DO REMITTANCES “ MAKE IT ” FOR THEIR MEMBERS LEFT BEHIND ? URBAN TO RURAL REMITTANCES BY MIGRANTS IN FOUR INDONESIAN CITIES

This paper examines the nature, extent, and effect of remittances sent by urban migrants to home villages. More specifi cally it examines: the magnitude of remittances, type and background of migrants sending remittance, the extent of remittances sent, and the use of remittances. This study is carried out in four major cities in Indonesia that have experienced signifi cant infl ows of rural–urban migrants over the past few decades: Tangerang in Banten, Samarinda in East Kalimantan, Medan in North Sumatra and Makassar in South Sulawesi. Using survey data collected in 2008-2009, this study found that, over time, rural-urban migrants have an opportunity to save portion of their income that subsequently send the remittances out to home village. Rural-urban migrants seem need time to be older and reach a certain level of income before they begin to send remittance. Statistically, age is among personal backgrounds that have signifi cant effect on propensity to send or not send remittances. One of important social dimensions of rural-urban migration in many developing countries is its effect on household members, particularly household members or children left behind. This does not seem to be the case in four cities under study, where many urban-rural migrants are accompanied by their members (wives and children).


Introduction
In the last decade the effect of international migration, especially remittances to home countries, has been the focus of some studies.In turn accurate data on international migration and remittances are available (World Bank, 2009a).International remittances are a small part of migrants' incomes that can be cushioned against income shock by migrants (World Bank, 2009b).Remittances are crucial elements of such household strategies to overcome livelihood diffi culties and to cope with basic need, not only for the migrants but for the household members they left behind as well.This is especially true when the household is really affected by crisis such as Asian economic crisis in 1997, like the increase of oil prices and global crisis in 2009.However, recently studies which examine internal migration led to information with regard to the effect of remittance sent by migrants to the household members left behind are limited.
The present paper, based on the data of voluntary rural-urban migration collected in four Indonesian cities survey 2009 in the context when global crisis started to have effects on Indonesian economy, is to examine the nature, extent, and effect of remittances sent by the urban migrants to their household members in home village.More specifi cally questions examined are: what is the magnitude of remittances, what is the type and background of migrants sending remittance, what is the extent of remittances sent, and what are the remittances used for.This paper is divided into seven sections and organized as follows.In fi rst section, we overview and highlight a few literature on migration and remittances.In the following section we describe four cities' setting.The next section, we discuss the effect of 2009 crisis on the incidence of remittance.In the fourth section we discuss urban migrants and remittances.The fi fth section describes the main demographic and social characteristics of migrants as well as the economic status of households involved in sending remittances.Then, we discuss the amount of remittances sent to the home villages.In the last section we focus on the utilization of remittances.
In this study remittance refers to cash and goods sent by urban migrants, both recent and lifetime migrants, to household members and families in home village.We defi ne migrants as people who lived continuously in a rural area for at least fi ve years before the age of 12, and currently reside in an urban area.Recent migrants are those who moved to a city in the fi ve years preceding the survey, and lifetime migrants are those who moved to a city more than fi ve years before survey undertaken.Nonmigrants are people who were born and raised in a city; therefore, this category includes urban to urban migrants.

