The Impact of Economic Depression on Health Status in Indonesia 1. Hasbullah Thabrany School of Public Health, University of Indonesia

Transcription

1 The Impact of Economic Depression on Health Status in Indonesia 1 Hasbullah Thabrany School of Public Health, University of Indonesia 1 The Indonesian Economy A decade before Asian economic crisis that hardly hit Indonesia, the Indonesian economy has achieved a remarkable growth of 7% a year was a nightmare of the Indonesian economy. In the Middle of 1997 Indonesia suffered from severe drought and extensive forest fire lasted for several months and devastated million of rainforests. Due to severe drought Indonesian agricultural sector failed and Indonesia had to import rice and others. In July 1997, the Rupiah (Indonesian currency) began to collapse with falling exchange rate 30-40% to the US Dollar. The collapse of Rupiah continued and reached the peak between March and July 1998 when Rupiah plunged 600%. The price shocks associated with the collapse of Rupiah and removal of the Government subsidies for rice and other staple foods had reduced real income of the Indonesian people. During its economic booming, just before the crisis, the Indonesian economies relied heavily on imported materials. For example, more than 95% of raw materials for pharmaceutical products were imported. Whilst purchasing power of Indonesian was declining due to prices shock, many industries were unable to import raw materials and sell their products. In turn, this rendered mass laid-off. It was estimated that more than 5.5 million people lost their jobs in The government and international community began to offer financial assistance to recover the economy and to protect low-income citizens from adverse effects of this crisis. It was estimated that the number of people living below poverty line had increased from 27 million to 85 million 2. Many could not afford to buy foods and send their children to school. Indonesia had achieved significant progress to improve health status of its population. Significant improvements in economic development and serious efforts to reduce infant mortality rate have resulted in successful health improvement as shown in Figure-1. The Asian Development Bank, the World Bank, under the coordination of the International Monetary Fund (IMF) poured grants and loans for economic development and Social Safety Net (SSN) for lowincome earners. The social safety net program continues to be provided until The ADB provided with US$300 million loan, and will be added by another US$ 300, for health sector in eight provinces 3. The World Bank also provided Indonesia with loan for social safety net for the rest 19 provinces. The priority of this intervention was to protect the health status of the poor. To do that, the Government disbursed block grants money to more than 7,000 health centers in the country. The health centers may use the money to buy additional medicines and necessary medical supplies to provide health services for the poor. User charges for the poor were exempted. In addition, the Government also provided with medical services for high-risk pregnant women through midwives in each village. 1 Presented at the 2 nd World Conference: the International Health Economics Association, Rotterdam, the Netherlands, June 6-9, 1999

2 Economic crisis-health 2 H. Thabrany Figure-1: Per capita Income Growth and Improvement of Health Status in Indonesia IMR LE CDR Concerns about Health Status and the impact on the Access to Health Care Countries Population (1998, million) Table-1: Indonesia Among Its Neighbors, 1998 GNP per capita 1998 (PPP US. $) Infant Mortality rates (1998) Mat. Mort. Ratio (1990) Life exp. (1998).M/ F Health Exp: % GDP (1995) China 1, , / Philippines , / Indonesia , / Thailand , / Malaysia , / Singapore , / Source: World Health Organization, 1999 The Ministry of Health since the beginning of crisis has been aware that the falling Rupiahs might have severe impact on health. At the beginning of 1998, the Ministry of Health had been successful in obtaining Rp 700 billion for essential drug subsidies. The Ministry of Health predicted that the crisis would reduce household income, on the other hand the prices of drugs and medical supplies would increase due to almost 100% dependent on import, and the Government budget for health (in real terms would decrease. Those three direct impacts presumably would reduce health status and nutritional status of the population 4. A study by Hellen Keller International (HKI) reported that the evident of reduce nutritional consumption in

