giving Index Low corruption perception index, 2009

Source: Data on pro-social behaviour come from the World Giving Index, and are collected by Gallup. Corruption data come from the Transparency International Corruption Perceptions Index. 1 2 http://dx.doi.org/10.1787/888932546585

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 91 8. SOCIAL COHESION INDICATORS

CO4. Suicide

Intentionally killing oneself is evidence not only of personal breakdown, but also of a deteri- Definition and measurement oration of the social context in which a person Suicide rates are based on the crude number lives. Suicide is most likely to occur during crisis of deaths caused by suicide divided by periods associated with divorce, unemployment, the total population. Rates are presented per clinical depression or other forms of mental ill- 100 000 persons. The comparability of suicide data between countries is affected by a number ness. of reporting criteria. Caution is thus required in The average suicide rate for the Asia/Pacific interpreting variations across countries. Suicide region is close to 12 per 100 000 (CO4.1). However, rates disaggregated by gender and trends in sui- disparities in suicide rates are huge. In Kazakhstan, cide rates for Asia/Pacific OECD countries and for Sri Lanka and the Lao People’s Democratic the OECD average come from the WHO Mortality Database. These rates have been age-stan- Republic, suicide rates are more than six times dardised to the 1980 OECD population, to remove higher than the rates of Brunei, Azerbaijan, and variations arising from differences in age struc- Philippines. tures across countries and over time within each Average suicide rates have increased across country. Data for Macao (China) come from the Asia/Pacific countries and economies (CO4.1). In WHO Mortality Database, but the data has not been age-standardised. Data for Hong Kong Spe- the Republic of Korea, suicide rates almost tripled. cial Administrative Region comes from a country The economic downturn, weakening social inte- report to the WHO (www.who.int/mental_health/ gration and the erosion of the traditional family prevention/suicide/country_reports/en/ind ex.html). support base for the elderly have all been impli- Data for the other Asia/Pacific countries cated in Korea’s recent increase in suicide rates. and economies are from the WHO Global Burden of Disease Database that draws on a wide range In general worldwide, suicide rates are of data sources (www.who.int/healthinfo/ greater for men than for women (CO4.2). However, global_burden_disease/estimates_country/en/ in China, in 2004, there were almost 100 000 more index.html). The latest assessment of GBD is suicides committed by women than by men. China for 2004. Age-standardised death rates were accounts for 40% of female suicides worldwide. In used. However, information on the population countries with larger rural populations, such as year used for the standardisation is missing. Data on trends for the non-OECD Asia/Pacific India, China, Pakistan, Sri Lanka and Nepal, the countries and economies come from country most frequent method of suicide is poisoning by reports to the WHO. For China, data on trends pesticide (Gunnell et al., 2007). Ready access to and on age disaggregation come respectively lethal pesticides results in an increased number of from Yip et al. (2005) and from a country report to fatalities from low-intent suicide behavior. the WHO. For the age disaggregation, data for all countries except China come from the WHO In most countries, except Kazakhstan, New Mortality Database. Zealand and Australia, suicide rates are higher among older people. Older people seem to be at particular risk in Chinese populations (China, Further reading Hong Kong [China], Macao [China], Singapore) and Gunnell, D., M. Eddleston, M.R. Phillips and the Republic of Korea (CO4.3). In the Republic of F. Konradsen (2007), “The Global Distribution of Korea, suicide rates for those over age 65 are more Fatal Pesticide Self-poisoning: Systematic than five times higher than suicide rates of Review”, BMC Public Health, Vol. 7, No. 357. 15-34 year-olds. Yip, P., K.Y. Liu, Jianping Hu and X.M. Song (2005), “Suicide Rates in China During a Decade of Rapid Changes”, Social Psychiatry and Psychiatric Epidemiology, Vol. 40, pp. 792-798.

