The Asian economic crisis of 1997–98 had devastating effects on Indonesia's poor and near poor, the numbers of whom increased rapidly as a result. This development was particularly tragic since much progress had been made in poverty alleviation during thirty-two years of rapid and sustained economic growth under the New Order government (1966–98). To put adverse developments, particularly the rise in the incidence of poverty, in proper perspective, it might be helpful to have a brief historical overview of Indonesia's economic development and its impact on poverty during the three decades preceding the onset of the crisis and in the decade following it.

ECONOMIC GROWTH AND POVERTY ALLEVIATION DURING THE NEW ORDER 1966–98

As a result of political turmoil and the utter neglect of sound economic policies after the late 1950s, the Indonesian economy in the early 1960s experienced steadily rising inflation. As a consequence of the unrestrained printing of money to finance the rising government budget deficit, this spiraled into hyperinflation, which had reached almost 600 per cent by 1965. The economy had stagnated during the 1961–64 period; the modest growth that occurred in 1965 reflected only a good agricultural season (Hill 1996: 2). Since population growth exceeded economic growth in the early 1960s, per capita income declined during this period, particularly in 1962–63, with the economy contracting by three per cent in 1963 (World Bank 1998: 2.1).

With the deterioration in economic conditions, absolute poverty was quite high in the mid-1960s. In Java, 61 per cent of the population was very poor, while outside Java the very poor accounted for 52 per cent of the population. In 1961 no less than 68 per cent of the total population had no schooling at all, while only 0.1 per cent had enjoyed a tertiary education (Hill 1996:5). The available evidence also shows that per capita income in Indonesia in the mid-1960s was well below that of other Southeast Asian economies for which reliable data are available, and below even that of India. Indonesia's real per capita GDP in 1966 is estimated to have been only 535 international dollars (1985 prices), compared with US$650 in India (Booth 2000: 74).

INTRODUCTION

Mountain climbers scaling a sheer cliff face, understandably, want protection from falling. One method of protection is to place a safety net at the bottom of the cliff, to catch falling climbers just before they hit the ground. The alternative is a safety rope attached to a set of movable devices anchored at higher and higher levels as climbers ascend, so that falling climbers are caught after falling at most the length of the rope. The “safety net” guarantees against falls past an absolute level, while the “safety rope” guarantees against a fall more than a given distance, irrespective of the original height. For climbers near the bottom the safety net provides reassurance, but for climbers who have already made substantial progress scaling the cliff face, a safety net, which benefits them only after they have lost nearly all of their progress, is much less attractive than a safety rope.

The now ubiquitous metaphor of a “social safety net” conflates two distinct objectives in the design of transfer programs. One possible objective is to minimize a measure of income or expenditure poverty. An alternative objective is to mitigate risk, that is, reduce household vulnerability to the wide variety of potential adverse shocks that they could face (death, accident, fire, crop loss, job loss), whether or not the shocks push households below some absolute threshold. If program targeting is judged exclusively on static benefit (or poverty) incidence, then risk mitigation programs benefiting households that have suffered large shocks but are not “poor” may appear to have large “leakage” when in fact they are simply serving a different social objective — risk mitigation.

The undifferentiated metaphor of “safety net” also confuses thinking about the political economy of transfer programs. Governments may choose to implement “safety net” and “safety rope” programs for completely different reasons. While a “safety net” program might be more popular with the poor the more effectively it transfers wealth from richer to poorer households, a “safety rope” program will be more popular with middle-income groups if it serves an important insurance function in transferring resources from good times to bad times.

INTRODUCTION

The Indonesian experience of implementing social protection programs during the economic crisis of the late 1990s and also during the post-crisis period shows that targeting in programs of this kind is always difficult. As a consequence, social protection programs always suffer from the problems of undercoverage and leakage at the same time. Both problems cause these programs to become less effective and less efficient than they have potential to be (Sumarto et al. 2002). Hence there is a clear need to improve targeting at two levels simultaneously: the geographic and the individual levels. This paper describes attempts to develop more effective targeting tools than those already in use in Indonesia.

POVERTY MAPPING: A TOOL FOR BETTER GEOGRAPHIC TARGETING

The Advantages of Small-area Poverty Mapping

Ideally, geographic targeting should be based on a description of poverty incidence and other indicators of economic welfare in small areas or at low administrative levels. It is here that poverty mapping offers advantages. Detailed poverty maps of small areas can provide benefits to help address many of the shortcomings of aggregate poverty profiles and can greatly enhance and sharpen poverty analysis.

First, small-area poverty maps can obviously reveal the variations in local poverty levels. Almost all countries in the world have regions that are well off and others that have lagged behind. Such differences are often obscured in national-level statistics, a problem that is particularly critical in large and heterogeneous countries like Indonesia. Second, poverty maps can improve the targeting of interventions, which means that resources can be used more effectively. Poverty maps have the potential to reduce the risk that benefits may be leaked from a program to non-poor households. Similarly, they can also reduce the risk of undercoverage, that is, the possibility that poor households will be missed by a program.