3 Economic crisis-health 3 H. Thabrany several provinces were obvious. In the island of Kalimantan, HKI reported that mothers and children stopped consuming milk and eggs due to much higher prices than they could afford 5. In addition, HKI also reported that in Central Java, one of the poor Provinces, the average body mass index among reproductive women has dropped by 0.45kg/m 2, almost equivalent the body mass increase in the last 30 years 6. It is reported that serious undernourished children under five in West Sumatera, has increased from 2,825 in March to 8,369 in May. 37 of them died due to severe undernourished 7. Media in Indonesia and oversees often exposed severe conditions of the impact of the crisis on health, employment, and education. The Government also reported severe effects of the crisis on social life, especially among the poor and newly poor. Several experts criticized that the publication of the crisis was exaggerations of the real conditions. While many people have certainly been impacted by the crisis, many others are actually benefited from the crisis. Farmers and other exported goods producers, for example, experienced big gains by the collapse of Rupiah. Shrimp and paper producers in 1998 received revenues in Rupiah that were three to five times larger than they had received in the previous year, due to much higher exchange rates of Rupiah to US $ and other foreign currencies. The economic depression coupled with political instability after Soeharto resigned in May 21, 1998 raised pessimistic expectation of Indonesian economy. In 1998, the Indonesian economy contracted as much as 15% below the 1997 level. Many economists predicted the recovery time for the crisis would last five to ten years. The Government will continue to provide social protection programs, using loan money, for at least until the year The problems are what kind of social protection programs are needed? How to identify groups villages or Districts that deserve assistance from the Government. Are those concerns happening as severe and as extensive as predicted? Those kinds of questions need to be answered. This study aims at answering some of that kind of question in health sector, specifically at health status among villagers. 3 Goals and Objectives To evaluate the impact of the economic depression on health status, prevalence of acute symptoms and the utilization of health services To obtain simple and noticeable indicators at village levels that can be used by decision-makers to determine appropriate assistance and necessary actions. To estimate the extent of changing of economic conditions and health conditions. 4 Data and Methodology Data for this study are taken from the 100 Villages Survey, a longitudinal survey conducted by the Central Bureau of Statistics. The survey initially aimed at monitoring poverty alleviation programs that the Government had implemented several years before the crisis. The initial survey was done in August 1997 in 100 villages in 10 Districts of 8 Provinces. For each village, 120 household were surveyed. In August and December 1998 the same households were resurveyed to obtain panel data on economic and socio-welfare conditions of the community. In 1997, 12,000 households were surveyed, in August ,999 households were survey, 8,000 of them were resurveyed of the previous round. And in December, about 97% of the 12,000 households surveyed in August 1998 were resurveyed. Some households moved to other villages when interviews were conducted. Due to high proportion of households in the first round were not resurveyed in the second and third round, this study focus on village level characteristics to maintain validity of information. The purpose of these data collections are to have snap shot on socio-economic conditions

4 Economic crisis-health 4 H. Thabrany over time, especially during the economic crisis (depression). The samples were selected on purpose to represent developed and underdeveloped villages, mountain and shore areas, urban and rural. Each village has 500 to 1,000 households. The sampling method of 120 households for each village was designed to represent the village conditions. The surveys were not designed to represent the whole community, rather it aimed at providing some clues for decision makers about trend in socio-economic variables. Initially, the main objective of this study is to identify certain population groups that hardly hit by the crisis. However, due to non-matched households surveyed between the first round and the second and third data collection, the change of the same people could not be studied. Therefore, this study looks at village level changes during the economic depression. The study is trying to relate some of tangible characteristics of villages such as informal sector employment and percentage of the head of households hold high school diploma with health status. Any significant correlation will be examined to see the appropriateness of using the village characteristics to predict significant changes in health status and further assistance needed. All related variables were aggregated to represent village conditions, for example, the proportion of respondent in a particular village who reported having chronic symptom and the average household expenditure for certain food or for health. Figure 2. Timing of surveys and the exchange rate of Rupiah to US $ Rp to US$ Round I Round 2 Round Jul-97 Sep-97 Nov-97 Jan-98 Mar-98 May-98 Jul-98 Month Sep-98 Nov-98 Jan-99 Mar-98 May-98 Health status is measured by several simple characteristics designed in the survey to obtain adequate responses from all community groups. Since most Indonesian are having relatively low education, the majority of them have only six years of schooling, this simple questions provide better validity. Respondents were ask of whether he/she and his/her dependent

