101666 A N U N FA IR STA RT h oW U n e qUA l o PP o rTU n iTi e s A F F e c T i n do n e s i A’ s c h i ld r e n 2 NOV EM BER 2015 The World Bank The World Bank Printed in Office Jakarta 1818 H Street NW November 2015 Indonesia Stock Exchange Washington, DC 20433, Building Tower II/12th Floor USA Jl. Jend. Sudirman Kav. T (202) 458-1876 52-53 F (202) 522-1557/1560 Jakarta 12910 W www.worldbank.org P (6221) 5299-3000 F (6221) 5299-3111 W www.worldbank.org/id An Unfair Start: How the Government they concerning the legal status unequal opportunities represent. The World of any territory or the affect Indonesia’s children Bank does not guarantee endorsement or acceptance is a product of the staff the accuracy of the data of such boundaries. of the World Bank. The included in this work. For any questions regarding findings, interpretations, The boundaries, colors, this report, please contact and conclusions expressed denominations, and other herein do not necessarily information shown on any Vivi Alatas reflect the views of the map in this work do not (valatas@worldbank.org) Board of Executive Directors imply any judgment on and Matthew Wai-Poi of the World Bank or the part of the World Bank (mwaipoi@worldbank.org). AN UN FAIR START 3 NOV EM BER 2015 Acknowledgements An Unfair Start was prepared by the World Bank’s Poverty This paper was prepared by Matthew Wai-Poi (Senior Global Practice team in the Jakarta o ce. The team, led by Economist, GPVDR), Grace Hadiwidjaja (Consultant, Vivi Alatas (Lead Economist, GPVDR), provides technical GPVDR), Laura Wijaya (Research Analyst, GPVDR) and and policy advice based on sound empirical research and Taufik Indrakesuma (Consultant, GPVDR), and produced analysis to the Government of Indonesia in support of its under the overall guidance of Ana Revenga (Senior e orts to reduce poverty, vulnerability and inequality. Director, GPVDR), Carlos Silva-Jaurequi (Lead Economist, GPVDR) and Salman Zaidi (Practice Manager, GPVDR). Financial support for this background paper was provided Strategic guidance and comments were provided by by the Australian Department of Foreign A airs and Trade Cristobal Ridao-Cano (Program Leader, EACIF). This (DFAT) through the trust fund for the Partnership for work builds on the 2013 working paper Multidimensional Knowledge-based Poverty Reduction. The trust fund is Child Poverty in Indonesia by Grace Hadiwidjaja, Cindy under the strategic oversight of Bambang Widianto, Executive Paladines and Matthew Wai-Poi. Secretary of the National Team for the Acceleration of Poverty Reduction (Tim Nasional Percepatan Penanggulangan Kemiskinan, or TNP2K) and Rahma Iryanti of the National The paper was edited by Peter Milne. Layout for this paper Development Planning Agency ( Badan Perencanaan was done by Bentuk Team (Andreas Pranoto, Muhammad Pembangunan Nasional, or Bappenas). Kamal, Phoebe Wathoel, Randy Kurnia). AN UNFAIR START 4 Table of Contents 03 ACKNOWLEDGEMENTS 04 TA B L E O F C O N T E N T S 05 L I S T O F A C R O N Y M S , A B B R E V I AT I O N S , A N D I N D O N E S I A N T E R M S 06 EXECUTIVE SUMMARY 08 I . W H Y D O E S I N E Q U A L I T Y O F O P P O R T U N I T Y M AT T E R ? 15 I I . W H AT D O E S T H E L I T E R AT U R E T E A C H U S A B O U T I N E Q U A L I T Y O F O P P O R T U N I T Y ? 17 I I I . W H I C H I N E Q U A L I T I E S O F O P P O R T U N I T Y M AT T E R M O S T I N I N D O N E S I A? 28 I V. W H AT D O E S I N D O N E S I A N E E D T O D O ? 32 REFERENCES 35 A N N E X : D ATA A N D M E T H O D O L O G Y AN UNFAIR START 5 NOV EM BER 2015 List Of Acronyms, Abbreviations, And Indonesian Terms TERM DEFINITION TERM DEFINITION CONEVAL Consejo nacional de evaluacion de la politica MHH Male-headed household de desarrollo social (national council for the evaluation of social development policy) MPI Multidimensional poverty index desa Village PKH Program Keluarga Harapan (Family Hope Program) DKI Daerah Khusus Ibukota (special capital region) PLN Perusahaan Listrik Negara (State Electricity ECED Early childhood education Company) FHH Female-headed household Podes Sensus potensi desa (village potential census) GDP Gross Domestic Product Q1 Poorest 20% of the population GFC Global Financial Crisis Q2 Second-poorest 20% of the population HDI Human Development Index Q3 Middle 20% of the population HOH Head Of Household Q4 Second-richest 20% of the population HOI Human Opportunity Index Q5 Richest 20% of the population IFLS Indonesian Family Life Survey SD Sekolah Dasar (primary school) Jamkesmas Jaminan kesehatan masyarakat (social SMA Sekolah Menengah Atas (senior secondary health insurance) school) kabupaten Regency SMP Sekolah Menengah Pertama ( junior secondary school) kelurahan Urban ward Susenas Survey Sosial Ekonomi Nasional KIP Kartu Indonesia Pintar (Indonesia Smart Card) (National Socioeconomic Survey) KM Kilometer UNDP United Nations Development Programme Kotamadya Township UNICEF United Nations Children's Fund AN UNFAIR START 6 NOV EM BER 2015 Executive Summary De s pi t e r a pi d e c on om ic gr ow t h , i n e qua l i t y i s i nc r e a s i ng i n I n d on e s i a . After recovering from the Asian financial crisis in 1997/98, Indonesia’s real GDP per capita grew at an annual average of 5.4 percent between 2000 and 2014. This robust rate of growth helped to halve the poverty rate from 23.4 percent during the crisis down to 11.2 percent by 2015. However, between 2003 and 2010, consumption per person for the richest 10 percent of Indonesians grew at over 6 percent per year after adjusting for inflation, while for the poorest 40 percent it grew by less than 2 percent per year. This disparity in consumption between different income levels has, in turn, given rise to a sharp increase in the Gini coefficient over the past 15 years, increasing from 30 in 2000 to 41 in 2013. I n e qua l i t y of opp ort u n i t y at bi rt h a n d i n c h i l dho od e x pl a i n s a s u b s ta n t i a l a mou n t of c on s u m p t ion i n e qua l i t y l at e r i n l i f e . Inequality of opportunity, when not everyone has access to the same opportunities in life, accounts for one-third of all consumption inequality. Access to opportunities—such as adequate health facilities and good quality schools— impact whether children are able to cognitively and physically develop, which in turn affects whether they can acquire the skills needed to find well-paying jobs. Thus, inequality experienced at birth perpetuates inequality in later life. AN UN FAIR START 7 E X ECU TI V E SU M M A RY NOV EM BER 2015 T h e mo s t a ppa r e n t c on t r i bu t or t o schools without electricity or running water. Another i n e qua l i t y of opp ort u n i t y i s ge o gr a ph y: example is that rural households are mostly (95 percent) e a s t e r n I n d on e s i a l ag s be h i n d t h e r e s t able to access primary health care facilities, such as of t h e c ou n t ry i n a l mo s t a l l h e a lt h a n d polyclinics or Puskesmas, but only 70 percent are able e duc at ion i n dic at or s . People living in eastern to access hospitals. Indonesia, particularly in Maluku and Papua, have much more difficulty accessing primary health care, hospitals, T h e e duc at ion l e v e l a n d i nc om e of and skilled delivery services, and even those who can t h e hous e hol d h e a d m at t e r s om e w h at access public health centers (Puskesmas) are unlikely to i n de t e r m i n i ng a c h i l d ’ s ac c e s s t o find ones with proper facilities. Schools in Maluku and opp ort u n i t i e s , bu t t h e ge n de r of t h e Papua are further away from homes, less likely to have hous e hol d h e a d d oe s n o t. As parents become access to electricity or proper sanitation facilities, and richer and better educated, their children are more likely less likely to have high quality teachers. Papua is the only to have access to higher quality health and education region without universal access to primary school (SD), services. There are also improvements in intermediary outcomes such as stunting and educational test scores. with 20 percent of households having to travel over 1km However, there are no significant differences in access to the nearest village with a primary school. Thus, children or outcomes between male- and female-headed who are born and grow up in eastern Indonesia start life households. with a huge disadvantage compared with those in the rest of the country. A m u lt i-pr onge d a ppr oac h i s n e c e s s a ry t o tac k l e i n e qua l i t y of opp ort u n i t y. These How e v e r , e v e n w i t h i n r e gion s , di f f e r e nc e s findings show that there is much to be gained from i n ac c e s s t o opp ort u n i t i e s be t w e e n u r b a n improving the quality of health and education facilities a n d ru r a l a r e a s a r e a ppa r e n t. While not as in eastern and rural Indonesia, as well as building jarring as the differences between regions, urban-rural roads to get to them. In addition, strengthening social differences in access to opportunities also exist within protection programs such as conditional cash transfers regions, especially in terms of the quality of available and education subsidies would help to improve access to facilities. For example, although both urban and rural opportunities for the poor and vulnerable. households have close to universal access to primary schools, rural children are more likely to find their AN UN FAIR START 8 NOV EM BER 2015 1. WHY DOES INEQUALITY OF OPPORTUNITY MATTER? I N E QUA L I T Y I N I N D ON E S I A I S R I S I NG R A PI DLY. were hit the hardest. However, since then the Gini has By most measures, inequality in Indonesia has reached increased from 30 points in 2000 to 41 points in 2014, historically high levels. In 2002, the richest 10 percent its highest recorded level in Indonesia (Figure 1). Even of Indonesians consumed as much as the poorest 42 this record level is likely to be underestimated, however, percent combined; by 2014, they consumed as much because household surveys tend to miss out the richest the poorest 54 percent. A popular measure of inequality households. Once relatively moderate by international is the Gini coe cient, where 0 equates to complete standards, Indonesia’s level of inequality is now equality and 100 represents complete inequality. During becoming high and increasing at a faster pace than most the Asian financial crisis in 1997/98, poverty rose sharply of its East Asian neighbors (Figure 2).1 while the Gini fell; everyone was a ected, but the richest After a long period of stability, the Gini began rising, SOURCE BPS, Susenas and World Bank calculations NOTE Nominal consumption Gini. The national poverty then fell with the Asian financial crisis, before rising line was changed in 1998, and the 1996 rate calculated sharply since the recovery. under both the new and old methodologies. Gini coe cient (points) and national poverty rate (percent) 1980–2014 (fig.1) S UHA RTO ER A ASIA N D E M O C R ACY, G LOB A L F IN A NCI A L F IN A N C IA L D E C E N TR A LIZ ATIO N A N D C R ISIS & A FT E R MAT H C R ISIS C O M M O D ITY B O O M 45 40 35 30 GINI 25 20 15 10 POV ERTY OLD POVERTY NEW 5 0 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 1 The report Indonesia’s Rising Divide: Why inequality is rising, why it matters and what can be done (World Bank 2015a) provides a detailed diagnosis on the trends and drivers of inequality in Indonesia. AN UNFAIR START 9 WHY DOES INEQUALITY OF NOV EM BER 2015 OPPORTUNITY MATTER? The increase in the Gini in Indonesia over the past two SOURCE Zhuang, et al. 2014. NOTE Consumption Ginis for all countries except Malaysia, which uses decades is one of the highest in the region. income. The periods for each country are: Indonesia 1990-2011; Malaysia Gini coe cient in East Asia 1990s & 2000s (fig.2) 1992-2009; Lao PDR 1992-2008; China 1990-2008; Vietnam 1992-2008; Thailand 1990-2009; the Philippines 1991-2009; and Cambodia 1994-2008. 