Third, poverty maps can help governments to state their policy goals objectively. If allocation decisions are based on observed geographic poverty data rather than on subjective rankings of regions, the transparency and credibility of government decision-making is increased. Poverty maps can therefore help limit the influence of special interests in allocation decisions. This is particularly relevant in the context of currently decentralized Indonesia.

INTRODUCTION

The economic crisis that began in mid-1997 had an overwhelming social impact on the Indonesian population, whose living standards deteriorated as the krismon increasingly affected their incomes. In examining this social impact, the present chapter focuses on changes in real household consumption expenditures, as they reflect actual changes in living standards and form a measurable proxy for income changes due to the crisis. At the same time, the use of consumption expenditures enables the evolution of poverty to be examined by tracking changes in the headcount measure of poverty.

The data used in this chapter are from the consumption expenditures module of the 100 Village Survey (Survei Seratus Desa or SSD), which was sponsored by UNICEF and carried out by Statistics Indonesia (Badan Pusat Statistik or BPS) in May 1997, August 1998, and December 1998. The Survey covered one hundred villages located in ten administrative districts (kabupaten) spread across eight provinces. The present analysis is based on three rounds of the Survey.

THE 100 VILLAGE SURVEY

In December 1998, the 100 Village Survey collected data from 12,000 households, many of which had previously been surveyed in August 1998 and May 1997. It surveyed 120 households in each of the 100 villages in each round. The selection of households was somewhat complicated. In the May 1997 round, 120 households were chosen randomly from two enumeration areas within the villages. The general method in the August 1998 round was to add a new enumeration area. Forty new households were then chosen randomly from this new enumeration area. In the two enumeration areas used in the previous survey, eighty households were chosen to be re-interviewed in return visits to the same dwellings. If the households could not be identified at those dwellings, other households from the original 120 interviewed in 1997 were selected and added to keep the sample size at 120. This was the planned methodology but it appears that there were some deviations from this sampling procedure in the field: in some villages more than eighty households were matched and in other villages many fewer than eighty households were matched. Meanwhile, in December 1998 all 120 households from the August 1998 round were supposed to be re-interviewed. Unidentified households were replaced by new, randomly selected households.

INTRODUCTION

The present study is an attempt to piece together a consistent series of data from various sources on the headcount measure of absolute consumption expenditure poverty during the crisis. There are many broad issues in defining poverty, which is intrinsically a complex social construct. Even within a narrow definition of poverty based on a deficit of consumption expenditures, there are numerous thorny technical issues in setting an appropriate poverty line (Pradhan et al. 2001; Ravallion 1994; Sen 1981). This study avoids those issues and is limited to examining how poverty, defined on a consistent, welfare-comparable basis, changed in Indonesia over the course of the series of crises after the middle of 1997. We use a variety of data sets as well as various studies to put together a consistent series of the evolution of poverty over a six-year period from February 1996 to February 2002.

As the issues surrounding poverty measurements are complex, we begin with two basic issues. The first is the deflation of nominal to real expenditures to maintain comparability in welfare levels, while the second is the responsiveness of poverty rates to changes in real expenditures. With these basics in hand, we can estimate changes in headcount poverty rates over time, using a range of price deflators. Based on the outlined methods, we create a consistent set of estimates of poverty over the course of the crisis based on various data sets and studies that are available.

DEFINING REAL EXPENDITURES

The deflation of nominal to real expenditures is central to a “welfare-comparable” basis for comparisons of poverty over time. For any given distribution of expenditures across households, the determinant of the poverty rate is the poverty line. The poverty line is expressed in rupiah terms and is simply the amount of expenditure above which households are considered not poor and below which households are in (varying degrees of) poverty.

A fruitful way of thinking about deflation of the poverty line in nominal rupiah, so that it represents “the same” amount of “real” rupiah in another period, is to use the standard microeconomic theory of consumer choice with individual welfare maximization. The consumer choice problem is to choose a consumption basket for a given expenditure budget and prices so as to maximize their utility (Varian 1992, Chapter 7).

Before the onset of the economic crisis in 1997, Indonesia was one of the fastest growing economies in the world. During the three decades before the crisis, the economy had grown at an average of around 7 per cent annually. This rapid growth had generated an unprecedented reduction in poverty within a relatively short period of time. Between 1970 and 1996, absolute poverty fell by around 50 percentage points, accompanied by substantial gains in education and health outcomes.

In mid-1997, however, after nearly a quarter of a century of rapid growth and welfare gains, a currency crisis struck Indonesia and by early 1998 the country was suffering from the combined effects of financial, economic, and political crises. Within one year, the value of the rupiah fell by 85 per cent, domestic prices soared by 78 per cent, nominal food prices increased threefold, and the economy contracted by almost 14 per cent. As the crisis worsened, mass rioting occurred in the capital Jakarta and in a few other major cities, culminating in the May 1998 fall of Soeharto, who had been in power for three decades.

The social impact of the crisis was enormous. By one account, the poverty rate doubled between mid-1997 and the end of 1998, implying that an additional 36 million people were pushed into absolute poverty by the crisis. More than half of the increase in poverty between 1996 and 1999 was due to an increase in chronic poverty, as the proportion of the chronic poor within the total population tripled during this period. Similarly, the vulnerable, that is, non-poor households that have a high probability of falling below the poverty line, tripled in number between 1996 and 1999. The crisis primarily affected the poor and the vulnerable through falling real wages and a large increase in the prices of basic commodities.