5 Economic crisis-health 5 H. Thabrany were having any of ten symptoms during one month of the interview. The symptoms asked were: fever, cough, cold, asthma, short breath, diarrhea, measles, jaundice, headache, tooth cavities, and other symptoms. In addition, the respondents were also asked if they were experiencing any symptom, was there severe enough so that they could not work or do normal daily activities. To simplified analysis in this study, health status is measured by percentage of the population in a village experiencing at least one acute symptom, percentage of sample experiencing at least one chronic symptom, and percentage of respondent reporting non-performing normal activities due to the existence of a disease symptom. First analysis was descriptive statistics to examine general conditions of the village and general trend of changes between the three surveys time. The study estimate the effects of dependent variables, such as education, economic, household, demographic, etc., on health status. General linear models were performed to test variations on health status among 100 villages for the three panel surveys. The General Linear Models were also applied to examine how certain predictors affected changes between the second and the third data collection, and the first data collection. The following table explains the dependents and independents variables. Table 2: Dependent and Independent Variables Dependent variables % reported acute symptom % reported chronic symptom % reported severe symptoms in percentage reporting acute symptoms between August 98 and August 97 in percentage reporting acute symptoms between December 98 and August 97 in percentage reporting severe symptoms August 98 and August 97 in percentage reporting severe symptoms December 98 and August 97 Independent variables Aug98 panel (dummy) Dec98 panel (dummy) % children under five years of age % elderly population (65 +) % head of household 12 + years of schooling or more % non-formal work % household with private latrine % household with clean water supply Log average HH total expenditure Log average per capita HH expenditure Log average HH expenditure for protein source (meat, fish, eggs, and milk) 5 Findings The table shows due to collapse of Rupiah, people in the villages ought to expend more money to sustain their lives. Household total expenditure increased from the average of Rp 200 thousands Rupiah per month to Rp 356 thousands a year later and further increased to double three months later. The very high proportion of those expenditures went for food. Non-food expenditure also increased from Rp 62 thousand to 84.5 in August 98 and 87.6 thousand Rupiahs

6 Economic crisis-health 6 H. Thabrany in December 1998, a modest increased. Surprisingly, the tobacco and alcohol consumption also increased quite substantial in the level about twice expenditure for health care across survey time. Table 3: Description of Household Expenditure And Health Conditions Expenditure (current 000 Rupiah) 1997 Aug Dec Total Food Source of Protein (fish, meat, eggs, milk) Non food Health Tobacco/alcohol Share health/total (%) 3% 4% 3% Share health/non food 11% 18% 17% Health Conditions and treatments: % acute symptom % chronic symptom % severe symptom Point prevalence severe % using modern drugs % using traditional medicines Average costs of drugs (current - 5,843 4,389 Rp) Average costs of medicines (Rp) Use public providers Use private providers Other health facilities HH out of pocket Insurance paid Other support Average Out of pocket (Rp) - 22,399 20,895 The prevalence of acute symptoms increased quite substantial. In August 1998 the proportion of people having acute symptoms increased from 16% to 24% a year later and this rate remains in December The proportion of villagers who claimed having chronic symptoms also expanded from 6.4% to 8.1% but this proportion declined to the level of August 97 again during the third round survey. Point prevalence, the proportion of respondents having any symptom during interview, extended from 3.2 to 5.7% in August 98, and this proportion remain relatively higher in December Despite of the increased of disease prevalence, the proportion of people who sought treatment, both modern medicines and traditional medicines, relatively unchanged. Surprisingly, the proportion of respondent seeking medical care from private doctors or private hospitals grew from 2.3% to about 5%, both in August and December

7 Economic crisis-health 7 H. Thabrany On the other hand, the proportion of respondents seeking treatment from other health care facilities, such as nurses and midwives, reduced quite dramatically from around ten percent to only five percent. Table 4: Health And Nutrition Related s Variable 1997 (%) Aug 98 (%) Dec 98 (%) % HH with latrine % HH clean drinking water 74* % reduce food quality % reduce food quantity % report reduce income % reported medicines at home *) only 60% HH matched The table above indicates that very modest changes had been observed in the household facilities and consumption pattern between August and December However, about half of the households surveyed reported that they have reduced the quality of food consumed compared with the condition at the beginning of the crisis. The proportions of respondents reporting reduce income decreased from 20% in August to only 10% in December The latest two surveys indicated that the proportion of households providing medicines at home, for emergency use, remain the same at 12%. Aug 97 Table 5: s In Per Capita Expenditure: Current Prices Aug98 Aug98 Aug98- Aug97 Dec98- Aug97 % % Protein % 180% Tobacco % 165% Health % 128% Food % 121% Non-Food % 38% Total % 260% From the table above we can see that generally households had adjusted their consumption to meet the needs of the population. The increases of expenditures vary among different needs. The highest increase of expenditure in August 1998 was for health services on which the average households spent 121% more compared at the previous year. The lowest increase was for non-food items. Overall total household expenditures increase 73% compared at a year earlier. Four months later, total increase of food expenditure quite substantial with the highest increase for protein consumption, which increased 180%, compared at the base expenditure. Health expenditure on the other hand, was relatively constant. Non-food expenditure