90'S 00' S Malaysia China Philippines Thailand Indonesia Cambodia India Laos Vietnam 0 10 20 30 40 50 60 I NC OM E I N E QUA L I T Y I S N O T A LWAY S A B A D restrictions that increase the cost of living most for the T H I NG ; I T C A N PR OV I DE R E WA R D S F OR T HO S E poor, or patterns of government taxes and spending that W HO WOR K H A R D A N D TA K E R I S K S . Hard work and fail to collect and channel su cient resources to help the innovation benefit society by creating new goods and poor and vulnerable, or those without equal access to services that everyone can enjoy, as well as contributing services. to a larger economy. This, in turn, can present the Government with a greater ability to provide public S OM E C ON S U M P T ION I N E QUA L I T Y I S DU E T O services to all. If this results in a gap between those who F OU R C I R C U M S TA NC E S OF BI RT H OR FAC T OR S work hard, take risks and innovate, and those who work T H AT A R E BE YON D T H E C ON T R OL OF T H E less hard, then some income inequality may be justified I N DI V I DUA L , OF W H IC H T H E MO S T I M P ORTA N T and even desirable. Indeed, many Indonesians share this I S PA R E N T S ’ E DUC AT ION . Four circumstances are view. When asked in a 2014 survey whether inequality examined: (i) province of birth, (ii) whether the individual is ever acceptable, 74 percent of respondents say that was born in urban/rural areas, (iii) the gender of the head “inequality is sometimes acceptable” so long as wealth of the household, and (iv) the parents’ level of education. acquisition is fair and meritocratic. 2 Figure 3 shows how much of current consumption inequality is due to di erences between groups for each HOW E V E R , I N E QUA L I T Y I S U N FA I R W H E N I T of these circumstances. For example, the di erence in I S DU E T O FAC T OR S T H AT A R E BE YON D T H E average per capita consumption between those born C ON T R OL OF I N DI V I DUA L S . There are many forms of in an urban district (kotamadya) and those in a mixed inequality: economic inequalities of income, wealth and urban-rural district (kabupaten) account for 9 percent of consumption, and inequalities of opportunity, when not di erences in all households’ per capita consumption everyone has access to the same opportunities in life. (between-group di erence). The other 91 percent of Factors beyond the control of an individual—where you inequality is due to di erences within each group; that are born, how educated or wealthy your parents are, and is, di erences in consumption within households living what access to public services you had when you were in kotamadya and di erences in consumption within growing up—can have a major influence on how your households living in kabupaten (within-group di erence). life turns out. Getting a healthy start in life and a quality Similarly, di erences between provincial averages education are fundamental prerequisites for getting account for 7 percent of consumption inequality, while a good job and earning a decent living in adulthood. 93 percent of inequality is due to di erences within each When economic inequality arises because of inequality province. The most important between-group di erence of opportunity, when not everyone has a fair start in life, is that for parental education, which explains 26 percent it is unfair. Other factors outside an individual’s control of inequality, while the least important between-group that can a ect incomes, standards of living and inequality di erence is for gender of the head of household, which include government policies, such as food import explains almost no inequality. 2 For more exploration of public perceptions of and attitudes to inequality, see A Perceived Divide: How Indonesians perceive inequality and what they want done about it (World Bank 2015b). AN UNFAIR START 10 WHY DOES INEQUALITY OF NOV EM BER 2015 OPPORTUNITY MATTER? ON E -T H I R D OF A L L C ON S U M P T ION I N E QUA L I T Y I S DU E T O J US T T H E S E F OU R C I R C U M S TA NC E S OF BI RT H C ON S I DE R E D A LT O GE T H E R . That between-group di erences of these four circumstances when taken together can explain at least one-third of consumption inequality. 3 Two-thirds of consumption inequality is due to within-group di erences, which may include other birth and early childhood circumstances, as well as aspects related to an individual’s characteristics, such as hard work, or random things such as luck. Role of individual birth circumstances in explaining Role of combined birth circumstances in consumption inequality (fig.3) explaining consumption inequality (fig.4) 100 92% 91% 100% 73% PE RC E NT OF I N E QUALI T Y 80 E XPLAI N E D 60 33% 67 % 40 20 26% 8% 9% 0 PROVI N CE URBAN HO H PARENTS’ /RURAL GE N DE R EDUCATI ON WITHIN GROUP DIFFERENCES WITHIN GROUP DIFFERENCES BETWEEN GROUP DIFFERENCES BETWEEN GROUP DIFFERENCES SOURCE Susenas. NOTE Decomposition of Theil L (GE(0)) Index (all individuals) into within and between group di erences. Birth circumstances are: head of household gender, parents’ education*, province of birth and whether the birth location was fully urban (whether kotamadya or kabupaten). *Adults’ own education is taken as a proxy for their parents’ education, which is not in the data. However, analysis of IFLS data shows that parents’ education and income are important determinants of children’s educational outcomes, as are availability of schools, all of which are themselves birth circumstances. Non-birth circumstances include children’s e ort. T H E R OL E OF C I R C U M S TA NC E S H A S R E C E N T LY BE E N I NC R E A S I NG . When the four aforementioned circumstances are disaggregated by the year in which the household head was born, it is evident that those children with a household head born between 1968 and 1977, or in other words children whose household head is currently aged between 47 and 38 years old, faced lower levels of inequality of opportunity due to the four circumstances already mentioned. Contribution of inequality of opportunity over time (fig.5) 40 CHI LDR EN WH EN INDONESIA PE RC EN T O F I N E QUA LI TY FIRST STARTS GROWI NG 38.6 % 37.3 % E X PLA IN E D 36.3% 35 3 4.7 % 33.9% F IRST TO B ENEF I T F ROM MASSIVE EDUCATI ONAL EXPANSION 30 HE AD O F HO U SE HO LD HE AD O F HOUSEH OL D H EAD OF H OUSEH OL D H EAD OF H OUSEH OL D HEAD OF HOU S E HOLD BO R N I N 1 9 48 5 7 B O R N I N 1958 67 BORN IN 1968 77 BORN IN 1978 87 AF TER 1 9 87 SOURCE Average of Susenas 2011-13. NOTE Decomposition of Theil L (GE(0)) Index into within and between group di erences. Birth circumstances are: parents’ education, region of birth and urban birth. 3 So the first group is males born in urban Aceh to parents with no education; the second group is females born in urban Aceh to parents with no education; the third group is males born in rural Aceh to parents with no education; an so forth. AN UNFAIR START 11 WHY DOES INEQUALITY OF NOV EM BER 2015 OPPORTUNITY MATTER? W I T H A N I NC R E A S I NG E M PH A S I S ON S K I L L S S HO C K S C A N A F F E C T I N E QUA L I T Y AT A N Y I N T H E MODE R N E C ON OM Y, T HO S E W I T HOU T S TAGE OF T H E F R A M E WOR K B Y E R ODI NG A T H E N E C E S S A RY S K I L L S A R E L E F T BE H I N D. HOUS E HOL D ’ S A BI L I T Y T O E A R N A N I NC OM E , The wages for high-skilled jobs are far higher than S AV E , A N D I N V E S T I N H E A LT H A N D E DUC AT ION . those for low-skilled jobs. At the same time, those Food price volatility plays a crucial role, especially when without the necessary skills to get good jobs are finding poor households spend a majority of their income on themselves trapped in informal or low-productivity jobs. consumption for food. Shocks due to non-communicable Both of these factors are contributing in turn toward diseases are also a cause of vulnerability as these a ect increases in wage inequality. an individual’s ability to work and at the same time may also place an additional cost burden on the household. I N E QUA L I T I E S I N F I N A NC I A L W E A LT H A L S O Furthermore, since the poor tend to work in the DR I V E DI F F E R E NC E S I N I NC OM E . Inequalities in agriculture sector, extreme weather changes caused by financial wealth can drive inequality in income today climate change can negatively a ect crop production. through high returns. Meanwhile, inequalities in financial wealth can drive inequality in income tomorrow through the transmission of greater human and financial capital to the next generation. AN UNFAIR START 12 WHY DOES THE LITERATURE NOV EM BER 2015 TEACH US ABOUT INEQUALITY OF OPPORTUNITY? 2. WHY DOES THE LITERATURE TEACH US ABOUT INEQUALITY OF OPPORTUNITY? P O V E R T Y I S A M U LT I D I M E N S I O N A L of opportunity has evolved out of assessments of the P H E N O M E N O N S U C H T H A T T H E A N A LY S I S O F political philosophy in the late 1980s such as the works I N EQUA L IT Y OF OPP ORT U N IT Y M UST CONSI DE R of Rawls (1971) and Arneson (1989). Sen’s capability D I M E N S I O N S O T H E R T H A N J U S T I N C O M E A N D/ approach (Sen 1976) di ers from previous utilitarian O R C O N S U M P T I O N . There is a general consensus evaluations (more generally, the ‘welfarist’ evaluation) in the literature on poverty that perceives it as a in that he makes room for a variety of human acts and multidimensional phenomenon. Consumption poverty states as important in themselves (not just because they is typically associated with deprivations in other may produce utility, nor just to the extent that they yield dimensions, such as health, education, housing utility). It also makes room for valuing various freedoms— infrastructure, and so on. A multidimensional approach in the form of ‘capabilities’. The capability approach is toward poverty analysis therefore examines each of based on a view of living as a combination of various these di erent dimensions of poverty. ‘doings and beings’, with quality of life being assessed in terms of the capability to achieve valuable functionings. T H E DI S C US S ION ON M U LT I DI M E N S ION A L The key challenge in utilizing Sen’s approach is in P OV E RT Y A N A LY S I S H A S E X T E N DE D BE YON D selecting a class of functionings in the description and AC A DE M IC S , A N D DE V E L OPE D I N T O A BR OA DE R appraisal of capabilities. The focus has to be related to P OL IC Y DE B AT E T O T H E E X T E N T T H AT S OM E the underlying concerns and values, in terms of which C OU N T R I E S H AV E C HO S E N T O A D OP T A some definable functionings may be important and M U LT I DI M E N S ION A L A PPR OAC H I N A N A LY Z I NG others quite trivial and negligible. P OV E RT Y. Mexico’s National Council for the Evaluation E QUA L I T Y OF OPP ORT U N I T Y M E A N S T H AT A L L of Social Policy (CONEVAL) has taken pride in making Mexico the first country to adopt a multidimensional PE OPL E S HOU L D H AV E T H E C H A NC E T O DE V E L OP index as the country’s o cial poverty measure. This E AC H C A PA BI L I T Y. Roemer (1998) writes that there means that poverty in Mexico is no longer narrowly are two guiding principles in the pursuit of equality of defined on economic grounds alone, but also takes full opportunity. The first principle is for societies to “level account of crucial social components of poverty, such the playing field” so that individuals have an equal as the quality of housing and access to health care, chance of obtaining success. An example would be education and food, which are all too often neglected to provide subsidized education for individuals from by established poverty measures.4 disadvantaged backgrounds. The second principle is that of nondiscrimination in the light of competition. T H E T H E OR E T IC A L L I T E R AT U R E ON E QUA L I T Y OF This means that individuals who possess the attributes OPP ORT U N I T Y H A S GR OW N OV E R T H E PA S T T WO relevant for the performance of the duties of a position DE C A DE S , BE GI N N I NG W I T H T H E C ONC E P T OF should be included in the pool of eligible candidates, and ‘ C A PA BI L I T I E S ’.The theoretical literature on equality those who obtain the position are determined by who has 4 See the Oxford Poverty and Human Development Initiative link in the references. AN UNFAIR START 13 WHY DOES THE LITERATURE NOV EM BER 2015 TEACH US ABOUT INEQUALITY OF OPPORTUNITY? the best attributes relevant to that position. An example average access (p¯), and (ii) increases in the equality of of this would be that in selecting an individual’s eligibility opportunity (1–D) of the existing opportunities. for a position race or sex become irrelevant, as the most important issue is whether the individual is able to fulfill The World Bank report on inequality of opportunities in the duties that come with the position. In short, before Latin America and the Caribbean (Barros, et al. 2009) the competition starts, opportunities must be equalized asks how much influence personal circumstances have on by social intervention if need be; but after it begins, children’s access to the basic services that are necessary individuals are judged based on their own performance. for a productive life. For example, is a girl’s probability of having access to clean water (a nutritional must), or piped SIMILARLY, ARNESON (2015) ARGUES THAT EQUALITY sewerage (a health shield), or electricity (a necessity for OF OPPORTUNITY IS A POLITICAL IDEAL THAT reading), or completion of the sixth grade (a predictor of IS OPPOSED TO CASTE HIER ARCHY, BUT NOT TO higher education) in any way a ected by her race, her HIER ARCHY PER SE. Arneson explains that in a caste mother’s literacy, or her father’s salary? As the answers are society the assignment of individuals to places in the social aggregated across services, children, and circumstances, hierarchy is fixed by birth. In contrast, when equality of a picture arises of how equitable (or not) a society is. opportunity prevails, the assignment of individuals to a position in the social hierarchy is determined by some form Utilizing data from the 19 largest Latin American of competitive process, and all members of society are countries covering around 200 million children and eligible to compete on equal terms. spanning roughly the past decade, a Human Opportunity Index is constructed for each of these 19 countries. R E C E N T LY, R E S E A R C H E R S H AV E F O C U S E D ON The report finds that about one-quarter (Colombia) T H E E M PI R IC A L E S T I M AT ION S OF M E A S U R E S OF and half (Guatemala) of the income inequality that is I N E QUA L I T Y OF OPP ORT U N I T Y. A few prominent observed among adults in Latin America is due to the approaches include the Human Opportunity Index (HOI) circumstances they faced when they started out in life—at and the Multidimensional Poverty Index (MPI), each the very outset, through no fault of their own. The report of which are analyzed in detail below. Subsequently, also finds that the most powerful circumstance are the this paper explores other ways in which inequality of mother’s education and father’s income. opportunity can be examined. The HOI focuses on the coverage and inequality of 1. THE HUMAN OPPORTUNITY INDEX HOI opportunities among children for three main reasons. The Human Opportunity Index (HOI) was first developed First, from an empirical standpoint, using children as based on the social welfare function proposed by Sen, a unit of analysis makes it possible to analyze how the which allows di erence weights to be placed on the di erent circumstances they are born in to a ect their outcomes of di erent individuals (Sen 1976). Barros et later achievements, as children (unlike adults) cannot al. (2009) then applied Sen’s framework and proposed a be expected to make the e orts needed to access Human Opportunity Index which provided estimations for these basic goods by themselves. Second, from a Latin America. Their HOI aims to measure the absolute policy standpoint, evidence indicates that interventions level of basic opportunities in a society and how equitably to equalize opportunities early in the lifecycle of an those opportunities are distributed within society. The individual are significantly more cost e ective and index essentially incorporates these measures into one successful than interventions later in life. Third, focusing single composite indicator. on children helps to put inequality of opportunity at the very center of the policy debate. On the day of their The HOI summarizes in a single composite indicator: birth, children cannot be held responsible for their family (i) how many opportunities are available, that is, the circumstances (Barros et al. 2009). coverage rate of a basic service; and (ii) how equitably those opportunities are distributed, that is, whether the The HOI aims to provide an ex ante evaluation of how distribution of that coverage is related to exogenous likely it is that children will do well. The HOI can be circumstances. The first component of the index—the used to track a country’s progress toward the goal of average coverage rate for a given basic opportunity— providing all children with equal access to these basic can be readily determined using household survey opportunities, simultaneously tracking both the overall data. The second component—the equity of opportunity coverage and the equity of their distribution. The HOI distribution—requires a more involved calculation. can therefore serve as a tool to help guide public policies Progress in the HOI can occur by: (i) increases in aimed at equalizing opportunity. AN UNFAIR START 14 WHY DOES THE LITERATURE NOV EM BER 2015 TEACH US ABOUT INEQUALITY OF OPPORTUNITY? 2. THE MULTIDIMENSIONAL POVERTY INDEX MPI The key criticism of multidimensional poverty indices The Multidimensional Poverty Index (MPI) was arises from the fact that the weights used to aggregate published for the first time in the United Nations Human across dimensions lack the intrinsic meaning associated Development 2010 Report. The MPI complements with prices, which are used to add the components of monetary measures of poverty by considering overlapping consumption expenditure (or, implicitly, its dual, the deprivations su ered by people at the same time. The incomes used to finance consumption). Under the law index identifies deprivations across the same three of one price, and given relatively weak assumptions on dimensions as the HDI and shows the number of people preferences, relative prices are equal to the rate at which who are multidimensionally poor (su ering deprivations consumers themselves—regardless of their income levels in 33 percent of weighted indicators) and the number of and allowing for di erent utility functions—are willing to deprivations that poor households typically contend with. trade one such component (e.g., bread) for another (e.g., It can be deconstructed by region, ethnicity and other a bicycle). groupings, as well as by dimension, making it an apt tool for policymakers. Ravallion (2011) suggests a dashboard approach where the focus is on developing the best possible distinct According to UNDP, more than 1.5 billion people, or one- measures of the various dimensions of poverty and a third of the people in the 91 countries covered by the credible set of ‘multiple indices’ rather than a single MPI, live in multidimensional poverty. That means, they ‘multidimensional index’. By utilizing a dashboard live with at least 33 percent of the indicators reflecting approach, it is possible to evaluate the importance of acute deprivation in health, education and standards of each dimension without making any arbitrary decisions. living. This exceeds the estimated 1.2 billion people in However, a key challenge in the use of the dashboard those countries who live on common international poverty approach is that it does not look at the interactions across measure of $1.25 a day or less.5 In addition, close to 800 di erent dimensions. Children who are deprived on one million people are vulnerable to falling into poverty if dimension are not necessarily the same children as those setbacks occur—financial, natural or otherwise. who are deprived on another dimension (Hadiwidjaja, et al. 2013). Sumarto and De Silva (2014) follow Alkire and Foster (2007; 2011) and make use of the National Socioeconomic Ferreira and Lugo (2012) propose three methodological Survey (Susenas) data from Indonesia to explore whether alternatives to the false dichotomy between scalar these multidimensionally deprived households are indices of multidimensional poverty on the one hand, necessarily income poor or not. Sumarto and De Silva and a “dashboard” approach that looks only at marginal (2014) conclude that the overlap between consumption distributions on the other. They argue that such a poverty and multidimensional poverty is extremely weak. dichotomous view misses the point and suggest that Consequently, their findings broaden the targeting space the analysis should examine the interdependence for poverty reduction, suggesting that poverty reduction between dimensions. These alternatives include simple programs should provide di erent kinds of assistance to Venn diagrams of the overlap of deprivations across the poor in di erent dimensions of poverty. dimensions; this can be termed the “associative” approach. The second alternative is the use of Alkire and Foster (2011) and Maasoumi and Lugo (2008) multivariate stochastic dominance analysis, which have proposed scalar indices that seek to combine, permits partial orderings across joint distributions that in a single number, information from those various are robust not only to poverty lines and welfare weights dimensions. One advantage of using a single number is (as in the unidimensional case) but also to dimension that it makes comparison across countries, regions or weights. The third alternative is the analysis of copula individuals, more feasible. However, one key limitation functions to capture the extent of interdependency in the use of multidimensional indices is that it requires across dimensions. Ferreira and Lugo (2012) highlight the use of relative weights for each dimension, which are the importance of examining the joint distribution chosen arbitrarily. between di erent achievements (or deprivations) so that just how closely related di erent achievements (or deprivations) are can be easily identified for the purposes of policymaking. 