4 In these later sections you sometimes describe the sources for funding consumption as

I think this is a very important contribution to the project and the NTA volume. You have elaborated on NTA in a very useful way by distinguishing economic flows by income level and residence. This can serve as a useful model for others hoping to undertake similar work. The results are very interesting and clearly demonstrate the value of looking at important population subgroups. In addition, this is clearly presented, well-organized, and meets the difficult challenge of doing it in only 19 pages. Thanks for your hard work on this. I think this is ready to go to Sandra Ward who is handling the copy editing of all of the chapters, but I have a few specific suggestions that I would like to see you implement. I don’t think it would take very long and I think it would be worthwhile making a few changes. I’ll mark these on the paper but provide some notes here.

1) On page 9 I would add a heading to make it clear that the estimates that follow are NTA estimates. I think it would be good to mention how they differ from the estimates in the preceding tables. I think it would be a very good idea to move Table 3 to this point in the paper and to make it a per capita table. Also I would make it a per capita table rather than an aggregate table. This would allow the reader to see how average values of some of these key components, e.g., consumption and labor, vary across the subgroups.

2) I found the paragraph that begins “The production at age 60 are approximately over 45 percent relative to production” on page 14 in your original difficult to follow.

3) A few comments are inserted in section IV. I’ve changed “finance” to “fund” and “financed” to “funded”. The term financed is often used to describe borrowing money and could lead to confusion. (I’ve used financed myself in the past, but don’t do this any longer.

4) In these later sections you sometimes describe the sources for funding consumption as income sources. You should drop the term income as saving is not an income source. )

Support System of the Indonesian Elderly:

Moving toward the Sustainable National Pension System

(National Development and Planning Agency, Bappenas)

Through Law No. 40/2006 on Social Security System in Indonesia (Jamsosnas), the Ministry of Finance issued Pension Fund Road Map that include improving the number of people registed into the pension program, improving the return pension coverage so that the elderly can rely on the pension program during their retirement period, and improving the management of the pension system in Indonesia. This paper investigates how the Indonesian elderly finance their retirement period without having sufficient pension program by specifically investigating how the support system differentiated by income level. To bridge the government plan and understand the implication of social security to the live of elderly, particulary those who are poor, this paper intensively discusses to what extent the existing support from the government reach the poor and fulfill their elderly consumption. For this purpose, we will use data developed by National Transfers Account (NTA) project lead by Lee and Mason (2006). The results show that the elderly finance their retirement age different by income level in which the poor elderly in both rural and urban areas rely heavily on public transfers. More importantly is that the poor elderly use this public cash transfers for supporting other household members.

1. 
   Background
   

Existing pension program covers only around 2.59 percent of total workefoces and majority are the government employees, army personnel and small coverage of fomal sector employees (Bapepam-LK, 2007). Lack of understanding the importance of pension program from both employees and private companies is one of main reason of the low enrollment rate of the pension program. In absence of pension support, there should be other means that the elderly can rely on to support for their retirement period.

Income level can be an important determinant of the elderly support system, particularly in the country without national coverage pension system. The elderly from low income families should hardly accumulate assets and need to find other means to support their consumption, such as extending their working period, relying government support, receiving support from other family members, or taking available private credits. Thus, this paper further analyzes the support system distinguishing by household income level to understand how the poor support their life and how far the government support their well being.

Data developed by National Transfers Account (NTA) project lead by Lee and Mason (2006) is used. Maliki (2008) has shown preliminary results of Indonesian National Transfers Account (NTA) for several years that are 1996, 1999, 2002, and 2005. He finds that the Indonesian elderly is characterized by longer period of working after retirement age, supporting their children or other family members, and depending on assets for their retirement.

This paper differentiated the support system by household income level and place of residents to bring several issues as a useful background for implementing comprehensive and sustainable social security system in Indonesia. These issues are; first is how the lowest income quartile elderly finance their consumption; second is how important self-employed income among other source of income for supporting the retirement period; third is how accumulated contributes to the retirement support. Two important issues are found from our analysis: first is public cash transfers are considered as an important source of income for elderly in poor households and second is that the non-poor, particularly in urban areas, rely more on the asset accumulation.

This paper is organized as follows; first, development of social security in Indonesia is discussed covering government regulation and long term planning program; second, description on public transfers and government program for the poor is presented to understand thoroughly on available assistance in supporting the elderly consumption; third is methodology that include data sources and estimation method; fourth is the support system discussion; and last is conclusion.

2. 
   Social Security Reform in Indonesia
   

The existing Social Security for Workers (Jamsostek) covers worker injury benefits, death benefits, retirement benefits, and healthcare benefits. In the implementation, Jamsostek only cover the formal sector workers, while the informal sector workers are still left out. The Jamsostek was not able to provide enough incentive for the workers to save through the program for their retirement because of considerably small benefits (Leechor, 1996). Not more than 20 percent of the population join the existing pension programs with considerably limited benefits.

The government of Indonesia attempts to increase the access of the pension program to all type of workers, health services, and protection for the vulnerable by issuing Law No. 40/2006 that regulates comprehensive coverage of social security system. The planned comprehensive social security system consists of 2 major programs; first is protecting the vulnerable by providing them minimum basic social services such as access to health services, people empowering program as well as the employment insurance. Second is establishing the insurance-based social security for the non-poor, both health insurance and pension program. In addition to the old age pension program, the individuals are needed to provide their retirement support by mandatory pension savings.

The old age pension takes a partial pay-as-you-go system that accumulates contributions for 15 years and starts to pay the benefits after the retirement age at 55. The formal workers are entitled for percentage of their income for the contribution, while the informal and self-employed workers are entitled for flat-rate contributions. The benefit paid is approximately 70 percent of the minimum wage. Widow and children respectively receive 40 percent and 60 percent of the minimum wage. Early retirement is compensated by the accumulated amount of the pension contributions with the returns in lump sum without monthly pension. The old age savings program is a retirement program in which participants are entitled to receive benefits before or upon reaching retirement age. The amount received is the accumulated amount plus returns.

3. 
   Data Sources
   

In addition to the core households’ information taken annually, Susenas also collect detail thematic informations every three years, such as consumption and income, education, health, etc. This three-year cycle Susenas is named Susenas-Module. In 2005, the Susenas-Module obtains detail information on consumption of the household, individual earnings, and other income of the households. These comprehensive information on consumption and income of individuals and households are important sources for constructing the National Transfers Account (NTA).

To adjust the information to national aggregate by type of household income, Financial Social Accounting Matrix (FSAM) or Sistem Nerasa Sosio-Ekonomi (SNSE) published by Central Bank and Central Statistical Bureau construct is utilized. FSAM 2005 published in 2008 has four types of household income and place of residents, which are poor-urban, poor-rural, non-poor-urban, and non-poor-rural. FSAM 2005 has information on consumption, production, and transfers that are useful for constructing the NTA variables.

Aggregate source of income from SNSEF 2005, consisting labor income, income on assets, and transfers, is illustrated in Table 1. NTA methods use labor income as earnings plus return for labor component of self-employed income. For this purpose, SNSEF 2005 counts the labor income by household income level and areas (rural and urban). As shown in Table 1, labor income is the most important income source of income for all types of households in both rural and urban areas. Labor income counts for 63 percent of total income, followed by income on assets and transfers.

Labor income of urban households are the largest and approximately 62 percent of total national labor income. Per capita labor income for poor urban areas is smaller than this of poor rural areas, while per capita labor for non-poor urban areas is as twice as larger than this of non-poor rural areas.

Households receive transfers, either from other households or from the government. Table 1 shows transfers received by households and does not consider transfers given by the households. Total transfers received by poor households in urban and rural areas contribute approximately 23-25 percent of their total income. While aggregate transfers for poor rural households are larger than those of poor urban households, per capita transfers received by poor rural households is smaller than this of poor urban households due to much larger poor population in rural areas. Transfers are almost equally important as income on asset as part of total income for poor rural households after labor income. On the other hand, the poor urban households depend more on income on assets rather than on the transfers. Further, non-poor urban households depend on transfers more than non-poor rural households do.

Note: USD 1 = Rp. 9500 ,-

1. 
   Income from assets is returns to assets (transfers for financial, non-financial institutions, and government (bonds and obligations) and part of non-labor income.
   
2. 
   Only include net transfers from/to other households, government, and ROW
   

Income on assets consists of return for assets as well as capital share of mixed income. Asset income is second largest income for both non-poor households, but not as large as transfers for poor rural households. Asset income contributes around 26 percent to the total income of non-poor households. Income on assets is also important part for the poor urban households and contributes around 38 percent of their total income.

Table 2 shows detailed transfers, both inflow and outflow from different type of sources. Households receive transfers either from government or from other households. Households receive transfers from government in forms of services and cash. Government services are considered as in-kind transfers, while cash transfers are government assistant in forms of scholarships for school children or any other direct cash transfers. The amount of in-kind services and cash transfers received by non-poor population are considered much larger than those received by the poor population. This is just because of larger non-poor population. But looking at the per capita figure, due to the methodology, all type of households receive almost similar per capita in-kind services, but receive different amount of per-capita cash transfers (Panel B of Tabel 2).

Not only poor households receive cash transfers from the government, but also non-poor households receive the benefits. The categories of poor and non-poor in this analysis is based on the standard used in Susenas determined by the Central Statistical Bureau (BPS Indonesia) that is also used by SNSEF 2005. The amount of sufficient food and non-food consumed per day per capita in the households is used as the criteria (SNSEF, 2005). Distribution of cash transfers or social assistance usually use additional criteria in addition to the above standard criteria. Thus, the government cash transfers program, eq. scholarships or other cash transfers program, is distributed not only to the poor indicated in this analisys but also to those who considered vulnerable to be poor that cannot be distinguished from our method. As shown in Table 2, as the government focuses on targetting the poor rural households for the social safety net programs or gasoline subsidy compensation, poor households in urban areas receive relatively less government cash than this of poor rural households, which is reflected by larger percapita net cash transfers.

Note:

1. 
   Cash received is estimated from cash transfers program and Susenas
   
2. 
   Income tax and estimated consumption tax
   

Scholarships and direct cash transfers are two main cash transfers program given to the poor as program of compensation of the increased price of gasoline in 2005. While scholarships are given to enrolled students from elementary school level to higher education, direct cash transfers or Bantuan Langsung Tunai (BLT) are cash given directly to the eligible households. Age profile of cash transfers recipients is estimated based on the household survey (Susenas) 2005. For this purpose, household head is assumed to be the recipient in the households. The head is assumed to redistribute the cash to support other household members consumption through intra-household transfers.

Not only households received transfers, households also give either cash or in—kind to other households or pay various taxes to government, such as income taxes, consumption taxes, and property taxes. As shown in Table 2, non-poor households in urban are net givers of government transfers, while the non-poor households in rural areas as well as the poor households in both areas are net receivers of government transfers. Per capita net government transfers received by poor household in rural areas are larger than per capita net govenrment transfers received by the non-poor households in the same areas.

Similar patterns occur for the inter-household transfers where the non-poor urban households are also net givers of the transfers, while other type of households are considered as net receivers. It is shown that only non-poor urban households transfer to the overseas, while poor households only receive small amount of cash from overseas. The poor households in both rural and urban areas barely transfer anything to the overseas and the non-poor households in urban areas give a small amount of transfers to the overseas.

National Transfer Account for Indonesia

The main components of National Transfer Accounts separately by income and residence level are reported in Table 3. The values are the per capita counterparts of aggregate controls used to construct To construct complete National Transfers Account (NTA). for all type of households, aggregate controls are required and it is shown in Table 3. SNSEF 2005, Flow of Fund, National Income 2005 and government financial statistics are main sources for constructing this aggregate controls. Based on this data, the support systems are estimated.

The NTA framework is quite different than the conventional framework presented in Tables 1 and 2. The conventional approach emphasizes sources of income. NTA places more emphasis on consumption and how it is funded. Labor, transfers, and assets are both income sources and consumption sources. Consumption also can be funded by dis-saving or by accumulating debt. NTA captures this source by introducing a new concept – asset-based reallocations. These consist of the combined flows generated by assets – asset income and saving. A second distinctive feature of NTA is that the basic unit of analysis is the individual rather than the household. For households private transfers are often relatively unimportant, but for individuals private transfers are enormous. Children depend heavily on intra-household transfers and, in some societies, the elderly may also depend heavily on intra-household transfers.

The values in Table 3 provide useful background information about how average values vary by residence and income. The emphasis in this paper, however, is not on averages but rather ison how economic flows vary by age and on the support system for the elderly.

Net public transfers are in-kind services and cash transfers received less taxes paid by the households. Public transfers inflow is estimated by sectors, while the tax age profile as transfers outflow is proxied by labor income and consumption profiles. Figure 1 shows per-capita net public transfers by household income level and their place of residents and net total government transfers. Net aggregate public transfers profile should follow the government transfers in Table 2, where the non-poor households, in general, are net givers and the poor households are net receivers.

Figure 1 also shows per capita inkind and cash transfers inflow. Inkind transfers are differentiated by health, education, and other sectors. Estimating transfers of education sectors to the recipients are based on the school enrollment reported in Susenas 2005 and transfers of health are based on the public health services utilization from the same source. The detail estimation methods are described in Maliki (2005). The other sectors are estimated by percapita (Mason and Lee et al. 2006).

The school age groups receive most of the government benefits. The government financial statistics show that the education spending is relatively large for both in-kind services and cash transfers program compare to health sectors (Ministry of Finance, 2005). Figure 1 illustrates that there two peaks occurs. The first peak occur because of the education services received by the school age children. The amount of the transfers are looked similar for all types of households. Health and other sectors transfers don’t shape the profiles significantly for all households’ type. The second peak occur because of the differences in cash transfers received by poor households. Poor rural households receive bigger cash transfers compare to other type of households that make this second peak is highest. The poor urban households also receive large cash transfers, while the non-poor households in both urban and rural receive much less than the poors.

Among the non-poor households, age groups of 20-62 show negative net public transfers meaning they pay taxes more than receive the public benefits. On the contrary, because household heads are assumed to receive all cash benefits from the government, high positive peak among the productive age group of the poor households is shown. High positive amount of net public transfers, even consist of large collective public goods received, means that the poor use public cash transfers as an important source of income.
Figure 1. Percapita public transfers Inflow and Percapita net total public transfers, Indonesia, in Thousand Rupiah, 2005

Figure 2. Percapita labor income relative to respective average production age 30-49, Indonesia, in Thousand Rupiah, 2005

Labor income

Figure 2 shows labor income by income level and place of residence. Labor income is normalized on the average labor income of the 30-49 age group for each of the four residence/income sub-populations. The normalization allows us to readily compare the shapes of the age profiles purged of differences in the average values across the four sub-populations.ts as relative to labor income of age 30-49. Figure 2 also shows the national average profiles of labor income relative to its average labor income of age 30-49 for comparison purposes. Labor income profile of all household income level is relatively similar, except for this of poor rural households. Labor income relative to their labor income of the prime age of non-poor urban and rural as well as poor urban households are relatively similar. Their peaks only vary from 1.08 to 1.1 of average production of prime age group (30-49). Ages of the peak for these profiles, however, are slighlty different between urban and rural. The non-poor rural households’ labor income reaches the peak from age of 42 to 45 years old, while the urban households are one year later than 45 years old and tend to have shorter period. Overall, their labor income profiles look similar with the shape of national average labor income.

The peak age of poor rural households’ labor income is at 45 years old with magnitude for about 1.18 of the average production of prime age group (30-49) of labor income of the same type of households. This makes the profiles are significantly different from the national average. The different start from early age, 20-39, where the profiles of poor rural households is slightly lower in magnitude with steeper profiles than any other profiles at the same age groups. The shape then deviates again from the national average at later age, starting from 55 years and older. The differences at the early age is mainly because differences at the earnings, while differences at the later age is mainly because of the self-employed income. Earnings at the earlier age of the poor rural households are slightly lower, while the self-employed income at later age tend to be higher than the national average. It shows that working opportunities for the poor youth in rural areas is lower than any other youth, but they have almost similar opportunity of self-employed labor with other type of youth. This shapes also clearly show that the elderly of poor rural households depend more on the self-employed income than any other types of households.

Differences of self-employed profiles between urban and rural are apparent. Those who live in rural areas have higher self-employed income contributing to their labor income compared to average national self-employed income profiles. Those who live in urban shows the opposite: their self-employed income is below the national average self-employed income profiles. Comparing poor and non-poor within place of residents, the self-employed income of the poor is always lower than this of the non-poor. It shows that the opportunities for working at self-employed business is more open for those who investment assets, which are the non-poor. The poor, particularly in rural areas, mainly work as the contract agriculture workers to cultivate the non-poor agriculture land.

The production of the elderly of 60 years old is relatively high, which is about 45 percent relative to prime age production. The production decline by age, in which for those who are 70 years old is only around 20-30 percent relative to prime age production. Unlike the non-poor households in urban areas, the elderly in rural areas either poor or non poor work more at the same age. Labor income profile of poor elderly in urban areas is also closer to non-poor households in the urban. Further, even though the different is not significant, the poor households have to work longer than their non-poor elderly counterpart.

4. 
   Support System for the Elderly
   

To focus on the support system of the Indonesian elderly, Figure 3 illustrates how much each income sources financefunds consumption of the elderly, 50 years and older, differentiated by household income level and their place of residents. Each funding source is expressed as a percentage of consumption. Consumption can be financefunded through 3 ways, which are labor income, transfers, and asset reallocation. Transfers consist of familial transfers and public transfers. Two types of familial transfers are made, which are intra household transfers and inter household transfers.

Panel A shows that labor income are main sources for all types of elderly. At age 50-54, labor income are still main source of income of the elderly for both poor and non-poor. The role of income declines substantially as the elderly getting older. At age 70 years and older, the labor income, in average, financefund around 20 percent of the elderly consumption.

In addition to the labor income, the poor elderly count on the public transfers as important source of income (Panel C). The 50-545 years old poor elderly in rural areas receive public transfers approximately 80 percent of their own consumption, and around 70 percent for the poor urban elderly. Even though the amount is still substantially large, net public transfers received decline as proportion of the consumption as the elderly getting older. The role of public transfers for the poor rural elderly is gradually replaced by small amount of familial transfers. The non-poor elderly at 50-59 years old in both rural and urban areas are net public transfers givers. The non-poor urban elderly is even net public givers until their late age of 65-69 years old.

The non-poor elderly reallocate their assets by dissaving for financing their consumption, paying taxes to the government, as well as supporting other member through familial transfers. On the other hand, the poor elderly accumulate assets from their labor income surplus as well as public transfers either to support their family members or paying other obligations, such as debt or even save the surplus. The definite use of accumulated assets among the poor elderly cannot be determined easily from the figure (Panel D) since the figure of assets reallocations are calculated as residuals and detailed components are not available at this time.

Figure 3. Support System of the Indonesian Elderly by Source Income, Income level, and Place of Residentceercentage s

5. 
   Conclusions
   

While the non-poor households in urban areas become net givers of government transfers, poor elderly in rural and urban areas use government transfers as an important source of income. The poor elderly use the accumulated income from public transfers to support other household member. Thus, not only the public transfers are important for the poor elderly, but the public transfers also are important for other age groups from poor families.

Three important findings can be summarized. First, public supports is important. Second, self-employed income is essential particulary in rural areas. Third, accumulated assets is alternative way for high income earners to support their retirement. High dependency of the poor, including to a certain degree the non-poor elderly in rural areas, to public transfers imply the importance of the social security program to suport the poor. Public transfers are even more important when the poor elderly transfer back the surplus to support consumption of other family members through intra household familial transfers or has to pay back the accumulated credit.

Self-employed income is important in supporting the elderly consumption at least until 65 years old. Self-employed income is an important source of income for middle to lower income households, particularly who work in agriculture field. To accomodate high varieties of self-employed workers requires flexibility of the pension system. For example, private institutions or associations could manage the pension system for the self-employed workers that operate at the same field, such as agriculture/farmer association, fisheries associations, or among other, Similar to health insurance system, the compulsory pension program requires thorough understanding on organizational issues and implementation because different nature of self-employed industries and formal industries.

To optimize the use of assets, particularly among the non-poor elderly, into more productive and less risky investment as well as to prepare their retirement period, financial education is urgently needed. The government need to provide guidelines for more systematic financial education programs, so that the workers understand how they invest their assets to get sufficient funds for their retirement.
Bibliography
Arifianto, Alex. (2004a). Social Security Reform in Indonesia: A Critical Analysis. Available at SSRN: http://ssrn.com/abstract=659821

Arifianto, Alex. (2004b). Social Security Reform in Indonesia:   An Analysis of the National Social Security Bill (RUU Jamsosnas). SMERU Working Paper.       

Bapepam-LK. 2006. 15 Tahun Dana Pensiun. Ministry of Finance, Indonesia.

Central Bank and Central Statistical Bureau. 2008. Sistem Neraca Sosial Ekonomi Finansial (Financial Social Accounting Matrix).

Leechor, C. (1996). Reforming Indonesia’s Pension System, Policy Research Working Paper, Washington DC: The World Bank: East Asia and Pacific Country Department III. Country Operation Division

Lieberman, S., M. Juwono, P. Marzoeki. (2001). Government Health Expenditures in Indonesia Through December 2000: An Update. World Bank Watching Brief, East Asia and the Pacific Region.

Maliki. (2008a). Health Card and Health Care Facilities Demand Among the Indonesian Elderly. The Singapore Economic Review, Vol. 53, No. 1, p. 103-119.

Maliki, 2008b. Supporting System Among the Elderly in Indonesia. NTA Working Paper.

Mason, Andrew, Lee R, Tung A, Lai M, Miller T, (2006). Population Aging and Intergenerational Transfers: Introducing Age into National Accounts, Economics of Aging Series, David Wise, ed. NBER and University of Chicago Press

Ministry of Finance. (2005). Laporan Keuangan Pemerintah Pusat (LKPP).

Trisnantoro, Laksono. 2008. “Evidence Based Policy Making Dalam Konteks Kebijakan Askeskin,” Kampus FK UGM, Forum Kepimpinan dalam Kesehatan.