To reduce the adverse social impact of the crisis, the Indonesian Government introduced Social Safety Net (Jaring Pengaman Sosial, or JPS) programs aimed at preventing the poor from falling more deeply into poverty and at reducing the exposure of vulnerable households to risks.

INTRODUCTION

At the onset of the Indonesian economic crisis, an important concern was raised over whether the achievements that had been made in the social sectors and poverty reduction over the previous decades could be sustained. Furthermore, there were some warnings about the looming social impacts of the crisis. This prompted the Indonesian Government to react rapidly and to institute a number of interventions aimed at safeguarding real incomes as well as access to social services for the poor.

To mitigate the social impact of the economic crisis, the Indonesian Government established a series of new and expanded programs known as the JPS (Jaring Pengaman Sosial, or Social Safety Net) programs. They were launched in early 1998, although many did not start until the second half of the year. It was hoped that through the implementation of these programs, the worst impacts of the crisis, such as widespread hunger, malnutrition, poverty, unemployment, and children dropping out of school, could be prevented or at least reduced.

This paper is an evaluation of how effective the various social safety net programs have been in reaching their intended target, namely, the traditionally poor and those newly poor due to the crisis. This is done by assessing the coverage of the programs among the poor as well as the way in which the benefits of the programs have been distributed between the poor and the non-poor.

THE SOCIAL SAFETY NET PROGRAMS

Indonesia's Social Safety Net Prior to the Crisis

The Indonesian people had never relied heavily on government-run safety net programs. The country has had neither the economic apparatus nor the political mechanisms necessary to deliver large-scale and widespread transfer programs. Instead, government social spending was largely focused towards “social services” such as health and education, while the family and communities provided “social insurance” in times of difficulty. There was some subsidized health care and a workers’ social security program, made compulsory for all formal sector employees by the 1992 law on Workers’ Social Security (McLeod 1993), but Indonesia did not have a social safety net system like the JPS. Establishing the social safety net programs in 1998 was, therefore, more like casting a new net rather than merely expanding an existing one.

INTRODUCTION

In response to the economic crisis of 1997–1998, the Indonesian Government introduced a series of poverty alleviation programs that together constituted the so-called Social Safety Net (JPS). But although the negative effects of the crisis on social welfare began to decline as the economy slowly recovered and macroeconomic stability was restored, the poor and vulnerable still very much needed assistance in the post-crisis period. Most of the components of the JPS programs were retained in the social protection programs that were adopted in and after 2000, although some underwent modifications in design. The JPS program itself was gradually phased out and by 2003 had ceased to exist in its original form.

POST-CRISIS SOCIAL PROTECTION INITIATIVES

Fuel Subsidy Reductions

In the year 2000 policy-makers launched the first phase in a lengthy process of reductions in the subsidy for fuel (gasoline, diesel fuel, and kerosene, known collectively as Bahan Bakar Minyak, BBM). The BBM subsidy, which had been increased in May 1998, had drastically reduced the cost of fuel products relative to international prices. The subsidy was highly regressive in absolute terms, however, as the majority of cumulative benefits accrued to non-poor households. It was only mildly progressive in relative terms in that BBM, particularly kerosene, comprised a slightly larger proportion of total expenditures among poorer households. The first subsidy reduction occurred near the end of 2000 and over the following years the price of BBM was gradually increased, but no government was willing to make adjustments on a scale that would have significantly altered this extremely regressive form of public spending. It was largely for reasons of political expediency that subsequent governments did not substantially reduce the subsidy on kerosene until several years later in the context of more dramatic overall price adjustments.

In October 2000 the Indonesian Government cut the fuel subsidy by approximately 12 per cent. It then reallocated Rp 800 billion in savings to a package of short-term compensatory programs known as PKPS-BBM (Program Kompensasi Pengurangan Subsidi Bahan Bakar Minyak or the Fuel Subsidy Reduction Compensation Program).

The printing press exerted great influence on the way religious authority was created and shaped. The cited quotation confirms what happened during the printing era in the Muslim community in the Netherlands Indies, in which the rise of various voices of Islam by various Muslim leaders — not merely the traditional ulama — became a leading feature. They engaged in a contest of interpreting and discoursing Islam. In Indonesia, the increase in the number of Muslims “who could claim to speak for Islam” began to occur in the early twentieth century. This proceeded alongside the rise and the development of Islamic reformism in an increasingly modernized colony as the consequence of the Dutch Ethical Policy of the period. And one of the most prominent features of reformism was the advance of the products of printing presses (journals, newspapers, and books) in social and intellectual live.

The use of the printing press meant that Muslims no longer had to turn to a recognized mufti only. In one issue of Islam Bergerak (10 June 1917), one of the newspapers published by Hadji Misbach (a leading leftist Muslim activist of the Sarekat Islam), there was a question from a reader about eating pork when one was starving and could not find any other food to eat.