5 US$1.25, but adjusted for differences in purchasing power in different countries. AN UNFAIR START 15 NOV EM BER 2015 3. WHICH INEQUALITIES OF OPPORTUNITY MATTER MOST IN INDONESIA? T O E X A M I N E T H E DI F F E R E N T C OM P ON E N T S born; (ii) whether the individual was born in a rural or T H AT A F F E C T I N E QUA L I T Y, T H I S PA PE R urban area6; (iii) the income level of the household; (iv) A D OP T S T H E DA S H B OA R D A PPR OAC H A N D T H E the education level of the individual’s parents; and (v) A S S O C I AT ION M E T HOD T O E X A M I N E HOW the gender of the head of the household7. This section BI RT H C I R C U M S TA NC E S A F F E C T C H I L DR E N ’ S analyzes how these five circumstances a ect a child’s OPP ORT U N I T I E S . The circumstances at birth examined access to education, health, and transportation services, & are: (i) which region in Indonesia the individual was how that access in turn can a ect the trend of inequality. 3.1 EDUCATION PA R ENTS’ EDUCATIONA L CIRCU MSTA NCES A N D parents have no education were enrolled in school. CONSUMPTION EX PEN DITU R E A R E BECOMING LESS Similarly, looking at 13- to 15-year-olds’ enrollment by R ELEVA NT TO THEIR CHILDR EN’S EDUCATIONA L parents’ quintile, this also suggests the same patterns ATTA IN MENT. Looking at enrollment of 13- to 15-year- of convergence. The enrollment of 13- to 15-year-olds olds by parents’ educational attainment, it is evident that from the lowest quintile in 2004 used to be less than over the years the enrollment gap between children with 70 percent, while in 2013 the enrollment was more than parents’ with higher levels of education and those with 80 percent. This shows that enrollment has increased lower levels of education is narrowing. This is mainly significantly even within a generation. This suggests because children whose parents have no education or only that if the older child was unable to enroll in junior high primary education are catching up with those children school, the younger child has a higher chance of enrolling. whose parents have higher levels of education. In 2013, Therefore, the trends reveal that mobility in educational more than 80 percent of 13- to 15-year-olds whose parents attainment persists in a positive way in that more have no education were enrolled in school. In contrast, in children are enrolling despite their parents’ educational 2004, less than 70 percent of 13- to 15-year-olds whose background and welfare status. 6 In this case, urban means an urban ward (kelurahan) and rural means a village (desa). 7 We look at five birth circumstances here, adding the household per capita consumption decile, which is known for children (which we analyze in this section), but not for adults (which we analyzed previously). AN UNFAIR START 16 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? Enrolment of 13- to 15-year-olds (fig.6) a) By parents’ education b) By parents’ quintile 100 100 PE RC E N T PE RC E N T 80 80 60 60 2 00 4 2007 2011 2013 2004 2007 2011 2013 N ON E SD SMP Q1 Q2 Q3 SMA T E RTI ARY Q4 Q5 SOURCE Susenas. SOURCE Susenas. NOTE Highest of both parents’ education; parents’ per capita household NOTE Highest of both parents’ education; parents’ per capita household consumption quintile. consumption quintile. W I T H R E S PE C T T O E N R OL L M E N T I N H IGH E R L E V E L S OF E DUC AT ION , PA R E N T S ’ E DUC AT ION A L C I R C U M S TA NC E S S T I L L H AV E A N E F F E C T BU T T H I S I S BE C OM I NG L E S S R E L E VA N T I N DE T E R M I N I NG T H E E DUC AT ION A L AT TA I N M E N T OF T ODAY ’ S A DU LT S . Adults are more likely to have higher levels of education than their parents. Adults between the ages of 44 and 53 whose parents have no education are 1 percent more likely to have a primary education, 12 percent more likely to have a junior secondary education, and 7 percent more likely to have a senior secondary education compared with the baseline (adults aged 54 to 63). Younger adults (between the ages of 34 and 43) whose parents have no education are 15 percent more likely to have a primary education, 13 percent more likely to have a junior secondary education, and 10 percent more likely to have a senior secondary education compared with the baseline (adults aged 54 to 63). Therefore, it is evident that even within one generation, people are more likely to have higher educational attainments than their parents and their older peers. Children’s attainment probability relative to the baseline (fig.7) a) Adults whose parents have no education b) Adults whose parents only have primary 15 15 10 10 5 5 0 0 5 5 P ER CE N T P ER CE N T 10 10 15 15 20 20 25 25 30 30 35 35 40 40 NONE SD SMP S MA NONE SD SMP S MA COHORT BORN 1962 71 COHORT BORN 1972 81 COHORT BORN 1962 71 COHORT BORN 1972 81 SOURCE IFLS. SOURCE IFLS. NOTE The baseline is adults born in 1952-61. NOTE The baseline is adults born in 1952-61. AN UNFAIR START 17 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? DE SPITE THIS TR EN D, HIGH ER EDUCATIONA L and 43) from the bottom quintile are 2 percent less likely ATTA IN M EN T H A S TR A NSL ATED ON LY to have a higher income than the baseline. This means MODER ATELY IN TO HIGH ER INCOM E S, E SPECI A L LY that even though parents’ education has only a small FOR YOU NGER A DU LT GROU PS W HOSE PA R EN TS influence over their children’s educational attainment, H AV E NO EDUCATION. Adults in the bottom quintile these children (the adults of today) still face di culties between the ages of 44 and 53 whose parents have no in obtaining a higher income status. Figure 8 also shows education are only 1 percent more likely to have a higher that, compared with adults from an older age group, adults income than the baseline (those aged between 54 and from the younger age group with better-educated parents 63). Meanwhile, younger adults (between the ages of 34 are more likely to have a higher income. Adult income quintile probability relative to the baseline: (fig.8) a) Adults whose parents have no education b) Adults whose parents only have primary 8 8 7 7 6 6 5 5 4 4 P ERCENT P ERCENT 3 3 2 2 1 1 0 0 1 1 2 2 3 3 Q1 Q2 Q3 Q1 Q2 Q3 COHORT BORN 1962 71 COHORT BORN 1972 81 COHORT BORN 1962 71 COHORT BORN 1972 81 SOURCE IFLS. SOURCE IFLS. NOTE Includes employed only. Results do not change significantly when the NOTE Includes employed only. Results do not change significantly when the unemployed are included. unemployed are included. T H E L I N K B E T W E E N I N C R E A S E S I N E DUC AT ION A L C E N T E R S T H AT H E L P T H E M DE V E L OP “ S C HO OL AT TA I N M E N T A N D I N C R E A S E S I N I N C OM E M AY R E A DI N E S S ”. However, those who live in rural areas B E W E A K A N D C OU L D B E DU E T O I N E QUA L I T Y I N are often left behind. More than 50 percent of children C I R C U M S TA N C E S AT BI R T H . Even though more in Papua do not have access to ECED centers compared children are enrolled in school and educational attainment with just 2 percent of children living in Java/Bali. A World is increasing, the quality of education may still be Bank evaluation of the ECED program finds that children inadequate. Good quality education starts with good who participated in an ECED program have higher levels of schools, but these may only be located in cities and urban school readiness at the age of six than their peers who did areas. Therefore, a child’s circumstances at birth in terms of not participate (World Bank, 2013). Thus, children who have where he/she was born will a ect where and what kind of access to an ECED center will have an advantage later on school he/she attends. Lastly, even if children successfully compared with those who do not have such access. enroll, this does not necessarily mean that they are able to do well in cognitive/national examinations and/or in their AC C E S S T O S C HO OL S AT T H E PR I M A RY S C HO OL subsequent performance in the labor market. L E V E L I S A L MO S T U N I V E R S A L , W I T H T H E E XC E P T ION OF PA P UA . In almost all parts of Indonesia S OM E C H I L DR E N GE T A H E A D S TA RT B Y H AV I NG there is a school within 1km of every village. However, AC C E S S T O E A R LY C H I L DHO OD E DUC AT ION ( E C E D) the situation is very di erent in Papua, where 20 percent AN UNFAIR START 18 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? of households have no primary school within 1km of their to a junior secondary school within 1km of their home. village, and 13 percent of households have no primary However, only 3.6 percent do not have access to a junior school within even 6km of their village. secondary school within 6km of their home. Once again, in Papua more than 25 percent of households do not even J U N IOR S E C ON DA RY L E V E L S C HO OL S A R E H A R DE R have access to junior secondary schools within 6km of T O R E AC H T H A N PR I M A RY S C HO OL S . On average in their home. Indonesia, 25 percent of households do not have access Access to di erent education institutions (fig.9) a) Without access to ECED b) Without access to SD 60 60 50 50 40 40 PERCENT P ERCENT 30 30 20 20 10 10 0 0 RURAL URBAN MH H FHH SUMATERA JAWA/BAL I K AL IMANTAN SUL AWESI NT MALUK U PAP UA RURAL URBAN MH H FHH SUMATERA JAWA/BAL I K AL IMANTAN S UL AWESI NT MALUK U PAP UA c) Without access to SMP 60 50 40 PE RC E N T 30 WITHIN 1 KM 20 AVERAGE 1 KM WITHIN 6 KM 10 AVERAGE 6 KM 0 SOURCE Susenas 2011 and Podes 2011. NOTE MHH are male-headed households, FHH are RU R AL U RB AN MHH FH H SU M ATE R A JAWA /BA LI K AL IMA N TAN S U LAW ES I NT MA LU K U PAPUA female-headed households. Without access to ECED and within 1km of village. Results very similar when considering access within 6km of village AN UNFAIR START 19 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? A LT HOUGH G A P S I N AC C E S S T O S C HO OL S M AY W I L L BE N E F I T MOR E F R OM T H E I R E DUC AT ION BE C L O S I NG OV E R T I M E , T H E R E I S A N E E D T O T H A N T HO S E W HO D O N O T. A teacher’s educational I M PR OV E T H E QUA L I T Y OF E DUC AT ION I N OR DE R attainment can be used as a rough proxy of their quality. T O R E DUC E I N E QUA L I T Y. Not all schools in Indonesia About 16 percent of children attending primary school in have full accreditation and some schools may not even rural areas do not have access to at least two teachers meet the minimum service standards. As a consequence, with an undergraduate degree (S1 degree) compared not all children have access to good schools and are with just 1 percent of children in urban areas. Moreover, therefore at a disadvantage to those who do. What looking at the per capita consumption deciles, it is constitutes a good school is somewhat vague. However, evident that children in the lower per capita consumption it can be defined in terms of the educational attainment of deciles are more likely to have less qualified teachers. the teachers, the availability of electricity and water in the This has policy implications as the Government will bathroom, and the availability of a laboratory. need to provide better qualified teachers with additional incentives to teach in rural areas. Alternatively, the T E AC H E R C OM PE T E NC Y PL AY S A K E Y R OL E I N Government may also need to create specific programs E N S U R I NG G O OD QUA L I T Y E DUC AT ION , S O T HO S E to provide teachers who teach in rural areas with an W HO D O H AV E AC C E S S T O G O OD T E AC H E R S opportunity to upgrade their qualifications. The percentage of children without access to a primary school with at least two teachers with an S1 degree (fig.10) a) By urban/rural, female/male-headed households, b) By household per capita consumption decile and island 50 50 40 40 30 30 PERCENT P ERCENT 20 20 N AT I O N AL AVE R AGE NATIONAL AVERAGE 10 10 0 0 1 2 3 4 5 6 7 8 9 10 RU RA L U RB AN M HH F HH SU M AT E RA JAWA /BA LI K A LI MA N TAN S U L AWE SI NT M ALU K U PAP UA H OUSEH OL D P ER CAP ITA CONSUMPTI ON D E C ILE SOURCE Susenas 2011 and Podes 2011. NOTE MHH are