Migration and Remittances
Recently many empirical studies have indicated that both internal and international migration remittances have a positive contribution to household welfare, nutrition, food, health and living condition in home villages.The extent of the contribution of international remittances is now well recognized in the literature on migration, remittances and development.This has been associated with increasingly accurate data on international migration and remittances to 194 developing countries since 2005 and outlook for remittances fl ow to developing countries for 2009-2011 that has been published by the World Bank (2009a and2009b).Human Development Report (2009:2) estimated that "approximately 740 million people are internal migrants--almost four times as many as those who have moved internationally--and most migrants both internal and international reap Qpqvmbtj Volume 23 Nomor 1 2015 gains in the form of higher incomes, better access to education health, and improved prospect for the children".
However, the contribution of remittances differs in different countries, communities and households (Human Development Report, 2009:71-860).Those depend on the extent remittances are remitted, where and how remittances are used for or spent.It also depends on the condition of social economic countries where the migrants exist and recipient of remittance countries and home village of households and families stay.For examples, Agunias (quoted in de Haas, 2007: 8) estimated that international remittances to Latin America countries accounted for 18% in Ecuador and 43% in Brazil. Another study (quoted in de Haas, 2007: 9) in Bangladesh and Nepal international remittances accounted at least half of total household incomes.A study by Nwajiuba (quoted in de Hass, 2007, 9) concluded that in Southeast Nigeria the contribution of those who migrate outside of the African continent may be up to 50% of household expenditure.All studies are contribution of international remittances on household incomes.However, the contributions of internal migration on the household income are lower than international migration.De Hass (2007:8) estimated that contribution of international migration on household income accounted for 33% while internal migration 10%.
In so far remittances spent largely for consumption and only in a small proportion use for productive activities led to the effect on communities are limited.However, Human Development Report (2009: 74) maintained "remittances use for consumption can be inherently valuable and often has long term, investment-like effects, especially in poor communities.Improvement in nutrition and other basic consumption items greatly enhance human capital and hence future income".Furthermore, results from a study using panel data and controlling for selection bias, examined the case of Indonesia between 1990-2000 found that "almost half of all households had an internal migrants, poverty rates for non-migrants were essentially stable for period (which included the Asian fi nancial crisis), falling slightly from 40 to 39 per cent, but declined rapidly for migrants from 34 to 19 percent" (Human Development Report, 2009: 73).

Four Cities Setting
This study is carried out in four major cities that have experienced signifi cant infl ows of rural-urban migrants over the past few decades: Tangerang in the province of Banten, Samarinda in East Kalimantan, Medan in North Sumatra and Makassar in South Sulawesi.In 2005 these cities covered about 33% of short and long-term migrants of rural origin in Indonesia2 .Two cities, Tangerang and Samarinda, became major urban settlements only quite recently along with the massive development of industries in Indonesia.The other two cities, Medan and Makassar, have a longer history as major urban areas since colonial period.
Tangerang is an industrial city at the outskirts of Jakarta with a population of around 1.5 million.After Batam in the Riau Islands, it is the city with the second-largest number of migrant workers in the manufacturing sector.Tangerang has rapidly grown since the 1980s in response to the twin effects of industrial growth to cater the demands for exports and the movement of Jakarta inhabitants to the Qpqvmbtj Volume 23 Nomor 1 2015 fringes of the city as urban congestion and rising land prices squeezed them out of the city centre.Samarinda is a medium-sized city with a much smaller population of around 600.000.This thriving provincial capital in the resource-abundant region of East Kalimantan is the largest destination for rural-urban migrants from other parts of Kalimantan.Both Tangerang and Samarinda expanded rapidly throughout the 1980s and the fi rst half of the 1990s, at rates above the national average (4-5 per cent per annum).The proportion of migrants (both male and female) engaged in manufacturing is higher in Tangerang than in Samarinda, refl ecting its importance as a major industrial centre.Its many factories are making textiles, garments, shoes, electronic goods, ceramics and other products that have attracted sizeable infl ows of migrants from countless remote rural regions over the past few decades (Warouw 2006).Although some of these factories closed down during the economic crisis in 1997, Tangerang is still an attractive destination for migrants in search of jobs.This is refl ected in its high population growth rates: during the period 1990-2000, Tangerang's population grew in average rate of 4.1 percent per annum, well above the national rate of 1.4 percent (Statistics Indonesia 2000: 161, 171).As in Tangerang, several factories have closed down in Samarinda in the past few years including at least fi ve large sawmills.The closure of these factories can be traced to the diffi cult economic conditions of the past few years, as well as the declining availability of timber (Forest Watch Indonesia 2001).In the second round survey of May 2009 we found that many of our respondents have moved out of Tangerang and Samarinda since they are no longer working due to closing down of fi rms as effect of global crisis.
The older cities of Medan and Makassar have grown at a more sedate pace over the past few decades.With a population of over 2 million, Medan is the largest city in the agriculturally rich province of North Sumatra.The region has been a centre of trade between the northern part of Sumatra, Singapore and Malaysia since colonial times.Medan was the main centre of manufacturing outside Java from the late 1960s through to the mid-1980s but since then its population growth has slowed.Medan, in the last decade, has experienced a reduction in fl ows of rural-urban migrants compare to two decades ago.The changes of employment opportunities in companies together with the slowdown of manufacturing activities due to the crisis economy in 1998 may not be a magnet for rural people from surroundings to look for jobs in Medan.Moreover, the increased price and demand of coconut oil, one of dominant agricultural activities in rural North Sumatera, widen up opportunities to more rural people to get benefi t from and that might be the cause that decelerate them move to urban areas.This has an effect on the current small number of migrants working in manufacturing industries in Medan.
The city of Makassar has a population of around 1 million.Although it lacks the strategic location of Medan, it is situated in one of the most densely populated areas of East Indonesian island provinces, South Sulawesi, and is the largest city in eastern Indonesia.In 1970s-1980s, Makassar city is the main destination area for migrants from other parts of eastern Indonesia, and has a relatively diversifi ed employment structures.Similar to Medan, Makassar has not experienced any signifi cant growth in its population for at least a decade.However, it stands out for the larger numbers of students who have moved to Makassar from surrounding confl ict areas (such as Maluku and Poso) and rural areas to continue their study.A signifi cant number of them were captured by the survey.Therefore, the proportion of recent migrants, lifetime migrants and non-migrants is relatively even in Qpqvmbtj Volume 23 Nomor 1 2015 Makassar and the two rapidly growing cities, Tangerang and Samarinda.