8 Economic crisis-health 8 H. Thabrany remained the lowest increase. Table 6: s of HH Expenditures: Constant at August 1997 Prices Aug 97 Aug98 Aug98 Aug98- Aug97 Dec98- Aug97 % % Protein % 55% Tobacco % 47% Health % 27% Food % 23% Non-Food (2.981) (3.985) -17% -23% Total % 100% The above table presents changes in HH expenditures at constant August 1997 prices. The table shows that health care expenditures increased substantial, 38%. However, three months later, health expenditure was actually decreased significantly. Household began to rationalize their consumption by reducing purchase on non-food item by 17% compared with a year before. The following three month, this non-food expenditure decreased further. Households began to increase the quality of food consumed by increasing consumption of food items contained more protein by December Total HH expenditure increased modestly a year after the Rupiah collapsed, but three months later total expenditure began to increase substantially. Table 7: Regression Coefficients on Health Status Parameter Acute Symptom Chronic Symptom Severe symptom INTERCEPT Agust-98 (dummy) ** ** Dec-98 (dummy) ** ** JAVA ** ** ** URBAN % high school ** * % non-formal work ** * % less 5 years ** ** ** % more 65 years ** ** ** % have closet at home % have clean water * * * Ln HH expenditure Ln per capita exp ** ** Ln protein exp ** ** F value R-square **) significant at p <0.01; * significant at p<0.05

9 Economic crisis-health 9 H. Thabrany From tabale-7 we can observe that a year after the crisis began, the prevalence of people complaining at least one acute symptom or severe symptom increase significantly than the prevalence before the crisis. This prevalence became larger 15 months after the crisis emerged. However the prevalence of chronic symptom did not change significantly both in August 98 and in December People who live in Java consistently had lower prevalence of disease symptom than those who live outside Java. The proportion of head of household who held high school degree or above significantly correlated with lower prevalence of reported acute and severe symptom. The same pattern was also observed for the proportion of people in non-formal work such as farmer and self-employed. Both the very young age group (<5 years) and the elderly population (65+) had significant relationship with higher prevalence of both acute and chronic symptom. The regression had consistently awkward positive correlation between the proportion of household with clean water source in the house with higher prevalence of acute and chronic symptom. The same relationship was also detected on log per capita household total expenditure. The relationship between consumption of protein sources of food consistently associated with lower prevalence of acute and severe symptom. It did not correlate with higher chronic symptom. Table 8: Regression Coefficients on s of Health Status Parameter symptom Au98-Au97 symptom Dec98- Au97 severe Au98-Au97 severe Dec98- Au97 INTERCEPT JAVA * URBAN ** % high school % non formal ** worker % age < years * % age >65 years % have closet at * home % have clean * water Ln HH Exp Ln per capita exp ** ** ** Ln exp on protein * ** ** F value R-square **) Significant at p<0.01; *) significant at P<0.05 To see which group had higher change in health status, regression analyses on change in the proportion (the difference acute and severe symptom the second and third round and the first round) were performed. The Table-8 above showed that villages in Java experienced fewer changes in acute symptom only in December Urban villages had also higher changes in