male-headed households, FHH are female-headed households. Access to public SD schools only. L AC K OF AC C E S S T O E L E C T R IC I T Y I N S C HO OL S I S S T I L L A PR OBL E M I N RU R A L PA RT S OF I N D ON E S I A . On average, more than 10 percent of schools in Indonesia do not have access to electricity and most are located in Papua. In fact, more than one in three schools in Papua does not have electricity. The lack of electricity in a school poses major problems, resulting in a lack of good lighting and the inability to use multimedia equipment such as the use of overhead projectors and computers. This means that lessons are delivered less e ectively than in schools that are able to use technology and various multimedia equipment. AN UNFAIR START 20 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? Percentage of children without access to a primary and junior secondary school with electricity (fig.11) a) By urban/rural, female/male-headed households, b) By household per capita consumption decile and island 40 40 30 30 PE RC E N T P E RC E N T 20 20 10 10 N AT I O N AL AVE R AGE NATIONAL AVERAGE 0 0 1 2 3 4 5 6 7 8 9 10 RURAL URB AN MHH FHH SUMAT E RA JAWA /BALI K ALI MANTAN SULAWE SI NT MALUK U PA PUA H OUSEH OL D P ER CAP ITA CONSUMPTI ON D E C ILE SOURCE Susenas 2011 and Podes 2011. NOTE MHH are male-headed households, FHH are female-headed households. ACCESS TO WATER IN SCHOOL BATHROOMS IS STILL A CHALLENGE IN INDONESIA WITH THE RESULT THAT STUDENTS ARE MORE SUSCEPTIBLE TO DISEASE. On average, 18 percent of schools in Indonesia do not have running water in their bathrooms. Some regions, such as Maluku and Papua, have double the level of schools without water in their bathrooms compared with Java and Bali. UNICEF (2005) contends that the lack of clean water and sanitation in schools jeopardizes the quality education, as clean water and sanitation are both essential to protect children’s health and their ability to learn at school. In this sense, they are as vital as textbooks to a child’s education. No water in bathrooms means that students do not wash their hands after they have been to the toilet, or wash their hands before they eat. This is highly unsanitary and increases the potential for the spread of diseases such as diarrhea and typhoid. Percentage of children without access to a primary and junior secondary school without water in bathroom (fig.12) a) By urban/rural, female/male-headed households, b) By household per capita consumption decile and island 35 35 30 30 25 25 PE RC EN T 20 20 N AT I O N AL AVE R AGE NATIONAL AVERAGE PE RC E N T 15 15 10 10 5 5 0 0 1 2 3 4 5 6 7 8 9 10 RU R AL U RB AN MH H FHH SU M ATE R A JAWA/B AL I K AL IM AN TA N S UL AWE SI NT MA LU K U PA PUA H OUSEH OL D P ER CAP ITA CONSUMPTI ON D E C ILE SOURCE Susenas 2011 and Podes 2011. NOTE MHH are male-headed households, FHH are female-headed households. AN UNFAIR START 21 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? L AC K OF AC C E S S T O L A B OR AT OR I E S I N J U N IOR S E C ON DA RY S C HO OL M E A N S T H AT I N D ON E S I A N S T U DE N T S H AV E L E S S A DVA N TAGE I N S C I E NC E -B A S E D S U B J E C T S . Access to science laboratories in schools is important in ensuring that students are not only learning theories, but also exposed to practical work. On average, 70 percent of junior secondary schools in Indonesia do not have access to a laboratory. Urban and rural di erences and di erences across a region are only 10 percent at most. This is worrying as it means that most Indonesian students do not have the facilities to enable them to do well in science. In fact, Indonesia’s OECD Pisa 2012 and 2015 results reveal that 15-year-old students in Indonesia (or those who are expected to be in the last grade of junior secondary school) score badly in science compared with other countries in the region. The percentage of children without access to a junior secondary school with a laboratory (fig.13) a) By urban/rural, female/male-headed households, b) By household per capita consumption decile and island 80 80 N AT I O N AL AVE R AGE NATI ONAL AVERAGE 70 70 60 60 50 50 PE RC E NT PE RC E NT 40 40 30 30 20 20 10 10 0 0 1 2 3 4 5 6 7 8 9 10 RURAL URBAN MH H FHH SUMATERA JAWA/BAL I K AL IMANTAN SUL AWESI NT MALUK U PAP UA H OUSEH OL D P ER CAP ITA C ONSUMPTI ON D E C ILE SOURCE Susenas 2011 and Podes 2011. NOTE MHH are male-headed households, FHH are female-headed households. Access to public SMP schools only. E V E N I F T H E R E I S AC C E S S T O S C HO OL S , T H I S D OE S N O T N E C E S S A R I LY M E A N T H AT E V E RY C H I L D I S E N R OL L E D. While most households do have primary schools in their village, nonetheless on average 2.5 percent of children fail to enroll in primary school (Figure 9b and Figure 14a). Therefore, the challenges in enrollment may go beyond supply constraints. Furthermore, household per capita consumption also reveals that children from households with higher per capita consumption are more likely to be enrolled than those who come from households with lower per capita consumption. This further points toward demand-based constraints. Percentage of children aged 7-12 years who are not enrolled (fig.14) a) By urban/rural, female/male-headed households, b) By household per capita consumption decile and island 25 25 20 20 PE RC E N T PE RC E N T 15 15 10 10 5 5 0 0 R U RAL U RB AN M HH F HH SU MAT E RA JAWA/ B ALI K AL I MAN TA N S U LAW ES I NT M ALU K U PAPUA 1 2 3 4 5 6 7 8 9 10 H OUSEH OL D P ER CAP ITA CONSUMPTI ON D E C ILE 2002 2002 AVERAGE 2011 2011 AVERAGE 2002 2011 SOURCE Susenas and Podes. NOTE MHH are male-headed households, FHH are female-headed households. Data not available for Maluku and Papua in 2002. AN UNFAIR START 22 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? E V E N I F T H E R E I S A G O OD S C HO OL N E A R B Y A N D evidence of the links between stunting and cognitive C H I L DR E N A R E E N R OL L E D, C I R C U M S TA NC E S AT ability (Mendez and Adair 1999; Levitsky and Strupp BI RT H M AY S T I L L PR E V E N T S T U DE N T S F R OM 1995). The likelihood of a child whose parents have less AC QU I R I NG T H E N E C E S S A RY S K I L L S . One example than a primary level of education experiencing stunting of such adverse circumstances at birth is stunting. is 38 percent compared with 27 percent for those Children with parents who have lower levels of education children who have at least one parent with a senior are more likely to be stunted. Meanwhile, there is strong secondary level of education. Stunting probability (fig.15) a) by parents’ education b) by parents’ income 50 50 40 40 PE RC E NT P E RC E NT 30 30 20 20 10 10 0 0 L ES S SD SMP SMA AND Q1 2 3 4 Q5 THAN SD ABOVE H OUSEH OL D CONSUMPTI ON QUINTILE SOURCE IFLS 2007. NOTE Extreme stunting as <-2 standard deviations z-score height-for-age using 2006 WHO standards. T H E R E I S A L S O E V I DE NC E T H AT C I R C U M S TA NC E S in the bottom quintile are more likely to have the bottom AT BI RT H , S UC H A S PA R E N T S ’ E DUC AT ION A N D score and only 14 percent are likely to have the top score. C ON S U M P T ION E X PE N DI T U R E , H AV E A N E F F E C T In sharp contrast, 19 percent of children in the highest ON C O GN I T I V E AC H I E V E M E N T S . Children whose quintile are more likely to have the bottom score, while 26 parents have lower levels of education are more likely percent are likely to have the top score. All in all, parents’ to have the bottom score and less likely to score higher income and educational attainment seem to have an in cognitive tests than children whose parents have impact on children’s cognitive scores. This may be due to higher levels of education. However, di erences are even several factors, for instance, parents with higher incomes more dramatic if scores are disaggregated by parents’ can a ord additional lessons outside school that improve income. Looking at di erences in scores in relation to their children’s ability, or if the parents have higher levels parents’ income, it appears that 40 percent of children of education they are able to tutor their children. Cognitive score probability (fig.16) SOURCE IFLS 2007. NOTE Extreme stunting as <-2 standard deviations a) by parents’ education z-score height-for-age using 2006 WHO standards. 50 40 30 P E RC EN T 20 10 0 Q1 2 3 4 Q5 HOUSEH OL D CONSUMPTI ON QUINTI L E BOTTOM SCORE QUINTILE TOP SCORE QUINTILE AN UNFAIR START 23 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? Cognitive score probability (fig.16) cont. SOURCE IFLS 2007. NOTE Extreme stunting as <-2 standard deviations z-score height-for-age using 2006 WHO standards. a) by parents’ education b) by parents’ income 40 40 35 35 30 30 25 25 PE RC E N T P E RC E N T 20 20 15 15 10 10 5 5 0 0 LESS SD SMP SMA AND P OOREST Q2 Q3 Q4 R IC HE ST THAN SD ABOVE BOTTOM QUINTILE OF TOP QUINTILE OF BOTTOM QUINTILE OF TOP QUINTILE OF COGNITIVE SCORES COGNITIVE SCORES COGNITIVE SCORES COGNITIVE SCORES S I M I L A R LY, N AT ION A L E X A M I N AT ION S C OR E S A L S O R E V E A L T H AT PA R E N T S ’ E DUC AT ION A N D C ON S U M P T ION E X PE N DI T U R E A R E P O S I T I V E LY C OR R E L AT E D W I T H E X A M I N AT ION S C OR E S . Observing the examination scores of students aged 10-15 years of age, it is evident that students whose parents have lower levels of education have lower test scores than those who have parents with higher levels of education. Similarly, students who come from poorer households have lower test scores than those who come from households with higher levels of consumption. All this means that some children cannot gain the same advantages from schooling, even if they have access to schools. National test scores of 10- to 15-year-olds (fig.17) a) by parents’ education b) by parents’ income 8 8 7 7 6 6 AV ER AG E SCO R E AV ER AG E SCO R E 5 5 4 4 3 3 2 2 1 1 0 0 B A H ASA MAT HE MAT ICS SCI E N CE SOCI AL B AH ASA MATH EMATICS SCI ENCE S OC IAL H OUSEH OL D CONSUMPTI ON QUI NT I LE LESS THAN SD SD SMP SMA+ Q1 Q2 Q3 Q4 SOURCE IFLS 2007. NOTE National test scores for 10- to 15-year olds. AN UNFAIR START 24 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? 