Global Crisis and Its Effect on Incidences of Remittances
Remittances data used in this paper are derived from the results of fi rst and second round survey of longitudinal (5 years) survey of Rural-Urban Migration in Indonesia (RUMiI).The fi rst survey was undertaken in April 2008 and the second survey was in May 2009 and they covered 2,417 and 2,080 urban households.In 2009, numbers of household interviewed have decreased 337 households or about 13.9%.Prior to the second round survey in May 2009, in January we done tracking in order to check whether household members interviewed in 2008 were still staying at the same address or not.In tracking we found that 183 respondents were already moved or back to home village (51.4%) for reasons they no longer work, had moved without any information (44.8%), refused admittance (3.8%).When we carried out survey in May 2009, following 4 months after tracking, there were 155 households who were not able to be interviewed owing to they had moved to other areas or back to home village (60.4%), moved devoid of any information (31.2%), and refused to be interviewed (8.4%).Those migrant households have moved to other areas or back to home villages.This indicates that they no longer work and have to looking for other job or temporarily stay in home villages waiting for other job.This is mostly the cause of manufacturing fi rms went out of business or had to reduce production activities since the 2009 global crisis affected Indonesian economy.In Tangerang, for example, this study found that many of recent migrants were working as wage labour in manufacturing based on contract labourers amounting to 33%.During 2009 crisis many of contract labourers were retrenched.In February 2009 Indonesian exports decreased about 32.26% and non-oil export decreased about 28% (Hadiz, 2009: 7).Many companies have retrenchment of employees.In June 2009, Ministry of Labour and Transmigration (www.nakertrans.go.id) reported that about 54,698 employees were dismissed and about 24,580 employees had sent home.Firms that dismissed and sent home their employees were electronic and timber manufactures.43.3% and in 2009 it decreased to 33.6%.These indicated that urban to rural remittances with a bit high proportion remittances have increased during crisis.In contrast, during crisis recent migrants that involved in sending remittances decreased.It seems that the crisis of 2009 have affected on recent migrants than lifetime migrants.In turn many recent migrants were not involved in sending remittance.
As indicated elsewhere that Tangerang as a centre of manufacturing activities in effect of crisis incidence of remittances have decreased, in 2008 incidence of remittances was about 57.8% and in 2009 it decreased to about 50.7% (see Table 1).In Samarinda, incidence of remittances had increased from about 16% in 2008 to 24.8% in 2009.Although in Samarinda, manufacturing activities reduced but coal mining exploration activities increased during crisis that may create other activities and it able to gain other opportunities for the migrants.Unlike younger cities, crisis in 2009 appear not much affected on the older cities.These conditions may have effects on the pattern incidence of remittances.In 2009, the pattern incidence of remittances in younger cities and older cities were different.In younger cities, in 2008 and 2009 incidence of remittances among recent migrants had decreased while lifetime migrants had increased.On the contrary, in older cities in 2008 and 2009 incidence of remittances among recent migrants had increased while lifetime migrants had decreased.

Urban-Rural Migrants and Remittances
Since the incidence of remittances between 2008 and 2009 is not much different, in analysing we focus on both recent and lifetime migrants that sent and do not sent remittances in 2009.In 2009 recent and lifetime migrants of 1220 households were interviewed.Of these, 351 or 28.8% migrants were involved in sending remittances to home village and 71.2% did not send remittances (see Table 1).From those 351 (28.8%) household heads that sent remittances to their home villages about 30.3% were recent migrants and 28% lifetime migrants (see Figure 2).As might be expected, a high proportion of remittances were sent out by household heads from younger cities, Tangerang and Samarinda, approximately 43.3% (see Table 2), while from older cities, Medan and Makassar, was just This rises a question why a high proportion of urban migrants do not send remittances to their home villages?This might be related to the fact that most of migrants of four cities are accompanied by their spouses and children in the city.One of important social dimensions of rural-urban migration in many developing countries is its effect on household members, particularly children left behind.However, this does not seem to be the case in four cities under study, where only a small number of household members are left behind.Household members of recent migrants and lifetime migrants of younger cities left behind in Tangerang are approximately 19% and 11.7 % and in Samarinda 7.2% and 14%.In fact in older cities recent and lifetime migrants household members left behind in a small proportion in Makassar was just 3.8% and 2% and in Medan the cases only 0.5% of lifetime migrants. 3ooking at migrants involved in sending remittances by migration status and city, we found that migrants of younger cities sent  and city it appears that in all cities education of migrants, both those who sent and did not send remittances, is not much different, on average it ranges 9-11 years, except in older cities (see Table 3).In older cities, recent migrants that did not send remittances their education on average were 14 years.As already discussed in the previous section that in older cities like Makassar many of recent migrants were still studying at the tertiary level and many of them were dependants of their parents.Table 2 also shows that in younger and older cities on average migrant household heads that sent remittances were younger than those who did not send remittances.Yet again exception for older cities, especially in Makassar, recent migrants that did not send remittances on average are a bit younger (25 years) than those who sent remittances.In all cities, on average household size of recent migrants that sent and did not send remittances is lower compared to lifetime migrants.For recent migrants it ranges1.9-2.7 persons while for lifetime migrants it ranges 3.4 -4.7 persons.
In older cities on average household size of recent migrants that did not send remittances is a quite lower than others.This was due to many recent migrants, especially in Makassar, were single and still studying in tertiary level.Lifetime migrants of younger cities that sent migrants on average household size were a bit lower compared to the others and this is because many of them were single (particularly in Tangerang).
In younger and older cities, recent and lifetime migrants that sent and did not send remittances on average the numbers of household members working are not much different.Excepting the older cities, numbers of working members of recent migrants that did not send remittances was just 0.5 on average.This is corresponding to the fact that many of recent migrants in older cities, especially in remittances, both recent migrants and lifetime migrants, a quite higher proportion amounting to 44.8% and 42.4% respectively and more than half do not sent remittances.In contrast, of recent migrants and lifetime migrants of older cities those sent remittances were just 10.3% and 15.6% respectively while more than two third of them did not sent remittances.A high proportion of recent migrant and lifetime migrant household heads from younger cities that sent remittances to home village compared to older cities might be related to the setting of the cities as discussed in the previous section that Tangerang and Samarinda are growing cities starting in 1980s along with the expansion of manufactures and the effect of mining (oil, coal) and timber activities.Migrants, both lifetime and recent migrants, can reap gains from the opportunities available there.Therefore, it is not surprising that about half of the recent and close to 53% of recent migrants and 49% lifetime migrants that sent remittances were working in manufacturing industry (see Table 2).In the last fi ve years, recent migrants have more access getting employed in manufacturing companies since they were young, single and educated (see discussion in the next section).Undoubtedly, many of migrant households are employed as factory workers. 4Such workers are generally having better salary and more secure than the workers in other sectors.This provides those workers an opportunity to save portion of their salary as sent remittances afterwards to home village.

Demographic and Social Characteristics
From comparison of migrants sending and not sending remittances by social and demographic characteristics, migration status, Qpqvmbtj Volume 23 Nomor 1 2015 Makasar, were young and single.In relation to gender, in younger cities recent migrants was quite different from lifetime migrants, since female of recent migrants quite high proportion (53%) were involved in sending remittances while lifetime migrant households was just 22%.Of the high proportion of young female, as discussed elsewhere, in younger city of Tangerang, many of young female of recent migrants were working in manufacturing.In contrast, in older cities those who sent remittances were females of both recent and lifetime migrants, working in manufacturing in a lower proportion of about 8% and 12%.It

Economic Status of Migrant Households
In this study household economic status refers to household expenditure per capita.Derived from annual household expenditure data, we divide household economic status into three categories, those are bottom 40%5 , middle 40% and upper 20% of total household expenditure equal to per capita per year.By applying these categories, the current study found that in younger cities, a high proportion (about 49.9%) of migrant middle household's economic status have sent remittances to home village while those who did not send remittances high proportion among recent migrants belong to bottom economic status of younger cities amounting to 43.2% (see Figure 3).In older cities, however, migrants that sent remittances appear not much different among bottom, middle and top household economic status: it accounted for more or less one third.Whereas, in older cities of migrants that did not send remittances a quite high proportion (46.2%) are middle incomes household.
Assessment of migrants that sent and did not send remittances by household economic status and city as presented in Table 4 appear that the highest proportion (51.8%) of those sending remittances were among lifetime migrants of middle economic status of younger cities.The next groups that have a higher percentage of sent remittances were recent migrants of middle and top household economic status of younger cities amounting to 41.6% and 47.5% respectively.As one might expect that in younger cities the highest percentages (52.7%) of those who do not sent remittances were lifetime migrants' those who of bottom household economic status.In older cities the discrepancy in sending remittances among recent and lifetime migrant household economic status of bottom, middle and top is not noticeable: more or less one third.However, migrants that did not send remittances a quite high proportion (57.4%) among lifetime migrants belong to middle household economic status group of older cities.
We have discussed descriptively variables of expected determinants of migrants who sent and did not send remittances.In this section, assessments through quantitative analysis is carried out in order to examine effects of demographic, social and economic variables on migrants that sent and not sent remittances.To accomplish the analysis we applied multinomial logistic regression.Dependent variables are migrants that sent and did not send remittances to their families in the home village.Independent variables include years of schooling, age, gender, marital status, household head occupation, household size, number of household members working, expenditure per capita/year, city, and migration status.The results estimation of regression is presented in Table 5.Table 5 shows that migration statuses variables do not give the impression to be effect on the propensity of migrants send or do not send remittances.Although, proportionally as discussed above the difference between migrants that sent and do not sent remittances appear signifi cant (see Figure 2) but are not statistically signifi cant.To explain recent and lifetime migrants sent and did not send remittances needs taking other variables into account.City setting, especially younger cities, seems to have a signifi cant effect on migrants that sent or did not send remittances and statistically the value of coeffi cient for Tangerang 1.807433 and Samarinda 0.8533276 are signifi cant at the level of 1%.As indicated in the previous section, of younger cities in term of industrial development are more developed, especially Tangerang, as a centre of industries.
Individual variables such as the age of migrants, both recent and lifetime migrants, has a negative coeffi cient (-0.0347717) and is signifi cant at the 1% level.This indicated that the likelihood of young recent and lifetime migrants to send remittances is greater than older migrants.This also means that increasing age of urban migrants is likely to reduce the likelihood to send remittances.This may be caused that beyond a certain age, the migrants no longer that need to support their families (parent or children) in the home villages.As one might expected that household size had effect on sending remittances, the results of regression shows that household size has signifi cant positive (0.710229) effect on sent remittances at the level 1%.However, since recent and lifetime migrants of four cities under study a high proportion accompanied by members (spouses and children), percentages  (send remittance==no is the base outcome/category) *significant at the 1% level Omitted categories: dummy for migration=lifetime migrant; dummy for years of schooling=no schooling; dummy for marital status=single; dummy for gender= male; dummy for expenditure=45-60jt; dummy for have house at village=no; dummy for occupation=others; dummy for location=Makassar.Source: Rural-Urban Migration in Indonesia, Second Round Survey, 2009 Qpqvmbtj Volume 23 Nomor 1 2015 of those who did not send remittances are higher than those sending remittances (Figure 2).Moreover, numbers of household members working statistically the value of coeffi cient 0.3733008 is signifi cant at the level of 1%.It is understandable that since more household members are working, migrant households are more likely to save some portion of their income in turn they have opportunities to send remittances to home village.Another variable that has infl uence on migrants in sending remittances is possession of a house in home village.Statistically it has positive coeffi cient (0.7007738) which is signifi cant at 1%. Undoubtedly this possession of house in home village requires migrants to go back to the village and possibly bring good or cash to their families.
To sum up, after taking into account some variables, it might be expected infl uenced on propensity send or not send remittances we found that personal variables that have statistically signifi cant infl uence is only age.Household variables that have statistically signifi cant infl uence on sending and not sending remittances are household size, number of household members working, and possession of a house in home villages which are signifi cant at the level of 1%.This suggests that recent and lifetime migrant households that send or do not send remittances appear to have been infl uenced by household than personal background.

Amount of Remittances
Data on average, the share proportion amount of remittances sent by migration status indicate that recent migrant households sent remittances approximately 8.6% while for lifetime migrants about 5.9% per year of total household expenditure (see Figure 4).In the previous discussion it is already indicated that many households that sent remittances were among middle household economic status.Consequently the capability in sending remittances to families in the home villages might be limited.It is not surprising therefore we found that on average the share proportion amount of remittances sent by all households, although there is a variation, overall it was less than 10% and it ranges from 4.5% -9.3% of total household expenditure a year (see Table 6).The lower proportion amount of remittances that was sent by lifetime migrants of older city Makassar in average is about 4.5% of total household expenditures.The share proportion amount of remittances sent by recent and lifetime migrants of older cities is lower compared to recent and lifetime migrants of younger cities.The highest is the share proportion amount of remittances sent by recent migrants from younger cities, especially of Tangerang, on average amounting to 9.3%.This fi gure is not much different from the result of a study done by Hugo (1978: 275) who found that urban permanent migrants in Jakarta sent remittances to the West Java villages on average accounted for 10.2% of total urban in come.Is there any different outcome if we analyse amount of remittances sent by migration status by household economic status?Data on Table 7 shows that, as one expected, the highest was sent by top of household economic status and next by middle and bottom, these amounting to 10.3%, 8.2% and 4.9%.Furthermore, the share amount of remittances on average the highest was sent by top household economic status of younger cities accounted for 10.7% and the lowest sent by lifetime migrant of bottom economic status of younger cities which is just 3.2%.From comparison of the share proportion amount of remittances sent by recent and lifetime migrants among bottom, middle and top of younger cities and older cities it appears that older cities is lower than younger cities.