92 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 8. SOCIAL COHESION INDICATORS

CO4.1. Suicide rates and variation in suicide rates

Panel A. Trends Panel B. Changes in suicide rates Levels (deaths per 100 000 population) () Trends (1990 to mid-2000s) 28.2 Kazakhstan 7.8 28.1 Sri Lanka 26.226.2 Lao PDR 22.4 Maldives 21.5 Korea, Rep. 13.6 19.4 Japan 4.9 18.6 India 1.6 16.9 China 15.5 Bhutan 15.2 Hong Kong (China) 3.7 14.5 Bangladesh 13.1 Kyrgyzstan -3.5 13.1 Mongolia 12.9 Pakistan 12.3 New Zealand -1.1 12.0 Papua New Guinea 11.7 Asia/Pacific 1.5 11.7 Nepal 11.7 Viet Nam 11.4 Indonesia 11.1 Singapore -2.8 10.9 OECD -3.0 10.1 Myanmar 9.8 Thailand 3.8 9.6 Timor-Leste 8.2 Malaysia 7.5 Australia -5.0 6.9 Macao (China) 5.5 Samoa 5.0 Cambodia 4.7 Korea, DPR 4.5 Tonga 3.9 Fiji 3.7 Armenia -0.4 2.6 Tajikistan -1.8 1.9 Philippines 1.4 Azerbaijan -1.0 1.4 Brunei Darussalam 30 25 20 15 10 5 0 -5 0 5 10 15

CO4.2. Men commit suicide more CO4.3. Suicide rates are highest among than women older people Difference in suicide rates between men and women in Suicide rates (deaths per 100 000 population) by age category selected countries (per 100 000 population) in selected countries and economies

15-34 35-64 65+ 80 20 19 18 72 70 15 60 51 50 10 8 9 7 40 31 33 5 30 29 27 22 20 18 18 18 19 17 17 0 14 13 10 10 11 12 11 12 11 0 10 10 9 -2 -5 0 a n n d d n a) a n dia n n n n sta n ralia Chi I t Japa OECD Japa Chi us ea, Rep. Paki A rea, Rep. Thaila g (Chi r n Ko Asia/Pacific Ko New Zeala Asia/Pacific g Ko n Ho

Source: OECD Health Data 2010; for non-OECD: WHO, country reports: www.who.int/mental_health/prevention/suicide/country_reports/en/ index.html; Macao (China): data from WHO Statistical Information System (WHOSIS): http://apps.who.int/whosis/database/mort/table1.cfm. 1 2 http://dx.doi.org/10.1787/888932546604

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 93 8. SOCIAL COHESION INDICATORS

CO5. Voting

A high voter turnout is a sign that a country’s is measured in terms of the Giving Index (CO3), political system enjoys a strong degree of partic- which incorporates information on helping a ipation. While low voter turnout might reflect sat- stranger and giving time or money. The observed isfaction with the country’s management, it also relationship is, however, relatively moderate. implies that the political system reflects the will of Countries with high voter turnout show a wide a limited number of citizens. A number of coun- range of other pro-social behaviour. For example, tries in the region do not choose to hold demo- pro-social behaviour is very high in the Lao cratic elections and thus cannot be included for People’s Democratic Republic, as is voter turnout. consideration. High Laotian voter turnout is shared with neigh- Voters turnout rates vary hugely across the bouring Viet Nam, where pro-social behaviour is region (Panel A, CO5.1). Over nine in every ten notably low. people turn out to vote in the Lao People’s Demo- cratic Republic, Viet Nam, Australia (where voting is a legal obligation) and Singapore, compared to fewer than one in every two people in the Republic of Definition and measurement Korea and Pakistan, the two lowest turnouts in the Voting in national parliamentary elections is region. In all other countries for which there is data one indicator of people’s participation in their on voting turnout in parliamentary elections, more community’s national life. The indicator used than half the eligible population vote. here to measure the participation of individuals in the electoral process is the “voter turnout”, Voter turnout has generally declined in most i.e. the number of individuals who cast a ballot countries in the last generation (Panel B, CO5.1). during an election as a share of the population The pattern of voting decline has been uneven and of voting age – generally the population aged 18 far from universal. Nepal, the Republic of Korea or more – as available from administrative and Mongolia have had declines on the order of records of member countries. Different types of one-fifth to one-quarter of the voting population. elections occur in different countries according However, over one-third of countries have actually to their institutional structure and different experienced rises in voter turnout, in some cases geographical jurisdictions. For some countries, considerable ones. In Bangladesh, the rise has it should be noted, turnout for presidential elec- been by nearly one-third of the voting population. tions and regional elections may be higher than for national parliamentary elections, perhaps Large rises in turnout over the last generation because those elected through these ballots are were also a feature of the electorates in Thailand, constitutionally more important for how those Kyrgyzstan and Sri Lanka. countries are run. Data about voter turnout are Voter turnout is positively associated with extracted from the international database pro-social behaviour (CO5.2). One could look upon organised by the Institute for Democratic and Elec- voting – political participation – as an additional toral Assistance (IDEA). sort of pro-social behaviour. Pro-social behaviour