3.2 HEALTH C I R C U M S TA NC E S AT BI RT H I N T E R M S OF W H E R E no access to a midwife, the percentage of unskilled S OM E ON E I S B OR N H IGH LY A F F E C T S AC C E S S deliveries is significantly more than this, at over 40 T O A M I DW I F E A N D T O A N E X T E N T U N S K I L L E D percent. Similarly, for the richer deciles, having no access DE L I V E RY. People born in eastern Indonesia—namely to a midwife only partially explains the level of unskilled Papua, Maluku, Nusa Tenggara, and Sulawesi—are more birth deliveries, albeit less dramatically than in the likely to have no midwife in their village than people born poorest decile. in the rest of Indonesia. Papua has the lowest access to a midwife, with 40 percent of Papuans having no access BE I NG P O OR I S A S S O C I AT E D W I T H U N S K I L L E D to a midwife. Correspondingly, unskilled birth deliveries DE L I V E R I E S , BU T T H I S I S N O T T H E ON LY are most common in rural areas. The percentage of E X PL A N AT ION . More than 40 percent of households unskilled deliveries in Maluku is the highest in the country, from the poorest decile have unskilled deliveries representing close to 60 percent of total births. compared with around 5 percent of all households. As the household per capita consumption increases, there A LTHOUGH ACCESS TO A MIDW IFE IS A CH A LLENGE , is a decrease in unskilled deliveries. However, of the THIS FA ILS TO E X PL A IN THE L A RGE PERCEN TAGE 24 percent unskilled deliveries across Indonesia as a OF U NSK ILLED DELI V ER IES, ESPECI A LLY FOR whole, only 5 percentage points are poor households. HOUSEHOLDS FROM THE POOR EST DECILE . While This points toward other reasons behind the high level of only 10 percent of households in the poorest decile have unskilled deliveries. Unskilled birth deliveries (fig.18) SOURCE Susenas and Podes, 2011. a) By urban/rural, female/male-headed households, b) By household per capita consumption decile and island 60 60 50 50 40 40 PE RCENT PE RCENT 30 30 20 20 10 10 0 0 1 2 3 4 5 6 7 8 9 10 R U RA L U R BA N MH H FH H S U MATE RA JAWA/ BA LI K A LIM AN TA N SU L AWE SI NT MALU K U PA PUA H OUSEH OL D P ER CAP ITA CONSUMPTI ON D E C ILE IN SHORT, E V EN W H EN ACCE S S E X ISTS A N D PEOPL E levels of unskilled delivery may be rather more complex. A R E ECONOMICA L LY SECU R E , OTH ER BA R R IER S For instance, unskilled deliveries may be due to the limited A LSO SEEM TO PR E V EN T TH E USE OF H E A LTH CA R E number of midwives compared with the number of pregnant SERV ICE S. Looking at the relationship between being poor, women, which points toward a supply constraint. It could having no midwife in the village and unskilled delivery, only 1 also be because people do not trust the competence of percent of the population has all of these three circumstances. midwives, or it could be due to cultural beliefs that favor more Interestingly, only 3 percent of the unskilled deliveries may traditional methods. Thus, the first challenge is to identify the be related to having no midwife in the village. Similarly, only constraints to equal opportunities where access does exist, 5 percent of unskilled deliveries are associated with being and once these are known the second challenge is then to poor. This therefore suggests that the reasons behind the high decide how to address these constraints. AN UNFAIR START 25 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? C I R C U M S TA NC E S AT BI RT H I N T E R M S OF PA R E N T S ’ E DUC AT ION A L AT TA I N M E N T PR OV I DE S A BET T E R E X PL A N AT ION F OR U N S K I L L E D DE L I V E R I E S T H A N E C ON OM IC S TAT US OR T H E A B S E NC E OF A M I DW I F E . More than one-third of unskilled deliveries are associated with parents not completing even primary education. This means that the type of intervention that is needed may require extensive socialization of the importance of skilled delivery, as well as the provision of qualified midwives. This also begs the question as to what other factors influence unskilled delivery. Relationship between poverty, midwives Relationship between parents’ educational attainment, and birth (fig.19) midwives and birth (fig.20) 13 % 5 % No midwife 21 % 5 % No midwife Poor 8% Parents 1% in village no SD 13 % 1% 2% in village 2% 1 % 1 % 4% 2% 2% 6% 17% 24 % Unskilled 15 % 24 % Unskilled delivery delivery SOURCE Susenas and Podes, from Hadiwidjaja, Paladines and Wai-Poi (2013). SOURCE Susenas and Podes, from Hadiwidjaja, Paladines and Wai-Poi (2013). NOTE Poorest 40 means the child lives in a household that is in the poorest NOTE Poorest 40 means the child lives in a household that is in the poorest 40 percent of households nationally. 40 percent of households nationally. L O OK I NG AT AC C E S S T O PR I M A RY H E A LT H C A R E F OR T H E W I DE R P OP U L AT ION OF I N D ON E S I A , MO S T PE OPL E H AV E AC C E S S T O PR I M A RY H E A LT H C A R E , A N D Y E T U R B A N/ RU R A L , R E GION A L A N D W E A LT H G A P S S T I L L E X I S T. People living in eastern Indonesia, especially those living in Papua, are more likely to experience di culties in accessing primary health care. More than 25 percent of Papuans still face challenges in reaching primary health care services such as polyclinics, Puskesmas, Puskesmas Pembantu, or a physician’s practice. As expected, the poorest deciles are more likely to have di culties in accessing primary health care than richer deciles, possibly due to the fact that rural areas mostly consist of poorer households. Percentage of people who cannot access primary health care (fig.21) a) By urban/rural, female/male-headed households, b) By household per capita consumption decile and island 30 30 25 25 20 20 P ER CE N T PE RC EN T 15 15 10 10 5 5 N AT I O N AL AVE R AGE NATIONAL AVERAGE 0 0 U RB AN MHH FHH SU MAT E RA JAWA/ BA LI K AL IM AN TAN SU LAW E SI NT M ALU K U 1 2 3 4 5 6 7 8 9 10 RU R AL PAP UA H OUSEH OL D P ER CAP ITA CONSUMP TION DE C I LE SOURCE Susenas 2011 and Podes 2011. NOTE Share of the population that cannot easily reach facilities. Even when access exists, barriers can remain to uptake: policy needs to identify and address these barriers. AN UNFAIR START 26 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? T H E C H A L L E NGE S I N AC C E S S I NG S E C ON DA RY H E A LT H C A R E A R E E V E N GR E AT E R T H A N T HO S E I N AC C E S S I NG PR I M A RY H E A LT H C A R E , PA RT IC U L A R LY I N E A S T E R N I N D ON E S I A . At the national average, 3 percent of people have di culty accessing primary health care compared with 16 percent for secondary health care. As before, eastern Indonesia has the most di culty in accessing secondary health care. In Papua, a little over 60 percent of people have di culty accessing hospitals, compared with 8 percent for those living in Java/Bali. This means that out of 10 people in Papua, only four have access to a hospital. Percentage of people who cannot access hospital (fig.22) a) By urban/rural, female/male-headed households, b) By household per capita consumption decile and island 70 70 60 60 50 50 PE RC E NT PE RC E NT 40 40 30 30 20 N AT I O N AL AVE R AGE 20 NATIONAL AVERAGE 10 10 0 0 1 2 3 4 5 6 7 8 9 10 RURAL URBAN MH H FHH SUMATERA JAWA/BAL I K AL IMANTAN SUL AWESI NT MALUK U PAP UA H OUSEH OL D P ER CAP ITA CONSUMPTI ON D E C ILE SOURCE Susenas 2011 and Podes 2011. NOTE MHH are male-headed households, FHH are female-headed households. F U RT H E R MOR E , E V E N W H E R E H E A LT H C A R E S E RV IC E S D O E X I S T, T H E Y OF T E N L AC K T H E FAC I L I T I E S N E C E S S A RY T O PR OV I DE PR OPE R S E RV IC E S . Not all Puskesmas have a doctor, or even the most basic facilities such as running water and electricity. More than one in four Puskesmas in Maluku and one in 10 in Papua do not have a doctor, compared with the national average of one in three. One in 25 Puskesmas in Indonesia does not have access to running water. This means that water is either obtained from a well, or even possibly from an unprotected source. Meanwhile, on average in Indonesia, one in 50 Puskesmas does not have electricity, and Papua has the worst access with more than one in four Puskesmas having no access to electricity. Percentage of people who cannot access hospital (fig.23) a) Without a doctor b) Without running water c) Without electricity 30 30 30 25 25 25 20 20 20 PE R CE N T PE R CE N T PE R CE N T 15 15 15 10 10 10 5 NAT ION A L AVE R AGE 5 NATIONAL AVERAGE 5 NATI ONAL AVERAG E 0 0 0 RU R AL U RB AN MH H FH H S U MATE R A JAWA/B ALI KA LI MA N TAN S U LAWE S I NT MA LU KU PAP UA RU R AL U RB AN MH H FH H S U MATE R A JAWA/B ALI KA LI MA N TAN S U LAWE S I NT MA LU KU PAP UA RU R AL U RB AN MH H FH H S U MATE R A JAWA/B ALI KA LI MA N TAN S U LAWE S I NT MA LU KU PAP UA SOURCE Susenas 2011 and Podes 2011. NOTE MHH are male-headed households, FHH are female-headed households. AN UNFAIR START 27 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? WHILE HEALTH CARE OUTCOMES ARE IMPROVING, THERE ARE OBVIOUS REGIONAL DISPARITIES WITHIN INDONESIA AND IN COMPARING INDONESIA WITH OTHER COUNTRIES IN THE REGION. On average, 4 percent of children in Indonesia have never been immunized, and the numbers are especially high for Papua with almost 18 percent of children never having been immunized. Lack of access to health care may also explain the high level of maternal mortality. In 2013, Indonesia’s maternal mortality ratio was almost four times of that of Vietnam’s, a country with a lower GDP per capita than Indonesia. In Indonesia, 190 out of 100,000 mothers die in childbirth, compared with 49 in Vietnam (World Development Indicators). WHILE CHILDREN FROM HOUSEHOLDS WITH A HIGHER PER CAPITA CONSUMPTION ARE LESS LIKELY NOT TO BE IMMUNIZED, THERE IS STILL ROOM FOR IMPROVEMENT AS SOME CHILDREN EVEN IN THE RICHEST DECILE HAVE NEVER BEEN IMMUNIZED. Almost 8 percent of children from the poorest decile compared with 1 percent of children from the richest decile have never received any form of immunization. Immunization plays a significant role in ensuring that children are protected from highly preventable diseases such as polio. Percentage of children under 5 who have never been immunized (fig.24) a) By urban/rural, female/male-headed households, b) By household per capita consumption decile and island 20 20 15 15 P ERCENT PERCENT 10 10 N AT I O N AL AVE R AGE NATI ONAL AVERAGE 5 5 0 0 1 2 3 4 5 6 7 8 9 10 RURAL URBAN MH H FHH SUMATERA JAWA/BAL I K AL IMANTAN SUL AWESI NT MALUK U PAP UA H OUSEH OL D P ER CAP ITA CONSUMPTI ON D E C ILE SOURCE Susenas 2011 and Podes 2011. NOTE MHH are male-headed households, FHH are female-headed households. AC H I E V E M E N T S I N I M M U N I Z AT ION C OM PL E T ION C A N BE I M PR OV E D A S T H E C U R R E N T AV E R AGE L E V E L I S S T I L L L OW. Furthermore, with the exception of Papua, there are no significant di erences between complete child immunization across regions, while di erences in completion across welfare status are negligible. At the national average, less than 17 percent of children have complete immunization. Di erences across regions are not too significant, although less than 6 percent of children under five in Papua have complete immunization. The percentage of children who have completed immunization are not dramatically di erent between di erent deciles. Percentage of children under five who have completed immunization (fig.25) a) By urban/rural, female/male-headed households, b) By household per capita consumption decile and island 25 25 20 N AT I O N AL AVE R AGE 20 NATIONAL AVERAGE P E RC EN T PE RC EN T 15 15 10 10 5 5 0 0 1 2 3 4 5 6 7 8 9 10 RU R AL U RB AN MHH FHH S U MAT E RA JAWA/B AL I K AL IM AN TA N SU LAW ES I NT MA LU K U PA PUA HOUSEH OL D P ER CAP ITA CONSUMPTI ON D E C ILE SOURCE Susenas 2011. NOTE MHH are male-headed households, FHH are female-headed households. All immunizations except fourth dose of HepB. AN UNFAIR START 28 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? L O OK I NG S PE C I F IC A L LY AT E DUC AT ION A N D H E A LT H AC C E S S A N D S E PA R AT I NG U R B A N F R OM RU R A L A R E A S , I T I S C L E A R T H AT L AC K OF AC C E S S T O H E A LT H C A R E A N D E DUC AT ION I N RU R A L A R E A S I S L A R GE LY A S S O C I AT E D W I T H L AC K OF T R A N S P ORTAT ION I N F R A S T RUC T U R E . In rural areas, 20 percent of children have poor education, and health access and poor transportation. By contrast, only 1 percent of children in urban areas have all three negative indicators. Therefore, the solution in rural areas may not necessarily lie in creating new schools or health care facilities, but rather in improving the transportation infrastructure so that people are able to get to these facilities more easily. Meanwhile, interventions for urban areas that focus on improving transportation services will be less e ective in improving access to health and education, as only 5 percent of urban children who lack access to health and education also lack access to transportation. Relationship between poor health access, poor education and poor transportation (fig.26) a) Urban b) Rural 7 % 41% 9% 5% 8% 4% 3 % 2 % 3 % 6% 1% Poor Poor 2 Education Education % 2% Poor Poor 20% 6% Health Access Health 5% Access Access Access 22% 26% 18 % 50% Poor Poor Transportation Transportation SOURCE Susenas and Podes, from Hadiwidjaja, Paladines and Wai-Poi (2013). NOTE Children are deemed to have poor physical access to health and education infrastructure if they live in a district whose index for each of these is in the lowest 40 percent. Transportation Infrastructure T H E PE R C E N TAGE OF M A I N R OA D S N O T GR AV E L E D OR A S PH A LT E D I S H IGH E S T I N RU R A L A R E A S , A N D I N PA P UA I N PA RT IC U L A R . In Papua, over 35 percent of main roads are not graveled or asphalted, while the figure is less than 1 percent in Java/Bali. Lack of good roads means that it takes longer for goods to be delivered, and makes the delivery of perishable goods more challenging. Moreover, lack of good roads means that travel will also take longer, whether this is to schools or to health care centers. Therefore, good road infrastructure plays a crucial role in development as it better connects people with services, food, and markets. AC C E S S T O GR AV E L E D OR A S PH A LT E D M A I N R OA D S D OE S N O T DI F F E R T O O S IGN I F IC A N T LY AC R O S S HOUS E HOL D W E L FA R E S TAT U S . At the national average, around 5 percent of roads are not graveled or asphalted. For the poorest decile, however, the level 50 percent higher than the national average. Percentage of main roads not graveled or asphalted (fig.27) a) By urban/rural, female/male-headed households, b) By household per capita consumption decile and island 40 40 30 30 PE RC E N T PE RC E N T 20 20 10 N AT I O N AL AVE R AGE 10 NATIONAL AVERAGE 0 0 1 2 3 4 5 6 7 8 9 10 PAP UA RU R AL U RB AN MHH FHH SU MAT E RA JAWA/ BA LI K AL IM AN TA N SU LAW E SI NT M ALU K U H OUSEH OL D P ER CAP ITA CONSUMP TION DE C I LE SOURCE Susenas 2011 and Podes 2011. NOTE MHH are male-headed households, FHH are female-headed households. AN UNFAIR START 29 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? I N A DDI T ION T O R OA D S , BR I D GE S A R E A L S O I M P ORTA N T A N D T H E N E E D F OR A DDI T ION A L BR I D GE S I S MO S T PR E VA L E N T I N RU R A L A R E A S . The need for additional bridges in Indonesia a ects around 17 percent of people and is significantly higher in eastern Indonesia. Interestingly, Nusa Tenggara is in need of more bridges than Papua, with 35 percent and 28 percent, respectively. Bridges are particularly useful in connecting people and services. The existence of a bridge can significantly cut the time needed for distribution and allow people to access services more easily. T H E N E E D F OR A DDI T ION A L BR I D GE S DI F F E R S S L IGH T LY AC R O S S HOUS E HOL D PE R C A PI TA C ON S U M P T ION . Around 22 percent of households in the poorest decile need more bridges compared with half this level for households in the highest decile. Di erences across households may stem from the fact that poorer households are mostly located in rural areas, while richer households tend to be mostly located in urban areas. Percentage need of additional bridges (fig.28) a) By urban/rural, female/male-headed households, b) By household per capita consumption decile and island 35 35 30 30 25 25 PERC E NT PERC E NT 20 N AT I O N AL AVE R AGE 20 NATI ONAL AVERAGE 15 15 10 10 5 5 0 0 1 2 3 4 5 6 7 8 9 10 PAPUA RURAL URBAN MH H FHH SUMATERA JAWA/BAL I K AL IMANTAN SUL AWESI NT MALUK U H OUSEH OL D P ER CAP ITA CONSUMPTI ON DE C ILE SOURCE Susenas 2011 and Podes 2011. NOTE MHH are male-headed households, FHH are female-headed households. T H E I N A DE QUAC Y OF P U BL IC T R A N S P ORTAT ION I N I N D ON E S I A I S C L E A R , A LT HOUGH S E RV IC E S A R E S L IGH T LY BET T E R I N JAVA A N D B A L I . One way to identify the availability of public transportation is to estimate the percentage of households without access to regular transport to their local district government o ce. On average, 27 percent of households in Indonesia have no access to regular public transport to their local district o ce. As previously, eastern Indonesia has the worst access to a public transportation system, with 55 percent of households in Papua and 49 percent of households in Kalimantan without regular transport to the district o ce. What is interesting is that, looking at the per capita consumption decile, there are no significant di erences across deciles. Percentage with no regular transport to local district o ce (fig.29) a) By urban/rural, female/male-headed households, b) By household per capita consumption decile and island 60 60 50 50 P E RC EN T P E RC EN T 40 40 N AT I O N AL AVE R AGE NATI ONAL AVERAGE 30 30 20 20 10 10 0 0 1 2 3 4 5 6 7 8 9 10 PA PUA RU RA L U RB AN M HH F HH S U MAT ER A JAWA/B AL I KA LI MA N TAN S U LAWE S I NT MA LU KU H OUSEH OL D P ER CAP ITA CONSUMPTI ON D E C ILE SOURCE Susenas 2011 and Podes 2011. NOTE MHH are male-headed households, FHH are female-headed households. AN UNFAIR START 30 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? OTHER FORMS OF INFRASTRUCTURE SUCH AS HOUSING BEHIND HOUSEHOLDS WITH HIGHER PER CAPITA AND SANITATION ARE JUST AS IMPORTANT IN CONSUMPTION. Poor quality materials represent the DETERMINING QUALITY OF LIFE AND ALSO INFLUENCE main reason for substandard housing across all deciles. INEQUALITY OF OPPORTUNITY. Without good housing In 2011, 70 percent of households from the poorest decile conditions, people are more susceptible to illnesses and had poor quality housing materials, compared with almost more vulnerable to adverse weather conditions. Moreover, half that for households coming from the richest decile. the absence of electricity also means that housework There are also large di erences across deciles in terms must be done manually, there is inadequate lighting for of sanitation, with 67 percent being substandard in the children to study, and it is more di cult to obtain water. poorest decile, but only 12 percent substandard for the richest decile. Around 30 percent of households from MOST INDONESIANS HAVE SUBSTANDARD HOUSING, the poorest decile have unclean water, while this figure ESPECIALLY THE POOR LIVING IN THE RURAL PARTS OF is only 5 percent in the richest decile. There has been EASTERN INDONESIA. Substandard housing means that a significant improvement in access to electricity for a house is either constructed from poor quality materials, the poorest households in recent years. Between 2002 is overcrowded, or lacks access to clean drinking water, and 2011, the percentage of households in the lowest electricity or proper sanitation. Around 83 percent of consumption decile with no electricity almost halved, from houses in rural areas are substandard, compared with 29 percent to 16 percent. 60 percent in urban areas. Java and Bali have relatively better housing than other parts of Indonesia, with 56 HOUSEHOLD PER CAPITA CONSUMPTION IS HIGHLY percent substandard housing, compared with 86 percent CORRELATED WITH THE QUALITY OF HOUSING, AND in Nusa Tenggara, 90 percent in Sumatra, 96 percent in THERE IS LITTLE CHANGE IN THE PERCENTAGE OF Kalimantan, 98 percent in Sulawesi, and 99 percent in SUBSTANDARD HOUSING FROM THE POOREST TWO both Maluku and Papua. QUINTILES. Between 2002 and 2011, the percentage of substandard housing of the poorest decile decreased from WHILE THE PERCENTAGE OF PEOPLE WITH 95 percent to 90 percent. What is interesting is that the SUBSTANDARD HOUSING HAS DECLINED OVER THE percentage of substandard housing for the richest decile YEARS, THE HOUSING CONDITIONS OF HOUSEHOLDS remained constant at 44 percent. WITH LOW PER CAPITA CONSUMPTION ARE FAR Percentage of substandard housing (fig.30) a) By urban/rural, female/male-headed households, b) By household per capita consumption decile and island 100 100 80 80 60 60 P E RC EN T P E RC EN T 40 40 20 20 0 0 1 2 3 4 5 6 7 8 9 10 PAPUA RU RA L U RB AN M HH F HH S U MATE RA JAWA/ B ALI K A LI MA N TA N SU L AWE SI NT MALU K U H OUSEH OL D P ER CAP ITA CONSUMPTI ON D E C ILE 2002 2 0 11 2002 2011 SOURCE Susenas. NOTE All results in this section are for children aged under 15 years. MHH are male-headed households, FHH are female-headed households. Data were not available for Maluku and Papua in 2002. Substandard housing means housing is either constructed from poor quality materials, is overcrowded, or lacks access to clean drinking water, electricity or proper sanitation. AN UNFAIR START 31 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? Housing conditions by household per capita consumption decile (fig.31) P O OR QUA L I T Y H O U SI N G PO O R SAN I TAT I ON P OOR WATER P OOR EL ECTRI C I T Y M AT E RI A L S 1 00 100 100 100 80 80 80 80 60 60 60 60 40 40 40 40 20 20 20 20 0 0 0 0 1 10 1 10 1 10 1 10 SOURCE Susenas. NOTE MHH are male-headed households, FHH are female-headed households. Data were not available for Maluku and Papua in 2002. Poor quality housing means walls, floor or roof are constructed from poor quality materials. T H E R E A R E NO T ICE A BL E DI F F E R E NCE S I N HOUSI NG CH A R ACT E R I ST IC S BET W E E N HOUSE HOL D S L O CAT E D I N RU R A L A N D U R B A N A R E A S I N T H AT CH I L DR E N L I V I NG I N RU R A L A R E A S A R E MOR E L I K E LY T O L ACK ACCE S S T O CL E A N WAT E R , S A N I TAT ION, A N D G O OD HOUSI NG SI M U LTA N E OUSLY T H A N CH I L DR E N L I V I NG I N U R B A N A R E A S . About 18 percent of children living in rural areas lack access to clean water, sanitation, and good housing, while only 2 percent of children in urban areas su er from these three poor housing conditions combined. In view of this, the interventions required will need to be di erent between urban and rural areas. Percentage of poor housing conditions (fig.32) SOURCE Susenas and Podes, from Hadiwidjaja, Paladines and Wai-Poi (2013). a) By urban/rural, female/male-headed households, and island b) By household per capita consumption decile 46% 6% 65% 27% Poor Unclean Poor 19 % Unclean Housing Drinking Housing 4 % Drinking Water Water 2% 32 % 2 % 1% 18 % 2 % 1% 3% 25 % 10% 10% 58% Poor 23 12 % % Sanitation Poor Sanitation AN UNFAIR START 32 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? B Y C OM PA R I NG T H E T WO E X T R E M E S —T H E R IC H E S T DE C I L E W HO L I V E I N U R B A N A R E A S A N D T H E P O OR E S T DE C I L E W HO L I V E I N RU R A L A R E A S —I T I S C L E A R T H AT I N D ON E S I A I S V E RY U N E QUA L . Compared with the other indicators analyzed, inadequate sanitation represents the greatest challenge. Various indicators by urban and rural (fig.33) SOURCE Susenas 2012 and DHS 2007. No phone Improper sanitation No clean water Dirt floor 16-18 y.o. not enrolled 7-15 y.o. not enrolled Primary completion Not fully immunized Low birth weight Unskilled birth attendant No antenatal visit 0 10 20 30 40 50 60 70 80 DECILE 10+ URBAN DECILE 1+ RURAL B O X .1 What specific policies are needed in eastern Indonesia? T H E R E A R E L A R GE DI S PA R I T I E S BE T W E E N E A S T E R N I N D ON E S I A A N D T H E R E S T OF I N D ON E S I A . Comparing urban Jakarta and rural Maluku or Papua, there are clear di erences in terms of inequality of opportunity across almost all indicators. Various di erent indicators by urban (DKI Jakarta) and SOURCE Susenas 2012 and Demographic and Health Survey 2007. rural (Maluku/Papua) (fig.34) No phone Improper sanitation No clean water Dirt floor 16-18 y.o. not enrolled 7-15 y.o. not enrolled Primary completion Not fully immunized Low birth weight Unskilled birth attendant No antenatal visit 0 20 40 60 80 100 URBAN, DKI JAKARTA, HOH SECONDARY EDUC, NON POOR RURAL, MALUKU/PAPUA, HOH NO EDUCATION, POOR A R OU N D H A L F OF A L L C H I L DR E N I N PA P UA L AC K on most dimensions. In fact, almost half of all children AC C E S S T O VA R IOU S G O OD HOU S I NG C ON DI T ION S . in Papua live in houses with no electricity from PLN, Around 60 percent of children are poor or vulnerable, 64 unclean drinking water, and have poor sanitation percent lack proper sanitation, 56 percent lack clean water, combined. Such conditions are far from optimal and and 62 percent have no electricity from PLN, the state- mean that these children will lag behind other children owned power utility. With such high poverty rates on each who have better housing conditions. dimension, this means that many Papuan children are poor AN UNFAIR START 33 WHICH INEQUALITIES OF NOV EM BER 2015 OPPORTUNITY MATTER MOST IN INDONESIA? B O X .1 C O N T. Percentage of poor housing conditions in Papua (fig.35) a) The relationship between poverty, unsafe drinking water, b) The relationship between the availability of PLN, and poor sanitation unsafe drinking water, and poor sanitation 60 % Poorest 40 62 % No PLN 56 % 3 % 4% 3% Unclean 7 % Drinking 3% 56 Water % Unclean 4% Drinking 11 % 46% 39 % Water 64 % 10 % 11% 64 Poor % 4% Poor Sanitation 6 % Sanitation 5% SOURCE Susenas and Podes, from Hadiwidjaja, Paladines and Wai-Poi (2013). NOTE Poorest 40 means the child lives in a household that is in the poorest 40 percent of households nationally. AN UNFAIR START 34 NOV EM BER 2015 4. WHAT DOES INDONESIA NEED TO DO? T H E G OV E R N M E N T C A N R E DUC E I N E QUA L I T Y OF the availability of doctors in health centers and qualified OPP ORT U N I T Y B Y S T R E NGT H E N I NG E X I S T I NG teachers in schools can be addressed. If the quality of S O C I A L A S S I S TA NC E PR O GR A M S A I M E D AT T H E services remains unequal, expanded access will only P O OR A N D V U L N E R A BL E . Addressing inequality of have limited e ects on inequality. opportunity begins with improving the access of poorer households to quality health and education services. T H E G OV E R N M E N T W I L L A L S O N E E D T O I M PR OV E This can start by expanding existing social assistance L O C A L S E RV IC E DE L I V E RY B Y I N V E S T I NG I N programs such as the “Family Hope Program” (Program T R A N S P ORTAT ION I N F R A S T RUC T U R E A S T H I S Keluarga Harapan, or PKH) conditional cash transfer, S U PP ORT S P OL IC I E S T O A DDR E S S I N E QUA L I T Y the “Indonesia smart card” (Kartu Indonesia Pintar, or I N A L L O T H E R A R E A S . Investing in transportation KIP) education subsidy for the poor, and the national infrastructure in the form of good roads, bridges, and health insurance program aimed at the poor (Jaminan transportation will improve access to health clinics Kesehatan Masyarakat, or Jamkesmas). and schools. Increased connectivity for remote areas and reduced logistics costs in general will also help I N T E RV E N T ION S I N T H E DE L I V E RY OF H E A LT H to reduce high and volatile rice prices and other food A N D E DUC AT ION S E RV IC E S N E E D T O F O C US prices, which disproportionally a ect the poor. It is N O T ON LY ON AC C E S S I BI L I T Y BU T A L S O ON T H E estimated that Indonesia has lost more than QUA L I T Y OF S E RV IC E S A N D FAC I L I T I E S . While 1 percentage point of additional GDP growth due increasing numbers of people are able to access health to underinvestment in infrastructure, primarily in the and education services, improvements in outcomes transportation sector (World Bank, 2014). Therefore, are constrained by the poor quality of these services. investment in infrastructure will not only reduce Interventions should begin with the most basic elements, inequality of opportunity, but also encourage such as the availability of water and electricity in health economic growth as such investment increases centers and schools. Then, once progress has been both connectivity and productivity. achieved here, the more complex challenges such as AN UNFAIR START 35 NOV EM BER 2015 References Zhuang, J., R. Kanbur, and C. Rhee. 2014. Rising Inequality in Asia and Policy Implications. ADBI Working Paper 463. Tokyo: Asian Development Bank Institute. Alkire, Sabina and James Foster. 2011. “Counting and multidimensional poverty measurement.” Journal of Public Economics 95 (7): 476–487. Arneson, R. 1989. “Equality and equal opportunity for welfare.” Philosophical Studies 56 (1): 77–93. Arneson, R. 2015. “Equality of Opportunity.” http://plato.stanford.edu/entries/equal-opportunity (accessed October 7, 2015). Banerjee, A.V., and E. Duflo. 2003. “Inequality and Growth: What Can the Data Say?” Journal of Economic Growth 8: 267–299. Barros, R.P.D, F.H.G. Ferreira, J.R.M. Vega, J. S. Chanduvi. 2009. Measuring Inequality of Opportunities in Latin America and the Caribbean. World Bank. Washington, DC. Chang, Mae Chu; Dunkelberg, Erika; Iskandar, Susiana; Naudeau, Sophie; Chen, Dandan; Cibulskis, Richard; Namasivayam, Puspavali; Josodipoero, Ratna; Soekatri, Moesijanti; Warouw, Elsa. 2006. Early childhood education and development in Indonesia: an investment for a better life. Working paper series; no. 2006-2. Washington, DC: World Bank. Dabla-Norris, E., K. Kochhar, N. Suphaphiphat, F. Ricka, E. Tsounta. 2015. Causes and Consequences of Income Inequality: a global perspective. Ferreira, F.F.G., and M. A. Lugo. 2012. “Multidimensional Poverty Analysis Looking for a Middle Ground.” Policy Research Working Paper 5964, World Bank, Washington, DC. Forbes, K.J. 2000. “A Reassessment of the Relationship between Inequality and Growth.” American Economic Review 90 (4) 869–887. Hadiwidjaja, G., C. Paladines and M. Wai-Poi, M. 2013. Multidimensional Child Poverty in Indonesia. (Working Paper). Washington, DC: World Bank Hongyi, L., and Z. Heng-fu. 1998. “Income Inequality is not Harmful for Growth: Theory and Evidence.” Review of Development Economics 2 (3): 318–334. Levitsky, D.A., B.J. Strupp. 1995. “Malnutrition and the brain: undernutrition and behavioral development in children.” Journal of Nutrition 125: 2212S–2220S. Kuznets, Simon. 1955. “Economic Growth and Income Inequality.” American Economic Review 45(1) 1–28. Maasoumi, Esfandiar and Maria Ana Lugo. 2008. “The Information Basis of Multivariate Poverty Assessments,” in Quantitative Approaches to Multidimensional Poverty Measurement, edited by Kakwani and Silber. Palgrave Macmillan. Mendez, M. A., L. S. Adair. 1999. “Severity and timing of stunting in the first two years of life a ect performance on cognitive tests in late childhood.” Journal of Nutrition 129: 1555–1562. Milanovic, Branko. 2011. The Haves and the Have-Nots: A Brief and Idiosyncratic History of Global Inequality. Basic Books. Oxford Poverty and Human Development Initiative. Launch of Mexico’s New Poverty Measure. http://www.ophi.org.uk/ launch-of-mexico%E2%80%99s-new-poverty-measure (accessed October 7, 2015). AN UNFAIR START 36 NOV EM BER 2015 References Persson, T., and G. Tabellini. 1991. “Is Inequality Harmful for Growth? Theory and Evidence.” American Economic Review 48: 600–621. Peirskalla, J. H., and A. Sachs. 2015. Unpacking the E ect of Decentralization on Conflict: Lessons from Indonesia. Ravallion, Martin. 2011. “On multidimensional indices of poverty.” Journal of Economic Inequality 9 (2): 235–248. Rawls, John. 1971. “A theory of justice”, Cambridge, MA: Harvard University Press. Roemer, J. 1998. Equality of opportunity. Harvard University Press. Cambridge, Massachusetts. Sen, A. K. 1976. “Poverty: An ordinal approach to measurement.” Econometrica 44 (2): 219–231. UNICEF (United Nations Children’s Fund). 2005. Press release: Lack of safe water and sanitation in schools jeopardizes quality education. Retrieved on October 7 2015 from http://www.unicef.org/media/media_24898.html (accessed October 7, 2015). World Bank. 2013. Early Childhood Education and Development in Indonesia. World Bank, Jakarta. World Bank. 2014. Development Policy Review 2014. Indonesia: Avoiding the Trap. World Bank, Jakarta. World Bank. 2015a. Indonesia’s Rising Divide: Why inequality is increasing, why it matters and what can be done. World Bank, Jakarta. World Bank. 2015b. A Perceived Divide: How Indonesians perceive inequality and what they want done about it. World Bank, Jakarta. Zhuang, J., R. Kanbur and C. Rhee. 2014. Asia’s income inequalities. In Zhuang, J., R. Kanbur and C. Rhee, eds. Inequality in Asia and the Pacific: Trends, Drivers, and Policy Implications. New York: Asia Development Bank and Routledge. AN UNFAIR START 37 NOV EM BER 2015 Annex: Data And Methodology This paper examines inequality in Indonesia, more specifically inequality of opportunity. It utilizes three di erent analytical tools to examine inequality through di erent dimensions. 1. INEQUALITY OF OPPORTUNITY DECOMPOSITION One way to analyze inequality of opportunity is to use decomposition analysis. Decomposition analysis essentially looks at how an individual’s current per capita consumption is a ected by various factors. The sample consists of households in which the head of the household was born after 1948. This person’s household per capita consumption is used to predict their welfare. Then a range of their circumstances at birth are examined to see which aspects have led them to have the per capita consumption expenditure that they have today. 2. DASHBOARD APPROACH The second analysis is done through a dashboard approach, which essentially provides di erent indices for each dimension. The dashboard approach allows for comparison of achievements (or deprivations) across di erent groups. For instance, it analyzes whether there are any significant di erences between: access to schools for those living in urban/rural areas, di erences between those who have male/female headed households, di erences based on the education attainment of the parents, di erences across di erent parts of Indonesia, and whether there are di erences by decile measured based on consumption expenditure. The dashboard approach therefore considers the marginal distribution between achievements (or deprivations) across di erent groups. 3. ASSOCIATIONS VENN Following Ferreira and Lugo (2012), this paper also examines the correlation between achievements (or deprivations) graphically through the use of Venn diagrams. It essentially aims to answer whether individuals who score negatively in one dimension also score negatively in other dimensions. For example, the associations between physical access to education, health, and transportation services, or education, the presence of a midwife, and unskilled delivery. These correlation/associations have important implications for program design and targeting. This paper utilizes data from the Indonesia Family Life Survey (IFLS), which provides reliable information on relevant dimensions such as health status, anthropometrics, education and consumption. IFLS is a panel survey so it allows for the examination of individuals over time, which in turn allows for the analysis of individuals’ consumption expenditure or education outcome in relation their circumstances at birth. The National Socioeconomic Survey (Survei Sosial Ekonomi Nasional, or Susenas) and the village potential data (Potensi Desa, or Podes) are also used to examine the existence of public infrastructure, such as hospitals and schools. AN UNFAIR START