Use of Remittances
This section discusses the use of remittances sent by internal voluntary rural-urban migrants who are living in four Indonesian cities.Data on the use of remittances, we collected from household head of rural-urban migrants that involved in sending by remittances, are not based on recipient of household members staying in the home villages.As indicated above that urban migrants, both recent and lifetime migrants who sent remittances to home villages, about less than one fi fth (16.9%) and on average the proportion amount of remittances sent it ranges about 3-10% of annual total households expenditure.This is hard to say that it is lower or higher but it can be attributed to their members or families staying in the village that need support from their members working in urban areas.This is parallel to the fact that a majority of remittances sent to home villages were utilized to meet basic need or daily consumption (see Figure 5).All households that have sent remittances 50.3% spent for daily consumption of household needs and parental care and medical cost come next in about 28.2%.Remittances spent for education accounted for 14% and traditional ceremonial cost only 3.85% and others 3.8% including house building (4 cases), agricultural activities (2 cases), regular social gathering (arisan) (1  (1978: 360) that urban permanent migrants remittances sent to the village more than half were used to purchase the staple subsistence food of rice and other food and the rest were spent for education, ceremonies, property investment and consumer goods.
The use of remittances breakdowns by migration status reveal that a high proportion of  remittances sent by recent migrants were used for daily consumption amounting to 55.3% and for parent health care and medical cost 27.2%, and education only 10.5% (Figure 6).The share use of remittances is not much different among recent and lifetime migrants, they also have a high proportion of remittances for daily consumption.The second large use of remittances is parent care and medical cost of household members: among recent migrants accounted for 28.7% and lifetime migrants 28.2%.The difference of the share use of remittances shows for education purpose, recent migrants the share about 10.5% whereas among lifetime migrants accounted for 15.7%.The lowest share of recent migrant household's in using of remittances for education might be because most of them are still single and young that have small numbers of children in schooling.
For the use of remittances there is no signifi cance change when we examine by household economic status as presents in Figure 7.It can be seen that more than half of recent migrants and lifetime migrants of bottom household economic status used remittances for daily consumption.On the contrary, remittances sent by middle household economic of recent migrants more than half were spent for daily consumption while for lifetime migrants about 44.6% were for daily consumption.Similar pattern also appears for the top household economic status.The use of remittances for parent care and medical cost shows a quite high percentage among bottom household economic status of lifetime migrants (33.9%) and middle recent migrants (32.6%).These suggest that many parents of migrants that stay in home village need support for medical cost from their members working in the city.The share use of remittances for education shows a quite high proportion among middle lifetime migrants (24.85) and recent and lifetime migrants of top household economic status accounted for 15.4% and 11.8%.The lower percentage of remittances used for education due to many of recent migrants are single, and for the married migrants many of their children, as indicated earlier, live together with their parents in the city.
Overall, assessment of the use of remittances by migration status and household economic status indicates that the utilization of urban migrant's remittances among bottom, middle and top of household economic status was to a great extent for daily consumption need oriented and parent care/medical cost, while only small portion of it was for education and business activities.The predominant use of remittances for daily consumption refl ects that families of migrants staying in home villages need support from their members working in city because of lack of income and may be live in poor condition.However, by the absence of rural survey it is impossible to evaluate household condition in home villages.This also has made it a limitation to analyze the effect of used remittances on social and economic of community.

Conclusion
This paper focus on remittances sent by rural-urban migrants; it especially examines the magnitude of remittances, type and background of migrants sending remittances, the extent of remittances sent, and the use of remittances.Incidence of remittances albeit in a small proportion had decreased during 2009 crisis.This decrease is a quite noticeable among recent migrants of younger cities.In contrast, the established migrants (lifetime migrants) of younger cities that sent remittances have increased in 2009.The city setting, especially Tangerang and Samarinda, have signifi cant effect on the propensity in sending remittances.The likelihood that urban migrant households would send remittances was greater among lifetime migrants rather than recent migrants.This suggests that, over time, rural-urban migrants have an opportunity to save portion of their income subsequently sent out of remittances to home village.Ruralurban migrants seem need time to be older and have to reach a certain level of income before they begin to send remittance.Statistically, age is among personal background that has signifi cant effect on propensity to send or not send remittances.However, the analysis indicates that among urban-rural migrants, both recent and lifetime migrants, a high proportion do not involved in sending remittances.One important social dimensions of rural-urban migration in many developing countries is its effect on household members, particularly household members or children left behind.This does not seem to be the case in four cities under study, where many urban-rural migrants are accompanied by their members.
The analysis indicates that migrants that send and do not send remittances are more determined by household variables such as household size, numbers of household members working, and possession of house in home village than personal variables.Only age of household head is personal variables that has signifi cant effect on sending or not sending remittances.This suggests that rural-urban migration may refl ect as one of strategies available to households in improving their livelihood.Given that situation, remittance is mainly used for daily consumption as well as household members care/medical cost, and only small portion is for education and business activities. 6This does not necessarily mean a useless use of remittances.Yet as improvement of quality of daily consumption in the case of bottom-middle households hence can improve nutrition and basic consumption items of household members of the migrants left behind.This in turn will be able to improve better health of household members whereby increase human quality (capital).
Notwithstanding the use of remittances may have or "make it" a positive contribution on health improvement for household members staying in the home village, but its effect on the village might vary, and it is hard to evaluate since the absence of rural survey.The urban questionnaire of this study was designed to obtain the purpose for which the amount and remittances were used.The data and information on amount of remittances sent by urban migrants may be able to provide the data accurately but they may not in position to give information on the use of remittances.It is possible that the information on the use of remittances are based on their perception, not on the actual made of the use of remittances.The absence of rural survey made it hard to evaluate to what extent the use of remittances has effect on the community.Similarly, it is not easy to evaluate the extent of effect of remittances on poverty, especially in receiving village.While the impact of urban to rural remittances on community may be vary, this study found that it seems has benefi cial livelihood impacts by strengthening human quality and compensating for lack of welfare.
What is the effect of those situations on incidence of remittances?Incidence of remittances by migrant households absolutely decreased.In 2008 numbers of migrant households were 1497 and of 406 involved in sending remittances.In 2009, numbers of household migrants were 1220 and of 351 involved in sending remittances.Proportionally incidences of remittances not much changed before crisis in 2008 was 27.1% and during crisis in 2009 it increased to about 28.7%.From those migrant households that have sent remittances to their home village in 2008 about 56.7% were lifetime migrants and it increased to 66.4% in 2009 (see sent by lifetime migrants both in 2008 and in 2009.This may be in consequence of migrants that need time to establish and have to reach a certain level of income before they initiated in sending remittances.(See, the established of lifetime migrants that sent Figure 1).In contrasts, in 2008 recent migrants sent remittances about

Figure 1
Figure 1 Percentage of Recent and Lifetime Migrant Households Send Remittances in 2008 and 2009

Figure 2
Figure 2 Percentage of Migrant Households That Send and do not Send Remittances to Their Home Village Figure 3 percentage of Migrants Hosueholds that Send and do not Send Remittances by Economic Status and City Figure 4 Average Proportion Amount of Remittances by Migration Status Figure 5 Percentages Use of Remittances

Figure 6
Figure 6 Percentages Use of Remittances by Migration Status

Table 1 Percentages of Recent and Lifetime Migrants that Send Remittance By City in 2008 and 2009
Source: Rural-Urban Migration in Indone2009irst and Second Round Survey, 2008 and2009Qpqvmbtj Volume 23 Nomor 1 2015 about 14.1%.Whereas, migrants that did not send remittances in younger cities were about 56.7% and 85.9% in older cities.

Table 2 Percentages of migrant Households That Send and
4 Ibid, p. 18

Table 4 Percentages of Recent and Lifetime Migrants That Send and Don't Send Remittances by Household Economic and City
QpqvmbtjVolume 23 Nomor 1 2015

Table 7 Average Proportion Amount of Remittance by Household Economic Status and City
Notes : ** Others, includes house building, agricultural activities, regular social gathering (arisan), and business activities, cemetery rebuild, pass away ceremonial Source: Rural-Urban Migration in Indonesia Second RoundSurvey, 2009

Figure 7 Percentages Use Remittances by Migration Status And Household Economic Status
Notes : ** Others, includes house building, agricultural activities, regular social gathering (arisan), and business activities, cemetery rebuild, pass away ceremonial Source: Rural-Urban Migration in Indonesia Second Round Survey, 2009 Qpqvmbtj Volume 23 Nomor 1 2015