10 Economic crisis-health 10 H. Thabrany severe symptom only in December The proportion of high school graduates did not significantly associated with more or fewer changes. The proportion of children under five year in villages correlated significantly with more prevalence in August 98. The same relationship was also detected for household having closet (toilet) at home. The proportion of household having clean water source at home correlated with change in acute symptom in August 98 but did not correlated in December Per capita expenditure in natural log transformation correlated positively with increasing prevalence of acute symptom both in August 98 and in December However, it relationship with higher prevalence severe symptom was significant only in December 98, in August 98 this relationship was not statistically significant. The consumption of high protein source of foods consistently associated with fewer proportion of acute and severe symptom. 6 Discussion Impact of non-random sampling The selection of villages was based on purposive sampling to represent certain conditions. In addition, the selection of villages was not aimed at representing the whole country. Therefore, the external validity of this study is not adequate to generate conclusion for the country. However, the changes of variables may be used as predictor for the villages. In other non-sampled villages, the findings may not be used cautiously. This study also examines whether there is a pattern of impacts across surveyed villages that may be used to predict those impacts in other villages. The characteristics of villages surveyed were not tested with characteristics of villages in the country. Effects of Crisis on Household Consumption Although there were concerns that households had severe impacts on their income and therefore expenditure would decrease, the study found that average total household expenditure increased 260% in December 1998 compared with the expenditure in August In 1997 constant prices, there were actually an increase of 100% in average per capita total expenditure in the villages studied. This study could not detect whether the increase of total expenditure was due to an increase in income or due to price increased that households were pushed to spend more. A study by Rand Corporation using the similar panel data from the Indonesian Family Life Survey found that there was decrease in per capita expenditure, mainly by those in upper parts of income distribution 8. The average per capita expenditure at village levels used in this study may explain the difference of expenditure pattern. In addition, this study may be biased toward rural villages, as the proportion of rural villages was only 20% in this study compared at 35% at the national level. Using constant prices, the largest increase of per capita expenditure for August 98 was for health, in which health expenditure increased by 38%, but this increase then declined in December of Compared at average the increase of drugs and medical supplies prices that ranged from % 9, this increase was vary low. A study by the National Institute of Health Research and Development found that health centers in East Java (that supposed not to increase prices) increased their prices by % 10. The Bureau of Central Statistics in Jakarta reported that the inflation rate for health during 1998 reached 80% compared with 1997 prices. So, even the average households in the villages surveyed spent 38% more on health, they may actually consumed less in the quantity or quality of health services. This might result in relatively still low health status (higher prevalence of acute and severe symptom) in three months later as shown in Table-3 and Table-7.

11 Economic crisis-health 11 H. Thabrany The important issue raised in this study is whether households had no money to spend on health or whether they did not place health as a priority. The study showed while household reduced their expenses for non-food item and smoking by August 1998, they spent more on tobacco both in August 98 and in December 1998 compared with their expenditure on health. In fact, the increase of tobacco expenses in constant prices was higher than the increase of health expenses between August 98 and December 98. In All three periods, households spent about twice more for tobacco than for health. It was unlikely that the relative low health expenditure between August 1998 and December 1998 was due to lack of money. It is more likely that households place higher priority for smoking than for health. The good news was that households spent much more for protein sources of foods such: meat, fish, eggs, and milk in December By December 1998, households in the villages surveyed seemed to have some recovery and showed rationing their expenditure by reducing non-food expenses and increasing protein consumption. The average expenditure for protein sources of foods increased by 55% in constant price compared to August 1997 level. This reaction may reduce the potential of losing generation due to poor nutrition as reported by HKI 11. Multivariate analyses showed that August 98 and December 98 had significantly higher prevalence of respondent reported having acute and severe symptoms than the period of August 97. However, for chronic conditions the test was not significant. This might be related with longer period for the development of chronic conditions. The December 1998 data showed small reduction of the proportion reported acute and severe symptoms. At the same time, at this time, consumption of nutritious foods as shown by the increase of 55% expenditure in constant 1997 prices. Accordingly, the average health expenditure also increased very significantly in August 98 but relatively reduced somewhat in December 98. Apparently, economic recovery began to pick up a little bit by the end of This findings consistent with macroeconomic indicators that have shown some improvements during the first semester of 1999, despite of political turmoil in Indonesia. Many analysts expect that the economic growth may pick up 1-2% this year. Villages in Java showed that a better performance in health status with consistently had fewer prevalence rates of acute, chronic, and severe symptoms. The change in health status, however, had been detected only significantly lower in Java for acute symptom in December 98 compared to August 97. It seemed that the effect of the collapse of Rupiah on health status was the same across villages in Java and outside Java. The urban-rural status also showed no significant changes, despite concerns that urban poor would be hardly hit by the crisis more than the rural villages. The result does not support a study by Poppele et.al. that reported urban population are hurting more 12. This study does not examine individual people living in urban area, rather it analyzes urban versus rural villages. The proportion of high school degree holders and non-formal workers had significant correlation with lower prevalence rates but had no significant difference in changes of the prevalence rates. It was not clear that non-formal workers had negative correlation with higher prevalence rates of acute and severe symptoms. Ages had significant association with the high prevalence rates. The higher the proportion of children under give and the higher the proportion of elderly the higher the proportion of people reported acute, chronic, and severe symptoms. However, the higher proportion of children under five years of age groups had only positive correlation with the increase of prevalence of severe symptom. These findings also showed that the impact of crisis on higher reported disease symptoms occurred on all age groups but greater severity on children under five years of age. This finding should strengthen the Government program on revitalization of integrated health posts (posyandus) across the country. Posyandu is a grass-root effort to improve nutritional status and immunization for children and pregnant mothers. The study found consistent negative of expenditure on high protein sources both in the