94 SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 8. SOCIAL COHESION INDICATORS

CO5.1. Voting

Panel B. Changes in voter turnout Panel A. Most recent parliamentary election Percentage points change in voter turnout, Voter turnout, latest year available (%) () 1990 to latest year available 100 Lao PDR 0.5 99 Viet Nam 95 Australia -0.1 94 Singapore -1.0 93 Tajikistan 9.2 85 Bangladesh 29.8 84 Indonesia -6.8 83 Samoa 4.5 79 New Zealand -5.8 79 Bhutan 79 Thailand 19.2 76 Malaysia 6.0 76 Sri Lanka 12.4 75 Cambodia -11.6 75 Kyrgyzstan 14.1 75 Asia/Pacific -1.8 74 Mongolia -23.7 71 Maldives 2.6 70 OECD -7.2 68 Kazakhstan -7.7 67 Japan -5.9 66 Papua New Guinea -15.5 64 Fiji -14.8 64 Philippines -6.9 63 Nepal -24.8 60 Armenia 6.1 58 India 1.0 46 Korea, Rep. -25.9 45 Pakistan -0.9 100 80 60 40 20 0 -30 -20 -10 0 10 20 30

CO5.2. Voting is positively associated with other pro-social behaviour

Index of pro-social behaviour 0.6 NZL AUS LKA 0.5 LAO

THA 0.4 PHL IDN TJK MNG

0.3 KOR MYS NPL SGP ARM JPN 0.2 KAZ KGZ IND VNM PAK KHM BGD

0.1 405060708090100 Voter turnout (%)

Source: International database organised by the Institute for Democratic and Electoral Assistance (IDEA) www.idea.int/. Data on pro-social behaviour come from the World Giving Index, and are collected by Gallup. 1 2 http://dx.doi.org/10.1787/888932546623

SOCIETY AT A GLANCE: ASIA/PACIFIC 2011 © OECD 2011 95 ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT

The OECD is a unique forum where governments work together to address the economic, social and environmental challenges of globalisation. The OECD is also at the forefront of efforts to understand and to help governments respond to new developments and concerns, such as corporate governance, the information economy and the challenges of an ageing population. The Organisation provides a setting where governments can compare policy experiences, seek answers to common problems, identify good practice and work to co-ordinate domestic and international policies. The OECD member countries are: Australia, Austria, Belgium, Canada, Chile, the Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Israel, Italy, Japan, Korea, Luxembourg, Mexico, the Netherlands, New Zealand, Norway, Poland, Portugal, the Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Turkey, the United Kingdom and the United States. The European Union takes part in the work of the OECD. OECD Publishing disseminates widely the results of the Organisation’s statistics gathering and research on economic, social and environmental issues, as well as the conventions, guidelines and standards agreed by its members.

OECD PUBLISHING, 2, rue André-Pascal, 75775 PARIS CEDEX 16 (81 2011 08 1 P) ISBN 978-92-64-10569-0 – No. 59717 2011 Society at a Glance Asia/Pacific 2011

Contents Chapter 1. How do People in the Asia/Pacific Region Spend their Time? Chapter 2. Interpreting OECD Social Indicators Society at a Glance Chapter 3. Society at a Glance: An Overview Chapter 4. General Context Indicators Asia/Pacific 2011 Chapter 5. Self-sufficiency Indicators Chapter 6. Equity Indicators Chapter 7. Health Indicators Chapter 8. Social Cohesion Indicators www.oecd.org/els/social/indicators/asia

Society at a Society Glance Asia/Pacific 2011 Asia/Pacific

Please cite this publication as: OECD (2011), Society at a Glance: Asia/Pacific 2011, OECD Publishing. http://dx.doi.org/10.1787/9789264106154-en This work is published on the OECD iLibrary, which gathers all OECD books, periodicals and statistical databases. Visit www.oecd-ilibrary.org, and do not hesitate to contact us for more information.

ISBN 978-92-64-10569-0 81 2011 08 1 P www.oecd.org/publishing -:HSTCQE=VUZ[^U: