econ

This is an R Markdown document. Markdown is a simple formatting syntax for authoring HTML, PDF, and MS Word documents. For more details on using R Markdown see http://rmarkdown.rstudio.com.

When you click the Knit button a document will be generated that includes both content as well as the output of any embedded R code chunks within the document. You can embed an R code chunk like this:

Socio-economic Development and Inequality in Decentralizing Indonesia Draft June 13, 2014

Abstract This paper describes the Indonesian socio-economic development and inequality in international and regional perspectives. Over the 30 years, Indonesia’s rank for socio-economic indicators across countries presents positive trends except income inequality which increases after Asian crisis. Among developing countries, Indonesia is catching up in terms of per capita GDP, literacy, mortality, fertility, access to sanitation and clean water. But, its variations of those indicators across provinces in Indonesia are higher except sanitation. From the regional perspectives, human development index, literacy ratio, access to clean water and sanitation tend to converge while mortality and fertility are likely to diverge among the provinces in Indonesia. Although higher per capita income could increase human development index faster, inequality in per capita income has increased at national, islands and provincial level since 2001. By using provincial data and employing club convergence analysis, we find there are two clustering groups in term of per capita GRP, two groups for Gini index and net enrollment ratio, and four groups of human development index. We also find a strong correlation between province’s economic growth rate and the initial per capita GRP. Meanwhile, the data show a strong relationship between per capita GRP and improvement in social economic indicators. Lastly, the cross province’s gap exists caused by natural resources endowment and concentration of economic activities.

Introduction

Indonesia has been known as the most diverse and large country with 247 millions inhabitants and ranked the 4th most populous country in the world. The total area is around 1,375,369 square km, stretching from west to east with a length of 5,200 km and width of 1,870 km (CBS Indonesia 2013). This archipelago consists of 17 thousand islands with five major islands i.e. Sumatera, Java, Kalimantan, Sulawesi and Eastern islands. Each island has its own unique characteristics of language, cultures, and also stage of development. In the aftermath of the Asian financial crisis, Indonesia has changed dramatically from the centralized to the decentralized country since 2001, known as “big bang” decentralization policy. After more than one decade, the authority is now delegated to 512 districts from previously 341 in 2000, almost doubled in number of local governments (MOHA 2013). Following that process, democratization is introduced in 2004 in which the president as well as the local heads are directly elected throughout Indonesia. Recently, Indonesia’s economy grows at 5.8 percent while GDP per capita measured by PPP is about USD5,000 (WDI 2013). Indonesia has played key role in Southeast Asian, maintained positive growth after global crisis, but showing signs of macroeconomic weaknesses during the last few years (ADB 2013). Above all facts, Indonesia provides an interesting research especially in the area of economic growth and development. The country has experienced high economic growth in the last four decades especially triggered by the pro-growth policy in the President Suharto’s era. However, critics from many scholars concerning the inequality have also escalated with pointing out that high growth does not necessarily mean lowering inequality in developing countries (Kanbur, 1998) although it may reduce poverty (Dollar and Kray, 2000). The similar argument can be found in Yusuf, Summer and Rum (2013) that rapid economic growth in Indonesia during the 1980s and 1990s is accompanied by large reduction in poverty incidence, however its impact on inequality is rather unclear. In this paper we analyze the socio-economic development and inequality in decentralizing Indonesia. We describe the overview of economic development of Indonesia from both international perspective and regional dimension. We provide the descriptive statistics in comparing some indicators with relevant countries in this study. The group of countries used in our analyses are ASEAN countries (Malaysia, Philippine, Thailand), Emerging countries (Brazil, India, China, South Korea), African countries (Kenya, Nigeria, South Africa), and Rich countries (Australia, Euro Area, Japan, Singapore, United States. On regional level we use the data of per capita Gross Regional Product (GRP) and other social indicators of 33 provinces. Besides providing descriptive statistics, we employ the sigma-convergence analysis to see the differences in per capita income and other relevant indicators. We borrow the most recent technique to identify the convergence clustering developed by Phillips and Sul (2007). In addition to sigma-convergence analysis, we also run the regression to identify the existence of beta-convergence of per capita GRP. We find that there is a wide income disparity between western and eastern Indonesia as well as an increasing inequality across provinces. There is also a widening gap in urban-rural, Java-outer Java in terms of income share after Asian financial crisis. Income share for the rich is increasing while for the poor is decreasing (Sakamoto, 2007; Yusuf and Resosudarmo, 2008; Mishar, 2009; Tadjoeddin, 2013; Chongvilaivan, 2013). Meanwhile, with longer period of observation Mishar (2009) found that income distribution was relatively constant during 1963-2005 with low fluctuations between 1.87 to 2.5. Nonetheless, Yusuf et al. (2013) find that a gap among income group in Java and urban areas measured by decile dispersion ratio is worse than that of its Gini index especially for urban areas and Java island. Also, inter-district inequality is higher than inter-province but there is no tendency to increase inter-regional disparity in Indonesia. From the convergence test of provinces’ per capita GRP, we find that there are two clubs i.e. the first club consists of relatively rich provinces and the rests are the member of second club. We also find that the existence of β-convergence in per capita GRP. Moreover, the variation of some socio-economic indicators across provinces tend to decrease. Human Development Index and Gini index are classified into four and two clubs of convergence, respectively.

Indonesia in International Perspectives

To see the relative position of Indonesia among other countries, we rank key socio-economic indicators such as per capita GDP, mortality, population growth, human development, access to clean water and sanitation from around 200 countries, and Gini index from 70 transition and developing countries. Figure 2.1 presents the percentage of rank of Indonesia where the lower rank means the better position of Indonesia relatively among other countries. Over the last 30 years, the rank of Indonesia is slightly higher except Gini index. Population growth is in the middle position with the percentage of rank around 50 percent, followed by mortality and human development index around 65-70 percent for the whole period. After Asian crisis, rank of per capita GDP is fairly stable as well as access to improved sanitation and clean water. Rank of income inequality in Indonesia as presented by Gini index, on the other hand, jumped from around 17 percent in 1999 to almost 43 percent in 2010. Figure 2.1: Indonesian Rank for Key Socio-economic Indicators, 1981-2011

Note: Percentage of rank is calculated based on number of countries with complete data. Data availability varies across indicators i.e. GDP per capita (183), Infant Mortality Rate-IMR (194), Population growth (212), Access to Improved Sanitation (198), Access to clean water (199), Human Development Index-HDI (180), Gini Index (70). Number of countries in sample is in bracket. Rank for Gini index is calculated among 70 transition and developing countries from 1990 to 2010. Source: WDI, UNDP, CBS Indonesia. However, in terms of absolute value of GINI index, Indonesian’s figure is much lower and fairly stable compared to ASEAN countries, emerging countries and African countries as depicted in Figure 2.2. It fluctuated between 33 and 38 for the past 20 years. The only similar pattern of Gini index Indonesia is shown by India. If we compare to ASEAN countries where the Gini index was ranging from 40 to 50 or to African countries with the range of 40 to 70, or even to Brazil with the index above 50 during the same period, the Indonesia’s figure seems uncommon. In respond to these findings, many scholars suggest that the low income inequality in Indonesia for long time was due to government policy i.e. subsidy in agriculture, infrastructure, education and health sectors in early 1970s (Booth and Sundrum, 1981; Booth, 1992).
On the other hand, Gini index in Indonesia is calculated based on household expenditure which leads to lower number compared to that of its income . Frankema and Marks (2009, 2010) propose alternative measurements in income inequality in Indonesia i.e. ratio of unskilled wages to GDP per worker, Theil coefficient of the inter-industry wage distribution in the manufacturing sector, and development of the relative size of the urban informal sector. Their results suggest that those major determinants of income inequality in Indonesia i.e. wage inequality and share of the self-employed in labor force appear similar to Brazil and Mexico which are known as the highest levels of income inequality in the world. Thus, a claim of income inequality in Indonesia were rather low by international standard, is not based on entirely convincing evidence (Zanden and Marks, 2012). Figure 2.2: Gini Index Indonesia and Selected Countries, 1990-2010

Note: GINI index Indonesia at national level and GINI index Java island are calculated by CBS based on household survey (SUSENAS) Source: WDI, UNDP, OECD, CIA, CBS Indonesia

Following above discussion, levels per capita GDP Indonesia for the last 32 years are increase, and its figures are higher than African countries but lower than Thailand, Malaysia, Brazil, South Korea, and China (Table 2.1). It appears that Indonesia is able to start catching up the richer countries within ASEAN and Emerging countries (Appendix A1). Hence, we may conclude that an increasing per capita GDP in Indonesia is not necessarily associated with an increasing Gini index. Further analysis on Indonesia’s Gini index will be discussed on regional dimension and convergence analysis section.

Table 2.1: Socio-economic Indicators Indonesia and Selected Countries 1980-2012 (1) Country Name GDP per Capita, PPP USD 2005 Human Development Index Poverty HCR at $1.25 a day (PPP) % of Population Literacy Rate, adult per total % of people age 15+ 1980 2012 annual growth (%) 1980 2012 annual growth (%) 1980 2012 annual growth (%) 1980 2010 % of change ASEAN COUNTRIES
Indonesia 1,371 4,272 3.6 0.422 0.629 0.8 62.8 16.2 -46.6 67.3 92.8 25.5 Java Island 734 3.548 5.0 0.678 0.725 0.2 .. .. .. 89.2* 92.6 3.4 Malaysia 5,063 14,822 3.4 0.563 0.769 0.6 3.2 0.5 -2.7 69.5 93.1 23.6 Philippines 2,807 3,801 1.0 0.561 0.654 0.3 34.9 18.4 -16.5 83.3 95.4 12.1 Thailand 2,226 8,463 4.3 0.490 0.690 0.7 21.9 0.4 -21.5 88.0 93.5 5.5 EMERGING COUNTRIES
Brazil 7,565 10,264 1.0 0.522 0.730 0.6 13.6 6.1 -7.5 74.6 90.4 15.8 China 524 7,958 8.9 0.407 0.699 1.0 84.0 11.8 -72.2 65.5 95.1 29.6 India 881 3,341 4.3 0.345 0.554 0.9 65.9 32.7 -33.2 40.8 62.8 22.0 South Korean 5,544 27,991 5.2 0.640 0.909 0.7 .. .. .. .. .. .. East Asia & Pacific 2,570 9,590 4.2 .. 0.683 .. .. .. .. 82.0 94.7 12.7 AFRICAN COUNTRIES
Kenya 1,375 1,522 0.3 0.424 0.519 0.4 38.4 43.4 5.0 82.2 72.2 -10.1 Nigeria 1,832 2,335 0.8 0.434 0.471 1.2 53.9* 68.0 14.1 55.5 51.0 -4.5 South Africa 8,763 9,860 0.4 0.570 0.629 0.2 24.3* 13.8 -10.5 82.4 93.0 10.6 DEVELOPED COUNTRIES
Australia 19,787 35,669 1.9 0.857 0.938 0.2 .. .. .. .. .. .. Japan 17,835 31,425 1.8 0.788 0.912 0.3 .. .. .. .. .. .. Singapore 15,116 53,266 4.0 0.756 0.895 0.8 .. .. .. 82.9 95.9 13.0 United States 26,086 45,336 1.7 0.843 0.937 0.2 .. .. .. .. .. .. Euro Area 18,982 29,859 1.4 .. .. .. .. .. .. 98.9 .. Note: * The earliest and the most recent data differ for some countries: Literacy Rate: the earliest data for Java island started in 2003 Poverty HCR: the earliest data for Kenya is 1992, South Africa 1993

In terms of socio-economic indicators, Indonesia has obtained higher improvements, also looks catch up other developing countries (Appendix A2). For instance, poverty alleviation, human development index, literacy rate, life expectancy, fertility, mortality are higher than the figures of India and African countries, but lower performance compared to Brazil and any other East Asian countries (see Table 2.1 and Table 2.2). Based on UNDP Report 2012, Indonesia’s HDI is 0.629, lower than East Asia and Pacific (0.683) and the world value (0.694). In addition to that, the rank of Indonesia in 2012 is 121 out of 194, below the rank of its neighboring countries such as Philippine (114), Thailand (103), Malaysia (64) and Singapore (18). However, if we look at Java island only, it outclasses with the HDI value of 0.725, much higher than most of developing countries. Java island apparently outpaces other area in Indonesia. If we delve into poverty alleviation in Indonesia, it recorded a substantial decrease in headcount ratio at $1.25 a day from 62.8 percent in 1981 to 16.2 percent in 2011 while the corresponding figures for China and India dropped from 84 percent to 11.8 percent and 65.9 percent to 32.7 percent respectively (Table 2.1). This strong performance in poverty reduction in Indonesia is due to various deregulation policies designed to promote non-oil exports, develop the financial sector, and give the private sector a larger role in the economy (Booth, 1992). Table 2.2: Socio-economic Indicators Indonesia and Selected Countries 1960-2012 (2) Country Name Life Expectancy Total fertility rates Infant mortality Sanitation Facilities Water Availability % of pop with access Years Children Per ’000 % of pop with access
1960 2011 annual growth (%) 1960 2011 annual growth (%) 1960 2012 annual growth (%) 1990 2012 % of change 1990 2012 % of change ASEAN COUNTRIES
Indonesia 44.8 69.3 0.9 5.7 2.6 -1.5 148.9 25.8 -3.3 35.3 58.7 23.4 69.8 84.3 14.5 Java Island 46.7 69.8 1.0 5.4 2.2 -2.2 140.2 23.3 -4.3 48.9* 58.6* 9.7 89.0* 93.2* 4.2 Malaysia 59.5 74.7 0.4 6.2 2.0 -2.2 67.0 7.3 -4.2 84.4 95.7 11.3 88.2 99.6 11.4 Philippines 57.8 68.4 0.3 7.1 3.1 -1.6 66.2 24.1 -1.9 56.8 74.2 17.4 84.8 92.4 7.6 Thailand 55.2 74.0 0.6 6.1 1.4 -2.8 102.0 11.8 -4.1 81.7 93.4 11.7 86.4 95.8 9.4 EMERGING COUNTRIES
Brazil 54.7 73.3 0.6 6.2 1.8 -2.3 129.7 12.9 -4.3 66.8 80.8 14.0 88.5 97.2 8.7 China 43.5 75.0 1.1 5.8 1.7 -2.4 82.7* 12.1 -4.4 23.7 65.1 41.4 66.7 91.7 25.0 India 41.4 66.0 0.9 5.9 2.5 -1.6 164.2 43.8 -2.5 17.7 35.1 17.4 70.3 91.6 21.3 South Korea 53.0 81.9 0.9 6.2 1.2 -3.1 79.0 3.4 -5.9 100.0 100.0 0.0 89.6 97.8 8.2 East Asia & Pacific 48.0 74.5 0.9 5.4 1.8 -2.1 87.5* 16.3 -3.8 37.4 69.8 32.4 70.9 91.5 20.6 AFRICAN COUNTRIES
Kenya 46.4 60.4 0.5 7.9 4.5 -1.1 117.0 48.7 -1.7 24.6 29.4 4.8 42.7 60.9 18.2 Nigeria 37.2 51.7 0.6 6.4 6.0 -0.1 182.6 77.8 -1.7 30.1 30.6 0.5 47.2 61.1 13.9 South Africa 49.0 55.3 0.2 6.2 2.4 -1.8 87.1* 33.3 -2.5 63.9 74.0 10.1 82.6 91.5 8.9 DEVELOPED COUNTRIES
Australia 70.8 81.8 0.3 3.5 1.9 -1.2 20.3 4.1 -3.0 100.0 100.0 0.0 100.0 100.0 0.0 Japan 67.7 82.6 0.4 2.0 1.4 -0.7 30.4 2.2 -4.9 100.0 100.0 0.0 100.0 100.0 0.0 Singapore 65.7 80.9 0.4 5.5 1.2 -2.9 35.5 2.2 -5.2 99.2 100.0 0.8 100.0 100.0 0.0 United States 69.8 78.6 0.2 3.7 1.9 -1.3 25.9 6.2 -2.7 99.5 99.6 0.1 98.4 98.8 0.4 Euro area 69.3 81.3 0.3 2.6 1.6 -1.0 38.8 3.5 -4.5 99.6 99.9 0.3 99.7 100.0 0.2 Note: * The earliest and the most recent data differ for some countries: IMR: the earliest data for China is 1969, East Asia & Pacific 1968 and South Africa 1974 Sanitation and Water: the earliest data for Java Island is 2007 and the most recent data is 2010 Demographic indicators for Java island started in 1971
Source: WDI, CBS Indonesia, Ministry of Health Indonesia (Riskesdas)

On the other hand, education in terms of literacy rate Indonesia is lower than China, Malaysia and even among East Asia and Pacific country group. Yet, Table 2.1. shows Indonesia has been catching up. Moreover, the positive changes in secondary school enrollment (net) from 1960 to 2011 and tertiary school enrollment (gross) from 1970 to 2011 indicated sound progress in education outcomes Indonesia which accounted for 58 percent and 22 percent respectively (see Appendix A3). Among other developing countries in Asian, South Korea, Thailand and Malaysia demonstrated the substantial improvement in both secondary and tertiary education. Furthermore, Indonesia, together with Malaysia, Thailand, China, Brazil and South Korea have decreased in fertility and mortality rate for the last 50 years but other countries did faster than Indonesia. The same pattern is also appeared from sanitation and water figures in which both Indonesia and China have increased in percentage of access to improved sanitation over the 25 years but China did faster. However, access to clean water and sanitation in Indonesia is generally very low compared to many other countries except Kenya and Nigeria. For instance, only 59 percent of Indonesian population with access to sanitation facilities and 84 percent with access to clean water (Table 2.2). Despite those socio-economic performance, there is a gap across provinces in Indonesia which measured by coefficient of variation for each indicator. For instance, if we compare regional income inequality, as measured by coefficient of variation of per capita GRP, Indonesia’s figure is the highest but decreasing trend among China, Thailand, and Philippines. Similar pattern also shown in literacy rate which is the highest trend among China and Vietnam, but it tends to converge (Figure 2.3). These patterns indicate that the income inequality as well as literacy rate in Indonesia varies across provinces but tends to converge. Figure 2.3: Variation of Province Socio-economic Indicators in Indonesia and Selected Countries

Source: CBS Indonesia, CBS China, CBS Vietnam, CBS Thailand, CBS Philippines In addition to that, the variations across provinces for fertility and mortality in Vietnam, and the variations in sanitation across provinces in China are higher than those are in Indonesia. In other words, disparities in fertility and mortality across provinces in Indonesia are lower than those are in Vietnam, and inequality in access to improved sanitation is lower than it is in China. To sum up socio-economic development, Indonesia is catching up in terms of per capita GDP, literacy, mortality, fertility, access to sanitation and clean water (Appendix A2). Meanwhile, most of variations across provinces for those indicators in Indonesia are higher except variations in sanitation. These patterns indicate that the disparities in socio-economic indicators Indonesia vary across provinces.

Regional Dimension

If we look at the cumulative distribution of provincial GRP measured by 2000 Rupiah which is shown by Lorenz cure, there is an improvement in equality since 1969 to 2012. But, the improvement seems very slow and the inequality still prevails. Our estimation of Gini index in 2012 by using provincial GRP is about 0.63 (Figure 3.1). The dot-and-dash line in the diagonal represents the imaginary of equality line, which shows the perfect cumulative distribution of income among the cumulative number of provinces. The dash line is the actual cumulative distribution of income in 1969, while the solid line in 2012. The distances between these lines with the diagonal line measure the inequality of the income distribution in 1969 and 2012. The distance of 2012’s line is narrower than the 1969’s means that provincial GRP distribution in 1969 is less equal to the 2012. The dynamics of Gini index will be explored further in this part and also in part 4 when we deal with convergence analysis. Figure 3.1: Cumulative distribution of provincial GRP, 1969 & 2012

                                   Source: Authors’ calculation from CBS Indonesia

3.1 Socio-economic Indicator We describe socio-economic performance from various indicators such as per capita GRP, human development index (HDI), literacy, mortality, fertility, access to drinking water and improved sanitation. The relationship between per capita GRP and HDI of 33 provinces in Indonesia, and the variations they have in 1996 and 2001 are presented in Figure 3.2. The positive trend indicates per capita GRP could increase HDI. If we compare the figures of 2011 are higher than those of 1996. However, the low HDI is not necessarily associated with the low per capita GRP. For instance, the lowest HDI in 1996 was West Nusa Tenggara while the lowest per capita GRP was East Nusa Tenggara. Some provinces could move to higher HDI or per capita GRP like West Nusa Tenggara and East Nusa Tenggara did. The former province was not the lowest HDI in 2011 but Papua, while the latter was also no longer the lowest per capita GRP in 2011 but Maluku. All provinces which experienced low HDI as well as per capita GRP were located in eastern part of Indonesia. Meanwhile, Jakarta performed the highest figure for those indicators. This indicates that there are high disparities across Indonesia not only in terms of per capita income but also human development. Detail figure of socio-economic indicators for each province can be seen in A4-A6. Figure 3.2: Relation Per Capita GRP and Human Development Index, 1996 and 2011

Note for legends: ENT (East Nusa Tenggara), WNT (West Nusa Tenggara), MAL (Maluku), PAP (Papua), WPA (West Papua), EKL (East Kalimantan), RIU (Riau), RIS (Riau Island), JKT (Jakarta) Source: CBS Indonesia Despite the gap between western and eastern part of Indonesia, the variations of socio-economic outcome indicators across provinces tend to decrease overtime except fertility and mortality (Figure 3.3). The noticeable figures are variation in per capita GRP falling from 0.45 to 0.14 while very small variation in population over the last 43 years. Akita and Lukman (1995) in applying Williamson index also suggest large decrease in inequality of per capita GRP among provinces in Indonesia during 1975-1992. However, inequality among non-mining per capita GRP remained stagnant during that period except in the mid of 1980 when export oriented reform was introduced. Akita, et al. (2011) show that the differences in inequality in terms of per capita GRP among Indonesia’s largest regions (Java-Bali, Sumatera-Kalimantan-Papua, and other regions in eastern area) are small compared with the levels of inequality within those regions, and that the levels of cross-regional inequality have been relatively constant throughout the years (1983-2004). They also find that an increasing level of inequality is not only occurred within regions but also among districts within provinces in those regions. Some studies confirm that regional inequality is relatively stable or slightly increase at district level during 1993-1998 (Tadjoedin, et al., 2001; Akita and Alisyahbana, 2002; and Hill, et al., 2008). In the recent study, Vidyattama (2013) finds inequality of GRP per capita increase slightly at both the province and the district levels in the period of 1999-2009, especially because of the growth of Jakarta during 2002-2008. Meanwhile, inequality in human development index decreases at both province and district level during the same period. He also suggests the rise of the less developed districts in Java and Sumatera to catch up is the main reason of declining inequality in human development. Figure 3.3. describes a decreasing trend on variations in human development index as well as per capita GRP across provinces, or in other words, inequality of human development index and per capita GRP among provinces is declining over time from 1996 to 2012. Figure 3.3: Variations of Province Socio-economic Indicators 1969-2012

Note for legends: pop (population), hdi (human development index), fert (fertility), lit15 (literacy rate for age15+), clnwtr (access to clean water), grpcap (gross regional product per capita), imr (infant mortality rate), sanitn (access to improved sanitation). The variations are calculated based on coefficient of variation (CV) of Province indicators i.e. (y), where CV=√(1/n ∑(i=1)^n▒((y(i-) y ̅)/y ̅ )^2 ) , y ̅≡1/n ∑_(i=1)^n▒y_i Source: CBS Indonesia, reproduced

3.2. Gini Index

As discussed in part 1, Figure 3.4 shows that income inequality in Indonesia which measured by GINI index fluctuates from time to time with a declining tendency until early 1990s then increasing until 2012. The similar pattern is also shown in GINI index by islands as well as by provinces (Figure 3.5 and Figure 3.7). There are several events which may cause this fluctuation. First, in 1970s Indonesia benefited from windfall profit of oil boom which increased GINI index during that period. The next incident is Asian Financial Crisis in 1998, which was followed by implementing decentralization in 2001. This urged GINI index lower and started increasing after crisis recovery in 2002. The last event is introducing democratization where national and local head as well as parliament members are elected directly by the voters. Besides those episodes, global crises in 2008 may affect the GINI index. An increasing trend of GINI index could be perceived in two folds i.e. as positive sign of the effectiveness of market mechanism and negative sign of widening income gap across Indonesia. In this paper we try to describe the possibility reasons GINI index tends to increase overtime after 2001, not to investigate the causality. Further research should be conducted to examine the pattern of GINI index.   Figure 3.4: GINI Index Indonesia 1964-2013

Note: GINI index at national level is calculated by CBS Indonesia based on household survey (SUSENAS) Source: CBS Indonesia

If we look at the percentage of the very rich tax payer in Indonesia, measured as 0.05-0.01 percent top of income level, it grows from 0.2% to 0.8% or a rise by 55% within 7 years (1999-2006). This increasing trend is in line with an increasing in GINI index which indicates that income disparity become higher in Indonesia (Figure 3.4a).
Figure 3.4a: GINI Index and Top 0.05-0.01% Income Share

Note: GINI index at national level is calculated by CBS Indonesia based on household survey (SUSENAS), Source: CBS Indonesia Top 0.05-0.01% income share is calculated based on tax data. Source: Alvaredo, Facundo, Anthony B. Atkinson, Thomas Piketty and Emmanuel Saez, The World Top Incomes Database, http://topincomes.g-mond.parisschoolofeconomics.eu/ , 09/06/2014y.

Since 1980s is known as period of high growth, Boediono (1990) concludes that high growth economy is slightly associated with declining GINI. However, Dick et.al (2002) argue that people perceived economic disparities were widening due to excessive self-enrichment of Suharto’s crony. This opposite views could be occurred because BPS’s data use household consumption instead of individual income as we discussed previously . Figure 3.5: GiniI Index Indonesia by Island, 1976-2013 Figure 3.6: Variation Gini Index Indonesia across Provinces per Island, 1984-2013

Note: Gini index island is the average of provincial Gini index which is calculated by CBS Indonesia based on household survey (SUSENAS) Source: CBS Indonesia

For further discussion on income inequality, we look at the variation of Gini Index across islands which becomes smaller and tends to converge (Figure 3.6). Yet, Eastern and Java islands show their variations increase after 2010. The variation among islands could be generally divided into three groups. First, an early 1980s until 1996 (pre-crisis) where the variations are relatively stable and tend to converge. Second, 1997-2007, a period of Asian Financial Crisis followed by the fall of Suharto when the variations tend to diverge (Mishra, 2009). Third, recent period (2008-2013) in which the variations tend to converge. The sigma convergence within islands looks like an U-shape for some islands (Sulawesi, Sumatera, Kalimantan) but not true in general i.e. for Eastern island. Figure 3.7: Gini Index Indonesia by Provinces, 1976-2013

Note: GINI index province is calculated by CBS Indonesia based on household survey (SUSENAS) Legends: ACH (Aceh), NSA (North Sumatera), WSA (West Sumatera), RIU (Riau), JBI (Jambi), SSA (South Sumatera), BKL (Bengkulu), LMP (Lampung), BBL (Bangka Belitung), RIS (Riau Islands), JKT (Jakarta), WJA (West Java), CJA (Central Java), YOG (Yogyakarta), EJA (East Java), BTN (Banten), BLI (Bali), WNT (West Nusa Tenggara), ENT (East Nusa Tenggara), WKL (West Kalimantan), CKL (Central Kalimantan), SKL (South Kalimantan), EKL (East Kalimantan), NSI (North Sulawesi), CSI (Central Sulawesi), SSI (South Sulawesi), TSI (Southeast Sulawesi), GOR (Gorontalo), WSI (West Sulawesi), MAL (Maluku), NML (North Maluku), WPA (West Papua), PAP (Papua) Source: CBS Indonesia
If we look at Gini index across provinces, there is an increasing trend in all provinces as well as islands. The highest Gini Index at provincial level in Sumatera, Java, Kalimantan, Sulawesi and Eastern island is Bengkulu (0.386), Yogyakarta (0.439), West Kalimantan (0.396), Gorontalo (0.437), and Papua (0.442) respectively. From Figure 3.7 income inequality looks faster in some provinces i.e. provinces at Eastern islands but relatively slower in provinces at Kalimantan island. Furthermore, the variation across provinces tends to decrease while the variation across islands tend to increase (Figure 3.8). Some government policies in 2003-2006 may affect a widening gap across islands while this is not so across provinces. For instance, in the period of 2003-2005, the government increased the rice price by 20 percent after very long time implementing rice price stabilization policies. This policy may reduce the real expenditure of the poor. The government also increased oil subsidy by 10 percent to respond an increasing of world oil price by 70 percent. The latter policy could be less favorable to the poor (Yusuf and Rum, 2013). Following an increasing retail fuel price particularly kerosene price, the government issued compensation policy to the poor namely conditional cash transfer in 2005-2006. This policy may lead to a slightly decreasing Gini Index in 2007 (Yusuf and Resosudarmo, 2008).
Figure 3.8: Variation of Gini Index Indonesia, 1984-2013

Note: the variations are calculated based on coefficient of variation (CV) of province and island Gini index (y), where CV=√(1/n ∑(i=1)^n▒((y(i-) y ̅)/y ̅ )^2 ) , y ̅≡1/n ∑_(i=1)^n▒y_i GINI index island is the average of provincial GINI index which is calculated by CBS Indonesia based on household survey (SUSENAS) Source: CBS Indonesia

Figure 3.8 also shows that the response of those policies to the individual province is less variation. But, when the provinces are grouped into islands, their responses vary. It seems that there is an island effect – the outstanding position of Java island which may come from the role of Jakarta. Akita et al. (2011) in their study of regional income inequality suggest that since the share of mining has decreased, the spatial distribution of manufacturing has played a more important role in the inequality of Sumatra and Kalimantan, while the primacy of Jakarta, with strong urbanization economies, facilitated by globalization and trade and financial liberalization, has determined much of the Java–Bali region’s inequality and, therefore, overall inequality in Indonesia. In addition to that, Gini Index at urban areas is much higher than rural areas while most of urban areas are located in Java island which could affect the whole variation across islands (Sakamoto, 2007; Mishra, 2009; Tadjoeddin, 2013; Chongvilaivan, 2013). Relationship of Gini Index and GDP (Kuznets’ Curve) Indonesia shows only some fraction of Kuznets curve where inequality increases when high levels of growth bring an economy to a certain level of income and then decreases with further increases in income. As we can see from Figure 3.9, Indonesia experienced relatively low and stable Gini index during the three decades with some fluctuations due to oil boom in the late 1970s. The peculiar trend in early 1960s until 1970s might be due to the new government of Suharto suppressed the economy to gain his legitimation among Indonesian people by providing subsidy to basic sectors. Hence, we could say that an increasing trend which follows Kuznets hypothesis started in early 1980s where Indonesia started to liberalize its economy i.e. more free market, higher role private sector, deregulation, etc. (Booth, 1992). Moreover, Tadjoedin (2013) suggests inequality among districts could follow Kuznets only if cities are integrated within their surrounding regencies. Figure 3.9: Relationship between Gini Index and GDP Indonesia 1964-2012

Note: Gini index at national level is calculated by CBS Indonesia based on household survey (SUSENAS) Source: CBS Indonesia

  Convergence Analysis The dispersions of Indonesia’s per capita GRP and some selected social indicators are declining for the last two decades with its neighbouring countries, especially with ASEAN and Asian countries. As discussed in part 2, the social indicators converge towards improvement. But, on the other hand, if we look at the regional dimension within Indonesia the Gini indices consistently increased in the last two decades. Moreover, the Indonesian Gini index is increasing starting from the beginning of year 2000. Analysing the behaviour of regional indicators then becomes interesting as we look at the convergence in provincial level of some indicators, especially the economic indicator. The study of economic convergence has been the heart of economic growth study after the neoclassical model of Ramsey (1928) and Solow (1956). Later, Sala-i-Martin (1990) distinguished the terminology of β(beta)-convergence and σ(sigma)-convergence. In this regards, β-convergence measures the “catch-up” effect of poor countries (regions) to the rich countries (regions), while σ-convergence measures the cross-sectional dispersion of per capita income of countries (regions) across the time. The convergence analysis has gained its popularity since the seminal papers by Barro (1991), and Barro and Sala-i-Martin (1992). Many literatures have also discussed the relationship between beta and sigma convergence, in which beta-convergence is a necessary condition for the existence of sigma-convergence. Later Furceri (2005) proves mathematically the relation between those two. Adding to the analysis of beta-convergence, authors have made distinction between absolute and conditional convergence to overcome the condition for steady state in per capita income. For example, see Sala-i-Martin (1995) and Islam (1995, 2003). On the Indonesian context, Garcia and Soelistianingsih (1997) was the first paper to test the existence of convergence by analysing the β-convergence methodology for both absolute and conditional convergence. Later, Resosudarmo and Vidyattama (2006) and Vidyattama (2013) use panel data of Indonesian provinces and districts level to analyse the disparity of provincial and district income in Indonesia. While Hill et.al (2007) provide analysis of Indonesia’s geographical changes in development. In the latest development of convergence analysis, Phillips and Sul (2007) (PS) have developed a methodology that allowing for individual heterogeneities. The methodology is based on clustering algorithm to identify if the non-convergent sample has convergence club. With this type of analysis we are able to identify the cross-sectional relative position of each individual in the sample across the time. In this section we apply the methodology developed by PS using Indonesia’s provincial data. Indonesia provides an interesting study of regional economic growth since it has been known as an archipelago with more than 17 thousands islands lie from the west to the east, with five main islands: Sumatra, Java, Kalimantan, Sulawesi, and Papua. Along with the island of Bali, Maluku, and Nusa Tenggara, the country comprises of 33 provinces (see table 4.1). Each Island has its own unique characteristics of languages, cultures, and also the development issues. In the term of regional development , Indonesia has been divided into two greater regions i.e. Western regions (Sumatra, Java, Kalimantan, and Bali) and Eastern regions (Sulawesi, Maluku, Papua, West Nusa Tenggara, and East Nusa Tenggara). Many have argued that there is a huge gap in development among the regions in Indonesia that the western regions are more developed than the eastern regions. This may be because the past development policy had only focused in Java. Whether the huge gap of development in the past persists until now and whether some poor regions are able to catch up the richer regions, the analysis of the income distribution need to be carried out further to answer these questions. The analysis of the role of geographical location is also important in this matter. Table 4.1: Provinces by Group of Main Islands Sumatra (10) Java (6) Kalimantan (4) Sulawesi (6) Nusa Tenggara and Bali (3) Papua and Maluku (4) Aceh North Sumatera West Sumatera South Sumatera Riau Jambi Bengkulu Lampung Bangka Belitung Riau Islands Jakarta Yogyakarta West Java Central Java East Java Banten West Kal. Central Kal. South Kal. East Kal. Central Sul. South Sul. Southeast Sul. Gorontalo West Sul. North Sul. Bali West NT East NT Maluku North Maluku West Papua Papua Note: Number of provinces in parentheses. Not only the eastern part of Indonesia is less developed compared to the western part, but also the inequality of provinces within western part is also higher than the provinces within eastern part. Figure 4.1 depicts the variation of provinces’ per capita GRP from 1969-2012 . This variation (measured by coefficient of variation) is the σ-convergence of the series. The figure shows per capita GRP gap has been declining since the end of 1970s for all provinces in Indonesia, as well as provinces in the western part. Meanwhile, the disparities of per capita GRP on the eastern-part show the up-side-down trend with declining trend from the end of 1970s to the mid of 1980s, but then increases afterwards with fluctuations over years until 2012. The 1998’s Asian economic crisis also worsened the equality which shown by the increasing of variation at those periods. We may argue that the development in the eastern part has remained low and not improved significantly more than four decades. The Java-centred development policy only benefited the neighbouring provinces at least until the end of 1990s, before the so called “the big-bang” decentralization policy which started in 2001.

Figure 4.1: Variation of Provinces’ Per Capita GRP

Note: the variations are calculated based on coefficient of variation (CV) of Provinces’ per capita GRP (y), where CV=√(1/n ∑(i=1)^n▒((y(i-) y ̅)/y ̅ )^2 ) , y ̅≡1/n ∑_(i=1)^n▒y_i ;

Convergence Test and Clustering Identification

We are interested to see whether some provinces share the long run common trend in term of per capita GRP. The notion that geographical location plays an important role in influencing the development of provincial per capita GRP. In this approach, grouping provinces by islands can be done. Here, we are interested to see if there exists multiple common trends of provinces’ per capita GRP. In this paper we apply the method proposed by Phillips and Sul (2007), hereinafter PS. The clustering groups are endogenously determined by using algorithm based on some statistics. This approach has been widely used in many literatures, for example in Phillips and Sul (2009), Panopoulou and Pantelidis (2009) apply the method to determine the club convergence for carbon dioxide emissions, while Herrerias (2013) extends its by looking the CO2 emissions by source of enery. Zhang and Broadstock (2014) use the method to analyse the club convergence in the energy intensity in China, and Kim (2014) analyses the energy use and economic development. More detail about this approach can be found in the paper by Phillips and Sul (2007). Before we do the cluster analysis, first we test the full panel of provinces’ per capita income of 33 provinces from 1969-2012. Define a series of provinces’ per capita GRP, Y_it, with period of t=1,…,N and number of province n=1,…,N. Based on factor model, Y_it can be composed as: Y_it=g_it+a_it (1) Where g_it represents the systematic component, and a_it the transitory component. Re-writing equation (1) with time varying factor we have: Y_it=((g_it+a_it)/μ_t ) μ_t=δ_it μ_t (2) Where δ_it is the idiosyncratic component and μ_t common component. Convergence is defined as δ_it→ δ. The convergence test based on Log t regression proposed by PS using the following equation: loga(H_1/H_t )-2logL(t)=c ̂+b ̂ loga〖t+u_t 〗 (3) The individual heterogeneity is measured by: h_it=Y_it/(□(1/N) ∑(i=1)^N▒Y_it )=δ_it/(□(1/N) ∑(i=1)^N▒δ_it ) (4)

Where h_it is the relative transition coefficient for each province with cross-sectional mean of unity. When δ_it converges to a constant, h_it converges to unity. The variance of h_it, H_t=1/N ∑_(i=1)^N▒〖〖(h〗_it-1)〗^2 converges to zero when t∞. The estimate (b ̂) from equation (3) is one-sided t-test with hetero autocorrelation consistent (HAC) standard error. The time period included in the regression equation starts from t=[rT],…,T. As suggested by PS, r is the fraction of T with r∊[0.2,0.3]. The full panel convergence exists if we cannot reject the null hypothesis of β≥0. The criterion of rejection is when t-stat < -1.65 (alpha=5%).

Identification of Convergence Clustering

If we reject the hypothesis of full panel convergence, according to PS, there might exist sub-sample convergence clustering. This notion can be tested by applying club convergence algorithm as follow: Ordering Member of panel will be ordered according to the last observation of each series. Last fraction can be used if there are substantial time series variability in the data, for example taking the last ½ or 1/3 of the last observations of the series. Forming the Core Group Maximizing the log t stat t_k=t(G_k) over 2≤kc^* (PS recommends c^=0 when t<50). The threshold c can be increased until the group has formed stable members. Lastly, if the first convergence group is formed, log t test will be applied to ensure t_b≻1.65. Stopping rule Individuals that are not in the first convergence club will be grouped into a second club and evaluated again using log t test to ensure the membership in the second club, and the steps are repeated.

Data

The data are Indonesia’s provincial gross regional product in real terms of 2000 prices from 1969-2012. Since we have different base year of nominal and real provincial GRP, the data cannot be merged directly. To deal with this issue first we re-calculate the base year and then merge the data. Meanwhile, the population data are incomplete which data only available from census year. The yearly full series of population data are interpolated from the available data provided by Indonesia’s Statistics. In addition to per capita GRP we also use some social indicators to compare the results. The data are human development index (HDI) (1996-2012), Gini Index (1976-2012), net enrolment ratio of primary and secondary schools (1996-2012), literacy ratio above 15 year old (1996-2012), and poverty rate (2002-2012). The convergence test requires balance panel but we have missing observations in our provincial data. The missing observations caused by new seven provinces that split-up in year 2000 and 2001, i.e. Bangka Belitung and Riau Islands from Riau, Banten from West Java, West Sulawesi from South Sulawesi, Gorontalo from North Sulawesi, North Maluku from Maluku, and West Papua from Papua. We apply imputation technique and process by using statistical software Stata 12. Each new province imputed from its reference province and run linear regression with reference province as independent variable. The same treatment also applies for the social indicators.

Results

We first test the full panel sample and find that convergence in per capita GRP is rejected. We simulate the various fraction (r) to check the stability of the regression. Table 4.2 shows the results of divergence of full panel sample of all b ̂ with negative signs and t-stat< -1.65. Table 4.2: Full Panel Convergence Test: Per Capita GRP r=0.2 r=0.25 r=0.3 b ̂ -1.198 -1.162 -0.947 t-statistics -2.561 -4.700 -4.474 Note: coefficients are significant at α=5%.

Figure 4.2 supports the results from full panel convergence test. The relative transition series are trend of per capita GRP after smoothed by using Hodrick-Prescott filter. For the yearly series, the parameter (λ) is set to λ=6.25 as suggested by Ravn and Uhlig (2002). The relative transition coefficient measures the relative position of each province to the cross-sectional mean, in this case the per capita GRP. We can see that at least there are two clubs of convergence with some province behave dramatically different with fluctuations and distances. For instance, Riau, East Kalimantan, Jakarta, and Riau Islands. Province of Aceh shows an very interesting evolution, increasing from 1970’s until 1990’2 and then decreasing afterwards to join another club. Meanwhile majority of the provinces move in another group of trend. Graphs in Figure 4.2 are calculated based on equation (3).

Figure 4.2: Relative Transition Coefficient (h_it) of Per Capita GRP (All Provinces)

Note: Series are Hodrick-Prescott trend of Per Capita GRP

Since we reject the hypothesis of full panel convergence, then we proceed to apply the club convergence algorithm to see if there exist the convergence clustering . Table 4.3 shows the results of club convergence test. We find that two clubs exist in provinces’ per capita GRP with the first club consists of 10 provinces and the second club consists of 23 provinces. Surprisingly there are two common trend in term of per capita GRP in Indonesia, slightly close to the notion that the country can be divided into two development regions. But unfortunately, if we look at the members of those clubs the provinces that form the clubs are mixed and there’s no strong signal of west-east dichotomy (see Figure 4.3).
Table 4.3: Convergence Club of Per Capita GRP Club Member Provinces Club 1 [10 Provinces] North Sumatera, Riau, Bangka Belitung, Riau Islands, Jakarta, East Java, Central Kalimantan, South Kalimantan, East Kalimantan, West Papua.

Club 2 [23 Provinces] Aceh, Jambi, Lampung, West Sumatera, South Sumatera, Bengkulu, West Java, Central Java, Banten, Bali, East Nusa Tenggara, North Sulawesi, Yogyakarta, West Nusa Tenggara, West Kalimantan, Central Sulawesi, South Sulawesi, Southeast Sulawesi, Gorontalo, West Sulawesi, Maluku, North Maluku, Papua.

Figure 4.4 explains the story behind the formation of those two clubs. We take the average of natural logarithm per capita GRP of both club and show the differences in their per capita GRP to estimate the common trend of both club. We can see that the second club members share lower average per capita GRP than member from the first club. The average trend of both clubs have been increasing in per capita GRP since 1969, but club 2 has smoother trend than club 1 which experienced volatility from 1970s to 1980s. However, provinces in both clubs suffered from the economic crisis that hit Asia in 1997-1998 which is shown by drop in the graphs. But, starting from 2000 both trends started to continue the increase in trend where Club 1 has higher performance than Club 2.

Figure 4.3: Club Distribution of Per Capita GRP

Note: Club 1 is the cluster of provinces with higher per capita GRP, while Club2 lower.

Figure 4.4: Estimated Common Trends of Per Capita GRP

Note: The series are average of provinces’ per capita GRP in natural log.

We compare the variation of both clubs and all provinces as shown by Figure 4.5, which is the σ-convergence for both clubs. This figure is modified from Figure 4.1 with difference in division of province members. In this figure we show the variations which calculated by their coefficient of variations. In general, between both figures, the variations are decreasing over time. In Figure 4.1, the connected scatter-plot between all provinces and western part province members almost intersected implying close variation between those two, while in Figure 4.5 the line of members of Club 1 show a significant distance from all provinces’ line. The descriptive statistics for both clubs are in table 4.4.

Table 4.4: Descriptive Statistics of per capita GRP in 2012 N Mean S.E. Min Max Club 1 10 5.07 0.19 4.53 6.08 Club 2 23 4.09 0.08 3.39 4.50 note: real per capita GRP in natural logarithm terms, measured in 2000 rupiah (million)

Figure 4.5: Variation of Provinces’ Per Capita GRP by Club

Note: the variations are calculated based on coefficient of variation (CV) of Provinces’ Per Capita GRP (y), where CV=√(1/n ∑(i=1)^n▒((y(i-) y ̅)/y ̅ )^2 ) , y ̅≡1/n ∑_(i=1)^n▒y_i .

Below we present the relative transition graph for both clubs. Figure 4.6 shows the behaviour of provinces in Club 1, while Figure 4.7 the provinces in Club 2. The benefit of transition coefficient allows us to capture the individual heterogeneities across time, even though there exists common trend. This coefficient is similar to the general σ-convergence, but it has the information on individual heterogeneity. It is interesting to see that Riau Province starts from a very high in term of per capita GRP in the late 1960s and consistently decreasing in its relative position to join another province in the beginning of 2010s. The similar pattern is East Kalimantan albeit it starts from the 1980s when the world’s oil price was booming at that time. Both East Kalimantan and Riau have been known as natural rich provinces since the independence of Indonesia. The decreasing pattern in relative transition graphs may be affected by the decreasing of their stocks in natural resources such as oil and gas, or mining. Another fact to explore is the Province of Jakarta which started to increase its relative position the beginning of 1980s and consistently increase until year 2012. Jakarta is known as a capital of Indonesia and centre of businesses with very strong evidence of economic agglomeration. In our data, Jakarta is included in Club 1. But, if we look at the graph we can see that in coming years Jakarta will be leaving the other members in Club 1 and to diverge. There are two possibilities in this concern. First, even though Jakarta has acceleration in per capita GRP but it has been known of over use of its economic sources. Land becomes scarce in Jakarta with price increases every year, traffic jams happen every day which causes the bottle neck of development. In this case, it is almost impossible for Jakarta to have an increasing growth in the future. Lastly, if introduction of new technology is possible is the future that can solve the problems of Jakarta then the increasing trend can be reached, assuming factor of productions are fixed. In club 1, provinces that showing upward trend in their relative positions are Bangka Belitung and West Papua. These two are new provinces formed in 2000. Bangka Belitung was a part of Riau and known of its mining activities especially in nickels. Since the mining industry is no longer the first source of income for Bangka Belitung, the province now focuses on tourism industry to generate income. Meanwhile, West Papua is also know of its mining especially gold and it was a part of Province of Papua. West Papua will be leaving behind its former province in its relative position from other provinces. On our data, West Papua shows an increasing trend in per capita GRP and join the first club.

Figure 4.6 : Relative Transition Coefficient (h_it) of Per Capita GRP for the First Club (10 Provinces)

Note: Series are Hodrick-Prescott trend of Per Capita GRP.

Even though the behaviour of provinces in Club 2 can be seen in Figure 4.2, the detail about its individual heterogeneities are clearer in Figure 4.7. The most interesting thing is to look at the Province of Aceh which shows the increasing position from 1969 to the late 1980, then started to decrease until 2012. Aceh is the unique province with rich history of its own. It was the province that gained Special Region while experiencing decades of separatism conflicts. It is also the province which most hit when Tsunami occurred in 2004. The mixed events has dragged Aceh into slower development relative to its counterparts. South Sumatra and Papua show their decreasing position across time with some fluctuations in the past. Papua is known as the Eastern part of Indonesia which lagged behind in development. The split-up of natural resource rich province of West Papua also affected Province of Papua. This can be seen from the fast decreasing relative position in the last decade. The rest of provinces in Club 2 show consistently low development for decades. To mention there are East Nusa Tenggara, Maluku, and some provinces in Sulawesi. Although Banten and Gorontalo have joined Club 2, but their relative positions show increasing trend in the future. If this trend persists, these provinces might jump from Club 2 to join Club 1 or may form another new club of per capita GRP. Figure 4.7 : Relative Transition Coefficient (h_it) of Per Capita GRP for the Second Club (23 Provinces)

Note: Series are Hodrick-Prescott trend of Per Capita GRP

Figure 4.8 shows the average of relative transition heterogeneity of Club 1 and Club 2. The Graphs tell the story of the clubs evolution from 1969 to 2012. From 1969 both clubs tend to converge each other until the beginning of 1990s, and do not change afterwards. With supported by Gini index, the income distribution is becoming unequal across provinces in Indonesia.

Figure 4.8 : Average-Relative Transition Coefficient (h_it) of Per Capita GRP for Two Clubs

Note: Series are Hodrick-Prescott trend of Per Capita GRP, average for each club. 4.2 Convergence of Social Indicators We also test the clustering convergence in some social indicators. The ideal approach is to select many social indicators that related to the economic development. Since it is difficult to obtain many indicators with long period of observation, then we select five social indicators that we consider to have enough period of observations. The two indicators we focus on are Human Development Index (available from 1996-2012) and Gini Index (available from 1976-2012). The other three indicators for comparison are school enrolment ratios (1996-2012), literacy ratio (1996-2012), and poverty rate (2002-2012). Since not all data are available for all province and the club convergence test requires balance panel data, then we interpolate data for some missing years for some provinces. More detail on individual province figures and group by island are on section 3, and also on Appendix A.4-A.6.
Similar to the test we apply on per capita GRP above, first we test the existence of full panel convergence. Table 4.5 shows that the hypothesis of full panel convergences are rejected since t-stat<-1.65 (for α=5%). Then, we proceed with club convergence test for both indicators. Table 4.5: Log-t test for full panel (33 Province) Indicators b ̂ t-statistics Human Developmen Index -0.460 -7.185 Gini Index -0.232 -4.581 Note: t-statistics indicate the null hypothesis of convergence are rejected for both indicators.

Human Development Index and Gini Index

Club convergence test for Human Development Index (HDI) resulted into four clubs, and two clubs for Gini Index. Table 4.6 and 4.8 show the classification of the tests, table 4.7 and 4.9 show the descriptive statistics, while Figure 4.9 and 4.10 show the relative transition coefficients for those indicators. The test results in social indicators show different pattern. Indicator of HDI shows four convergence clubs which provinces resulted from per capita GRP spread into all clubs in HDI. For example North Sulawesi now join Jakarta and Riau Islands in first club, while West Papua of first club in per capita GRP now has jumped into fourth club. The first club is the cluster for higher HDI, while the fourth club is the lowest HDI. This result tells us that being in club of higher per capita GRP does not guarantee to become the club of higher HDI. This fact can be explained as follows. Province with rich natural resources will have high gross domestic products and per capita products, but the products do not all go the total population. Instead, the products are only in the hand of a part of the population. This can be measured by the high income inequality among the population. The other reason is that per capita GRP is measured at aggregate level and a proxy to per person income. Measurement error exists in this matter, and household level data are needed to have a better measurement. Most provinces of the eastern part of Indonesia are in the fourth club, the lowest HDI club. Province of Banten is also in the fourth club, split-off from the main Province of West Java in 2000. This province has been accused of bad public management and services, which its governor has been involved and jailed in 2013 of bribery and corruption case.

Table 4.6: Convergence Club of Human Development Index Club Member Provinces Club 1 [7 Provinces] Riau, Riau Islands, Jakarta, Yogyakarta, Central Kalimantan, East- Kalimantan, North Sulawesi. Club 2 [8 Provinces] North Sumatera, West Sumatera, Jambi, South Sumatera, Bengkulu, Bangka-Belitung, Central Java, West Sulawesi. Club 3 [12 Provinces] Aceh, Lampung, West Java, Bali, West Kalimantan, Central Sulawesi, South Sulawesi, Southeast Sulawesi, Gorontalo, East Java, South Kalimantan, Maluku. Club 4 [6 Provinces] Banten , East Nusa Tenggara, West Nusa Tenggara, North Maluku, West Papua, Papua. Note: province name in bold and italic means membership in first club of per capita GRP convergence test, otherwise from the second club.