12 Economic crisis-health 12 H. Thabrany high prevalence rates of disease symptoms and high changes in health status. The higher the expenditure for protein the lower the prevalence rates and reduction of prevalence rates of disease symptoms. This finding indicates that there appear to be greater risk among the poor who usually consume less protein than those who are better off. This finding strengthen concerns raised by Hellen Keller International, especially among children and pregnant mothers. The positive correlation of per capita expenditure with higher prevalence rates and increase in prevalence rates might be due to perception biased. The prevalence rates were actually self-reported health status. It is a tendency that higher income (expenditure) highly correlates with higher education. Those who have higher education are usually have better awareness and perception about minor symptom than those who were in low education and low income. This pattern of correlation has been observed in many self-reported health status in the country such as in: Annual Soci0- economic survey and health household survey. 7 Conclusions and Recommendations The study use three panel data on 100 Village survey covering 12,000 households in Indonesia. Data collection were conducted in August 97, August 98, and December 98. Due to unmatched 40% households between the first round and the second round of data collection, the study use village level data as predictors to high or low prevalence of disease symptoms. Health status is measured in this study by using the proportion of people who reported having at least one acute symptom, one chronic symptom, and experiencing severe symptoms. There was significantly higher prevalence of disease symptom after the crisis than before the crisis. The high prevalence was positively correlates with Out side Java location, higher proportion of children under five, higher proportion of elderly, higher expenditure per capita, and lower expenditure for protein source of foods. The increase of prevalence rates was consistently related with the lower expenditure for high protein-contained foods. The study suggests that the poor might be hurting more seriously since usually they consumed lower protein. Despite the significant increase of prevalence rates, the study suggests that the health-impacts may be less dramatic than previously predicted, since there was indication that consumption of protein and food expenditure began to recover in December However, further evaluation on social impact of economic depressions needs to be continued. Revitalization of integrated health posts in the community is highly recommended to be continued until the sign of crisis has been disappeared. Acknowledgment The Graduate Programs, University of Indonesia has provided data for this study. The author also gratefully acknowledges the contribution of Mr. Hendro Hendratno for his sincere help in writing statistical programs. References 1) Employment Challenges of the Indonesian Economic Crisis, ILO-UNDP Report, ) Bureau of Planning, Ministry of Health. Estimation of Poor People, Jakarta, July ) Asian Development Bank. Health and Nutritional Sector Development Program (HNSDP): Policy Matrix, Manila, March ) Bureau of Planning Ministry of Health, Progress ReportL Managing Health Impact of the Economic Crisis, Indonesia. Jakarta December ) Hellen Keller International. Special Report, Nutrition and Health-Related Issues

13 Economic crisis-health 13 H. Thabrany resulting from Indonesian Crisis, September ) Hellen Keller International. Have 30 Years of Nutritional Improvement in South East Asia Disappeared in One Year of the Crisis?. Indonesia Crisis Bulletin, 1998, October, 1:4. 7) Kompas, June 1, ) Frankenberg, E; Thomas, D; and Beegle, K. The Real Costs of Indonesia s Economic Crisis: Preliminary Findings from the Indonesia Family Life Surveys. Rand, Santa Monica, USA, March ) Bureau of Planning, the Ministry of Health. Progress Report: Managing Health Impact of the Economic Crisis, Indonesia. Jakarta, December ) NIHRD. Effect of Monetary Crisis to Health Services at Puskesmas and Hospital in East Java, Jakarta, April ) Hellen Keller International. Alarming Rise of Iron Deficiency Anemia May Herald Lost Generation. Indonesia Crisis Bulletin, 1998 (1:3). 12) Poppele, J; Sumarto, S. and Pritchet, L. Social Impact of the Indonesian Crisis: New Data and Policy Implications. Mimeograph, January 1999.