Table 4.7: Descriptive Statistics for HDI (2011) N Mean S.e. Min. Max. Club 1 7 76.35 0.33 75.06 77.97 Club 2 8 73.18 0.48 70.11 74.65 Club 3 12 71.58 0.29 69.66 72.84 Club 4 6 68.24 0.88 65.36 70.95

Figure 4.9 : Relative Transition Coefficient (h_it) of Human Development Index

Club 1      Club 2
 
Club 3      Club 4

Note: Series are Hodrick-Prescott trend, with λ=6.25 for yearly data, as suggested by Ravn and Uhlig (2002)

Results from Gini index shows only Jakarta and West Papua left in the first club from high per capita GRP club. First Gini index club refers to high inequality of income, while the second club refers to lower inequality. Provinces from Sumatra are in the second club, while most of provinces from Sulawesi are in the first club. Table 4.8: Convergence Club of Gini Index Club Member Provinces Club 1 [10 Provinces] Jakarta, West Java, Yogyakarta, Bali, North Sulawesi, South Sulawesi, Southeast Sulawesi, Gorontalo, West Papua, Papua.

Club 2 [23 Provinces] Aceh, North Sumatera, Riau, Bangka Belitung, Riau Islands, Jambi, Lampung, West Sumatera, South Sumatera, Bengkulu, Central Java, Banten, East Nusa Tenggara, West Nusa Tenggara, West Kalimantan, Central Sulawesi West Sulawesi, Maluku, North Maluku, East Java, Central Kalimantan, South Kalimantan, East Kalimantan. Note: province name in italic and bold means membership in first club of per capita GRP convergence test, otherwise from the second club.

Table 4.9: Descriptive Statistics for Gini Index (2013) N Mean S.e. Min. Max Club 1 10 0.427 0.004 0.403 0.442 Club 2 23 0.364 0.005 0.313 0.407

Figure 4.10: Relative Transition Coefficient (h_it) of Gini Index

Club 1      Club 2

Note: Series are Hodrick-Prescott trend, with λ=6.25 for yearly data, as suggested by Ravn and Uhlig (2002)

4.2.2 Education outcomes and poverty rate The results from education outcomes are similar which consistently form two convergence clubs. Some variations emerge from provinces that jump from one to another club if we compare the net enrolment ratio from primary, junior secondary and senior secondary school. However, some provinces show persistently first club membership i.e. Aceh, North Sumatra, Riau, Lampung, Riau Islands, Jakarta, East Java, Bali, East Kalimantan and Maluku. This group of provinces are the members of per capita GRP club, which tell us the strong relationship between those two indicators. Table 4.10 shows one of the convergence club of education outcome, while table 4.11 shows the descriptive statistics. Table 4.10: Convergence Club of Net Enrolment Ratio (Senior Secondary School) Club Member Provinces

Club 1 [17 Provinces] Aceh, North Sumatera, West Sumatera, Riau, Bengkulu, Riau Islands, Jakarta, Yogyakarta, East Java, Bali, West Nusa Tenggara, East Kalimantan, North Sulawesi, Southeast Sulawesi, Maluku, North Maluku, West Papua.

Club 2 [16 Provinces] Jambi, South Sumatera, Lampung, Bangka Belitung, West Java, Central Java, Banten, East Nusa Tenggara, West Kalimantan, Central Kalimantan, South Kalimantan, Central Sulawesi, South Sulawesi, Gorontalo, West Sulawesi, Papua. Note: province name in italic and bold means membership in first club of per capita GRP convergence test, otherwise from the second club.

Table 4.11: Descriptive Statistics for Education outcome (2012) N Mean S.e. Min. Max Club 1 10 51.94 2.24 42.12 61.71 Club 2 23 49.89 1.68 30.06 64.02 Note: the selected indicator for education outcome is net school enrollment ratio for senior secondary school

Figure 4.11 shows the relative transition coefficient for the four educational outcomes. Even though provinces form two convergence clubs in those indicators, the relative position of most of the provinces do not diverge heavily among others with exception for Papua and East Nusa Tenggara. On primary school enrollment and literacy ratios the position of those two provinces are far below the others and show a decreasing trend in time. Meanwhile, the East Nusa Tenggara was on the lowest level of the Junior Secondary School enrolment but its relative position is getting higher and catching up the rest. The top member of all clubs are Aceh, West Sumatra, Jakarta, Yogyakarta and North Sulawesi.

  Figure 4.11 : Relative Transition Coefficient (h_it) of education outcomes

Net enrolment ratio (%) – Primary School        Net enrolment ratio (%) -  Junior Secondary School
 
Net enrolment ratio (%) -  Senior Secondary School      Literacy ratio 15+ year old (%)

Note: Series are Hodrick-Prescott trend, with λ=6.25 for yearly data, as suggested by Ravn and Uhlig (2002)

Figure 4.12 : Relative Transition Coefficient (h_it) of Poverty Rate

Note: Series are Hodrick-Prescott trend, with λ=6.25 for yearly data, as suggested by Ravn and Uhlig (2002)

Comparing Convergences Results  

The results from club convergence analyses do not always put one province in the same and consistent club from on indicator to another. In other words, the growth in per capita GRP does not assure development in social indicators even though it is commonly accepted that higher income is correlated with social development. Some provinces may have high per capita GRP and high growth rate, but may have lower social development indicators. Table 4.12 summarizes the results from club convergence tests.
We are interested to see the relation between the results from convergence club in per capita GRP and social indicators. There are consistently five provinces that belong to the elite club, high per capita GRP and developed social indicators. Meanwhile, some other five provinces with high per capita GRP club do not belong to high development in social indicators club. One of the reason is that the aggregate measurement of income (per capita GRP) does not reflect the actual income at individual or household level. Trickle-down effect does not occur and the wealth controlled by small fraction of the population. There are 16 provinces which are consistently low in club membership meaning that slower growth in per capita GRP and also in selected social indicators development. Provinces that confirm development in social indicators not always related to high per capita growth are West Sumatra, Bengkulu, Yogyakarta and North Sulawesi. North Sulawesi has been known as province with relatively good education and human resources since the Dutch colonization era although the economy is only supported by agricultural sector. Yogyakarta is the province heavily relied on tourism and supporting services but it has good infrastructures allowing people to commute within and between provinces.

Table 4.12: Relation Between Convergence Club in Per Capita GRP and Social Indicators Social Indicators Development Club High Low

Per Capita GRP Club High Jakarta, North Sumatra, Riau, Riau Islands, East Kalimantan Bangka Belitung, East Java, Central Kalimantan, South Kalimantan, West Papua Low West Sumatera, Bengkulu, Yogyakarta, North Sulawesi West Java, South Sulawesi, Gorontalo, Lampung, Jambi, South Sumatera, Central Java, West Kalimantan, Aceh, Maluku, Central Sulawesi, North Maluku, West Nusa Tenggara, East Nusa Tenggara, Banten, Bali. Note: Per capita GRP (Club 1=high, Club 2=low), Social Indicators Development (HDI (Club 1&2=high, Club 3&4=low, NER (Club 1=high, Club 2=low)).

The better analysis on the development clustering allows better policy to be taken by the government. But this might raise a question about which target indicator to pursue by the government: income generating or social development. Kenny (2005) compares income and some social indicators of India and United Kingdom and shows that India’s social indicators deviation is getting smaller compared to United Kingdom. He asks an intriguing question: if everything that matter is converging why do we worry about income? In the society’s point of view the higher income and better social development are wanted. But, in the policy maker’s point of view this could still raise a policy debate especially on the budget allocation policy. Conclusion This paper looks at the development and inequality in Indonesia by exploring the socio-economic data of Indonesia at international, national, and provincial level. We compare the position of Indonesia with selected international countries in terms of its size of economy, per capita income, social indicators such as human development index, Gini ratio, and other related social indicators for measuring development of a nation. We also use data of provincial level for analyzing the economic performance, the state of social developments, and the notion of economic convergence. On the international comparison, the world rank for socio-economic indicators the position of Indonesia is above 50 percent except for inequality. Gini ratio Indonesia was very low and stable for long time compare to selected countries which is peculiar. The strong government’s role in economy since 1970s until liberalization era early 1980s may cause this trend. Meanwhile, we find that Indonesia is catching up to ASEAN and emerging countries group in terms of per capita GDP, literacy, mortality, fertility, access to sanitation and clean water. But, among other developing countries, the variations across provinces in Indonesia are still higher for those indicators except sanitation. High income per capita is experienced by rich natural resource provinces and capital city of Jakarta. Furthermore, there is a huge gap of development across provinces. Income inequality is increasing at national, province and island level but the variation of inequality across islands tends to increase while the variation across provinces tends to decrease. To assess the behavior on regional level, we use the provincial per capita GRP, and some selected social indicators. We employ the newest sigma-convergence methodology and analyze the relative variation of the indicators over time. We use the club convergence test developed by Phillips and Sul (2007) which endogenously determined the club membership. We also analyze the catching-up effect of poor provinces to their rich counterparts.
On the provincial level we find that per capita GRP do not converge for all provinces, instead we find that the existence of clusters. Provinces in Indonesia are clustered into two clubs of per capita GRP, four clubs of human development index, two clubs of GINI index, and two clubs of literacy rate. Member of club in per capita GRP does not always belong in the same and consistent club from on indicator to another. The provinces that show robust development are Jakarta, North Sumatra, Riau, Riau Islands, and East Kalimantan. On the other hand, the provinces which are considered underdeveloped are West Java, South Sulawesi, Gorontalo, Lampung, Jambi, South Sumatera, Central Java, West Kalimantan, Aceh, Maluku, Central Sulawesi, North Maluku, West Nusa Tenggara, East Nusa Tenggara, Banten, Bali. The main findings of this paper are two things i.e. increasing inequality between islands while decreasing inequality within islands and per capita GRP as well as other socio indicators do not converge but clustered into two or more clubs. These suggest that there is dynamic process throughout Indonesia which needs to be further researched. The next study should investigate how this phenomenon occurred i.e. individual mobility and or firms choice location due to higher economic growth, or better public services, or better local government, or others. We have shown in our data that inequality at national level, measured by Gini index, is increasing especially after year 2001. Meanwhile, if we compare the variation Gini index among provinces with the variation of among islands (provinces added up in same islands), the latter shows lower coefficient of variation. This means the inequality at provincial level is higher than island level. Some explanation can be derived from this results. First, the impact of decentralization and democratization processes resulted the increase in inequality at national level. In these combined processes we may see the transition from highly subsidized economy into more open market economy. When the market becomes open, combined with decentralization, it creates the two faced opportunities: 1) the more productive labor or firms will benefit from the competition, or 2) rent-seeker emerges from the locals who gain power from decentralization process. The actors who neither these two groups would be left behind. This will worsen the distribution of income among citizens. Second, we may consider the island effect which resulted the lower variations of income between islands. The dynamics of economic actors: firms, labors, local governments only happens in one island where the economic actors belong to. The sorting process exists in one particular island where the economic activities are considered in one island. People easily commute between provinces in the same island for work but not between island. We might also think that the number of poor becomes rich people is increasing and the number rich stay the same or becomes poorer. Third, the process of migration between island. In this process the lower income group migrate to a richer island to seek better opportunities or the higher income group migrate to lower income island to invest or to obtain more lower price of land and properties. Our findings lead to interesting future researches on the dynamics of the economic agents and role of local governments in affecting the distribution of income. For instance, what might the drivers of convergence clubs? Is it resource endowment (natural resources or taxes), foreign direct investment (FDI), or infrastructure? FDI might stimulates agglomeration process. On the other hand, infrastructure might affect the behavior of labors in commuting where eventually contribute to the region’s income from remittance. If convergence club is driven by FDI, the next interesting thing to know is where the locations of FDI, industrialization, or agglomeration take place. We may also be interested to analyze why some regions are high in FDI while others are low. In case that one region is high with FDI, is there any association with better regulation of local government or not? Is high FDI associated with higher budget allocation on public services which leads to better human capital in particular region? Finally, if convergence club is driven by remittance, we are interested to see the pattern of urbanization and agglomeration. Moreover, how is the behavior of people in creating such pattern, how far they live from district center, how long they travel for work, and how they commute to the center.

References ADB, 2012, “Asian Development Outlook 2012: Confronting Rising Inequality in Asia”, Manila. Bhalla, S. Surjit, 2012, “Devaluing to Prosperity: Misaligned Currencies and Their Growth Consequence”, Peterson Institute for International Economics Booth, Anne and Sundrum, R.M (1981), ‘Income Distribution’, in Anne Booth and Peter McCawly (eds.,), The Indonesian Economy during the Soeharto Era, Kuala Lumpur: Oxford University Press Booth, Anne (1992), ‘Income Distribution and Poverty’, in Anne Booth (ed.,), The oil Boom and After: Indonesian Economic Policy and Performance in the Soeharto Era, Singapore: Oxford University Press Booth, Anne (2000) Poverty and Inequality in The Soeharto Era: An Assessment, Bulletin of Indonesian Economic Studies, Vol. 36, No. 1, pp. 73-104 Bolt, J. and J. L. van Zanden, 2013, The First Update of the Maddison Project; Re-Estimating Growth Before 1820. Maddison Project Working Paper 4. Chongvilaivan, Aekapol, “Individual Income Inequality and Its Drivers in Indonesia: A Theil Decomposition Reassessment”, ISEAS Working Paper No 2, 2013. Dalgaard, Carl-Johan and Jacob Vastrup, “On the measurement of σ-convergence”, Economic Letters 70 (2001), 283-287. Dollar, David and Arrt Kraay, “Growth is Good for the Poor”, Journal of Economic Growth 7.3 (2002): 195-225. Frankema, E and D. Marks (2009), ‘Was It Really “Growth with Equality” under Soeharto? A Theil Analysis of Indonesian Income Inequality, 1961-2002’, Economics and Finance in Indonesia, Vol. 57, No. 1, pp. 48-80 Frankema, E and D. Marks (2010), ‘Growth, Stability, and Equality? Re-assessing Indonesian Inequality from a Comparative Perspective’, Economic History of Developing Regions, Vol. 25, No. 1, pp. 75-104 Furceri, D., “β and σ-convergence: A mathematical relation of causality”, Economics Letter 89 (2005), 212-215. Hartono, D and Irawan, Tony, “Decentralization Policy and Equality: A Theil Analysis of Indonesian Income Inequality”, Center for Economics and Development Studies, Department of Economics, Padjadjaran University, Working Paper, 2008. Kanbur, R, 1998, “Income Distribution and Development” in A.B. Atkinson & F. Bourguignon, eds., Handbook of Income Distribution, North Holland. Mishra, Satish C, 2009, “Economic Inequality in Indonesia: Trends, Causes and Policy Response”, UNDP Report for Strategic Asia. OECD, Economic Outlook for Southeast Asia, China, and India 2014: Beyond the Middle-Income Trap, OECD 2013. Panapoulou, E. and Theologos Pantelidis (2009), “Club Convergence in Carbon Dioxide Emissions”, Environ Resource Econ, 44, 47-70.. Phillips, P.C.B. and Donggyu Sul (2007),”Transition Modeling and Econometric Convergence Tests”, Econometrica, Vol.75, No.6 (November, 2007), 1771-1855. Ramsey, Frank P.,1928, “A Mathematical Theory of Saving”, Economic Journal, 38, December 1928: 543-559. Sakamoto, H, 2007, “The Dynamics of Inter-Provincial Income Distribution in Indonesia”, The International Center for the Study of East Asian Development, Kitakyushu, Working Paper, 2007, 25. Sala-i-Martin, X., 1995, “The Classical Approach to Convergence Analysis”, Universitat Pompeu Fabra, Economics Working Paper, 117, June 1995. Sala-i-Martin, X., 1990, “On Growth and States” Ph.D Dissertation, Harvard University. Solow, Robert M., 1956, “A Contribution to the Theory of Economic Growth”, Quarterly Journal of Economics, 70: 65-94. Tadjoeddin, M. Z., (2013), “Miracle that never was: disaggregated level of inequality in Indonesia”, International Journal of Development Issues, Vol. 12 No. 1, 2013 pp. 22-35 Yusuf A.A, A. Komarulzaman and M. Purnagunawan and B. Resosudarmo. 2013, “Growth, Poverty and Labor Market Rigidity in Indonesia: A General Equilibrium Investigation”, Working Papers in Economics and Development Studies (WoPEDS) 201304, Department of Economics, Padjadjaran University (revised January 2013). Yusuf, A.A., Resosudarmo, B.P. 2008, “Mitigating Distributional Impact of Fuel Pricing Reform: Indonesian Experience”. ASEAN Economic Bulletin 25, 32–47. Zanden, J.L.van, 2012. “An economic history of Indonesia 1800-2010”

Appendix A1: GDP per Capita Indonesia and Selected Countries, 1980-2012

Source: WDI, CBS Indonesia

  A2: Socio-economic Indicators Indonesia and Selected Countries

Source: CBS Indonesia, CBS China, CBS Vietnam, CBS Thailand, CBS Philippines, WDI  

A.3: Education Outcome Indicators for Selected Countries 1960-2011 Country Name School enrollment, primary School enrollment, secondary School enrollment, tertiary % net % net % gross 1971 2011 % of change 1960 2011 % of change 1970 2011 % of change ASEAN COUNTRIES
Indonesia 70.6 95.8 25.2 16.4 74.4 58.0 2.9 24.9 22.0 Java Island 93.1* 95.3 2.2
Malaysia 83.9 95.9 12.0 32.7 68.6 35.9 3.9* 37.1 33.3 Philippines 96.8 88.3 -8.5 46.2* 61.6 15.4 17.7 28.2 10.5 Thailand 75.2 89.7 14.5 15.5 71.5 55.9 2.9 46.4 43.6 EMERGING COUNTRIES
Brazil 69.8 94.4 24.7 .. .. .. 4.7 25.6 20.9 China 96.1* 96.4* 0.3 .. .. .. 0.1 26.8 26.7 India 60.8 93.3 32.5 .. .. .. 5.0 17.9 12.8 South Korea 96.5 98.6 2.1 36.0 96.0 60.0 7.2 103.1 95.9 East Asia & Pacific 89.6 95.8 6.2 56.2* 73.0 16.8 2.9 30.1 27.2 AFRICAN COUNTRIES
Kenya 62.3* 81.8 19.5 33.3* 50.0 16.6 0.9 4.0 3.2 Nigeria 61.9* 56.2 -5.6 .. .. .. 0.7 10.4 9.7 South Africa 65.6 85.0 19.4 50.5 68.8 18.3 4.0* 15.0 11.0 DEVELOPED COUNTRIES
Australia 96.2 97.1 0.9 86.9* 85.5 -1.4 15.8 79.9 64.1 Japan 99.9 100.0 0.1 93.2* 99.5 6.3 17.6 59.7 42.1 Singapore .. .. .. .. .. .. .. .. United States 80.7 94.6 13.8 88.6* 89.5 0.9 47.0 94.8 47.8 Euro area 96.3 98.2 1.9 72.3 92.3 20.0 19.3 62.5 43.1 Note: * The earliest and the most recent data differ for some countries: Literacy Rate: the earliest data for Java Island 2003 School Enrollment, Primary: the earliest data for China (1987), Kenya (1999), Nigeria (1999), Java Island (2003) and the most recent data for China is 1997. School Enrollment, Secondary: the earliest data for Australia (1993), Japan (1978), Kenya (2000), South Africa (1994), Philippine (1979), United States (1987), East Asia & Pacific (1998) School Enrollment, Tertiary: the earliest data for Malaysia (1982) and the most recent data for South Africa is 1994
Source: WDI, CBS Indonesia   A.4: GRP per capita and social indicators (part 1) No. Province GRP per capita (Mio. Rupiah) Infant Mortality Rate Fertility rate 2000 2012 Annual growth (%) 2000 2010 Change (%) 2000 2012 Change 1 Aceh 10.1 8.9 -1.1 39.7 28.0 -11.7 2.44 2.79 0.4 2 North Sumatera 5.9 10.0 4.4 43.7 25.7 -18.0 2.84 3.01 0.2 3 West Sumatera 5.4 9.6 4.9 52.7 29.7 -23.0 2.95 2.91 0.0 4 Riau 14.1 13.2 -0.5 47.7 23.0 -24.7 2.45 2.82 0.4 5 Jambi 4.0 6.8 4.5 52.7 29.0 -23.7 2.37 2.51 0.1 6 South Sumatera 6.7 9.6 3.1 52.7 25.3 -27.3 2.33 2.56 0.2 7 Bengkulu 3.3 5.1 3.6 52.7 27.7 -25.0 2.49 2.51 0.0 8 Lampung 3.5 5.4 3.8 47.7 23.0 -24.7 2.42 2.45 0.0 9 Bangka Belitung Islands 6.8 11.4 4.4 52.7 26.7 -26.0 2.53 2.54 0.0 10 Riau Islands - 28.2 - - 20.3 - 2.38 2.4 11 Jakarta 27.3 49.6 5.1 24.8 14.0 -10.8 1.66 1.82 0.2 12 West Java 5.5 8.3 3.5 56.6 26.0 -30.7 2.28 2.43 0.2 13 Central Java 3.7 6.5 4.8 43.7 21.0 -22.7 2.14 2.2 0.1 14 Yogyakarta 4.3 6.7 3.7 24.8 15.7 -9.1 1.79 1.94 0.2 15 East Java 5.8 10.8 5.3 47.7 25.0 -22.7 1.87 2 0.1 16 Banten 5.6 8.9 3.9 65.6 24.3 -41.3 2.37 2.35 0.0 17 Bali 5.5 8.9 4.1 35.7 20.0 -15.7 2.03 2.13 0.1 18 West Nusa Tenggara 3.0 4.0 2.3 88.5 48.3 -40.2 2.69 2.59 -0.1 19 East Nusa Tenggara 2.1 3.1 3.5 56.6 38.7 -18.0 3.46 3.82 0.4 20 West Kalimantan 4.8 6.9 3.1 56.6 28.3 -28.3 2.62 2.64 0.0 21 Central Kalimantan 5.9 8.3 2.9 47.7 23.3 -24.4 2.21 2.56 0.4 22 South Kalimantan 5.8 9.5 4.3 69.6 34.3 -35.3 2.3 2.35 0.1 23 East Kalimantan 33.6 35.9 0.5 39.7 21.0 -18.7 2.32 2.61 0.3 24 North Sulawesi 5.3 9.1 4.7 27.8 25.0 -2.8 2.1 2.43 0.3 25 Central Sulawesi 4.0 7.7 5.6 65.6 45.0 -20.6 2.81 2.94 0.1 26 South Sulawesi 3.8 6.5 4.6 56.6 31.0 -25.7 2.55 2.55 0.0 27 Southeast Sulawesi 3.2 5.7 4.9 52.7 39.7 -13.0 3.14 3.2 0.1 28 Gorontalo 1.8 3.7 6.3 56.6 56.3 -0.3 2.63 2.76 0.1 29 West Sulawesi - 4.9 - 48.0 48.0 3.33 3.3 30 Maluku 2.4 3.4 3.1 60.6 45.0 -15.6 3.29 3.56 0.3 31 North Maluku 2.3 3.4 - 74.6 39.7 -34.9 3.04 3.35 0.3 32 West Papua - 18.1 - 28.0 28.0 3.18 3.2 33 Papua 10.1 7.3 -2.6 56.6 19.0 -37.7 2.38 2.87 0.5 Average 6.8 10.5 3.6 145.0 26.0 -119.0 2.27 2.41 0.4 Source: Authors’ calculation from CBS Indonesia data A.5: GRP per capita and social indicators (part 2) No. Province Literacy Ratio >15 year old (%) School Participation 6-12 yo (%) School Participation 16-18 yo (%) 2000 2010 Change (%) 2003 2012 Change (%) 2003 2012 Change (%) 1 Aceh 96.8 97.3 0.5 98.1 99.4 1.3 72.3 74.4 2.2 2 North Sumatera 95.6 96.7 1.1 98.3 98.6 0.3 63.2 69.7 6.5 3 West Sumatera 96.1 97.8 1.7 96.9 98.4 1.4 63.2 71.4 8.2 4 Riau 95.2 96 0.8 97.0 98.1 1.1 60.4 65.8 5.4 5 Jambi 95.2 96.9 1.7 97.3 98.7 1.4 52.9 59.1 6.2 6 South Sumatera 93.6 95.7 2.1 96.5 98.0 1.5 44.6 58.3 13.7 7 Bengkulu 91.6 95.1 3.5 96.3 99.0 2.7 52.4 66.7 14.3 8 Lampung 91.5 95.9 4.4 96.1 98.6 2.5 48.3 59.8 11.5 9 Bangka Belitung Islands 100 97.8 -2.2 96.1 97.7 1.7 44.2 50.9 6.7 10 Riau Islands 98.4 99.1 0.7 98.3 - 69.7 69.7 11 Jakarta 93.8 96.2 2.4 98.2 99.0 0.8 71.6 60.8 -10.8 12 West Java 85.8 90.4 4.6 96.3 98.3 2.1 43.5 55.7 12.2 13 Central Java 85.7 92 6.3 97.9 98.9 1.0 48.1 58.6 10.4 14 Yogyakarta 83.4 89.3 5.9 98.7 99.8 1.1 73.6 80.2 6.6 15 East Java 93.8 96.5 2.7 97.2 98.7 1.5 51.7 61.7 10.0 16 Banten 84.4 90.2 5.8 96.8 98.3 1.5 45.7 58.6 12.9 17 Bali 75.1 83.7 8.6 97.3 99.2 1.9 62.0 70.8 8.8 18 West Nusa Tenggara 84.9 88.7 3.8 94.7 98.2 3.5 43.0 60.8 17.8 19 East Nusa Tenggara 87.6 91.1 3.5 90.8 96.1 5.4 37.8 62.2 24.4 20 West Kalimantan 96.2 97.5 1.3 92.1 96.6 4.6 45.0 54.7 9.7 21 Central Kalimantan 93.5 96.4 2.9 97.7 98.5 0.8 50.3 54.1 3.8 22 South Kalimantan 94.9 97.5 2.6 96.3 97.9 1.6 41.4 57.6 16.1 23 East Kalimantan 98.9 98.8 -0.1 96.7 99.2 2.5 61.3 71.2 9.9 24 North Sulawesi 93.6 94.9 1.3 97.4 98.2 0.8 54.3 65.4 11.1 25 Central Sulawesi 83.4 88.7 5.3 96.6 96.5 -0.1 44.2 59.6 15.4 26 South Sulawesi 90.5 91.5 1 92.4 97.6 5.2 45.4 61.6 16.2 27 Southeast Sulawesi 94.7 95.2 0.5 95.5 97.4 1.9 47.3 65.3 17.9 28 Gorontalo 100 88.8 -11.2 90.6 97.5 6.9 35.1 57.8 22.7 29 West Sulawesi 97 97.1 0.1 95.7 95.7 56.4 56.4 30 Maluku 95.5 96.4 0.9 95.7 98.3 2.6 55.7 68.4 12.7 31 North Maluku 100 94.7 -5.3 97.4 98.2 0.8 50.9 68.3 17.3 32 West Papua 74.5 65.7 -8.8 95.6 95.6 67.2 67.2 33 Papua 89.8 93.2 3.4 85.8 75.3 -10.4 49.4 50.7 1.3 INDONESIA 91 93.4 2.4 96.4 98.0 1.5 51.0 61.1 10.1 Source: CBS Indonesia.

A.6: GRP per capita and social indicators (part 3) No. Province Poverty Rate (%) Gini Index Human Development Index 2000 2010 Change 1996 2013 Change (%) 1996 2011 Change (%) 1 Aceh 26.65 18.58 -8.1 0.259 0.341 0.08 69.40 72.16 2.8 2 North Sumatera 13.90 10.41 -3.5 0.301 0.354 0.05 70.50 74.65 4.2 3 West Sumatera 11.90 8.00 -3.9 0.278 0.363 0.09 69.20 74.28 5.1 4 Riau 11.20 8.05 -3.2 0.3 0.374 0.07 70.60 76.53 5.9 5 Jambi 10.27 8.28 -2.0 n.a 0.362 - 69.30 73.3 4.0 6 South Sumatera 19.15 13.48 -5.7 0.246 0.348 0.10 68.00 73.42 5.4 7 Bengkulu 22.13 17.51 -4.6 0.300 0.383 0.08 68.40 73.4 5.0 8 Lampung 22.19 15.65 -6.5 n.a 0.313 - 67.60 71.94 4.3 9 Bangka Belitung Islands 9.54 5.37 -4.2 0.273 0.386 0.11 - 73.37 - 10 Riau Islands 10.30 6.83 - 0.276 0.356 - - 75.78 - 11 Jakarta 4.61 3.70 -0.9 0.363 0.433 0.07 76.10 77.97 1.9 12 West Java 13.55 9.89 -3.7 0.356 0.411 0.06 68.20 72.73 4.5 13 Central Java 20.43 14.98 -5.5 n.a 0.399 - 67.00 72.94 5.9 14 Yogyakarta 18.99 5.88 -3.1 0.291 0.387 0.10 71.80 76.32 4.5 15 East Java 19.98 13.08 -6.9 0.353 0.439 0.09 65.50 72.18 6.7 16 Banten 9.07 5.71 -3.4 0.311 0.364 0.05 - 70.95 - 17 Bali 6.63 3.95 -2.7 0.309 0.403 0.09 70.10 72.84 2.7 18 West Nusa Tenggara 24.99 8.02 -7.0 0.286 0.364 0.08 56.70 66.23 9.5 19 East Nusa Tenggara 27.51 20.41 -7.1 0.296 0.352 0.06 60.90 67.75 6.9 20 West Kalimantan 12.91 7.96 -5.0 0.300 0.396 0.10 63.60 69.66 6.1 21 Central Kalimantan 9.38 6.19 -3.2 0.271 0.350 0.08 71.30 75.06 3.8 22 South Kalimantan 7.01 5.01 -2.0 0.292 0.359 0.07 66.30 70.44 4.1 23 East Kalimantan 11.04 6.38 -4.7 0.318 0.371 0.05 71.40 76.22 4.8 24 North Sulawesi 11.42 7.64 -3.8 0.344 0.422 0.08 71.80 76.54 4.7 25 Central Sulawesi 22.42 14.94 -7.5 n.a 0.437 - 66.40 71.62 5.2 26 South Sulawesi 14.11 9.82 -4.3 0.302 0.407 0.11 66.00 72.14 6.1 27 Southeast Sulawesi 21.33 13.06 -8.3 0.323 0.429 0.11 66.20 70.55 4.3 28 Gorontalo 27.35 17.22 -10.1 n.a 0.349 - - 70.82 - 29 West Sulawesi 19.03 13.01 -6.0 0.311 0.426 0.12 - 70.11 - 30 Maluku 31.14 20.76 -10.4 0.269 0.370 0.10 68.20 71.87 3.7 31 North Maluku 11.97 8.06 -3.9 n.a 0.318 - - 69.47 - 32 West Papua 39.31 27.04 -12.3 0.386 0.442 0.06 - 69.65 - 33 Papua 40.78 30.66 -10.1 n.a 0.431 - 60.20 65.36 5.2 INDONESIA 16.58 11.66 -4.9 0.355 0.413 0.06 67.70 72.77 5.1 Source: CBS Indonesia.

summary(cars)

##      speed           dist    
##  Min.   : 4.0   Min.   :  2  
##  1st Qu.:12.0   1st Qu.: 26  
##  Median :15.0   Median : 36  
##  Mean   :15.4   Mean   : 43  
##  3rd Qu.:19.0   3rd Qu.: 56  
##  Max.   :25.0   Max.   :120

You can also embed plots, for example:

Note that the echo = FALSE parameter was added to the code chunk to prevent printing of the R code that generated the plot.