Poverty and Social Impact Analysis
Indonesia rice tariff
Jennifer Leith
Catherine Porter
SMERU Institute
Peter Warr
March 2003
Background to the PSIA Studies
Poverty and Social Impact Analysis (PSIA) is an important feature of the new approach to
supporting poverty reduction in developing countries. PSIA is defined as the analysis of
intended and unintended consequences of policy interventions on the well-being or
welfare of different groups, with a special focus on the vulnerable and poor. Well-being or
welfare includes the income and non-income dimensions of poverty.1
The overarching objective of PSIA is to promote evidence-based policy choices, by
explicitly including poverty and social impacts in the analysis of policy reforms, and to
build country ownership of policies by informing a public debate on the trade-offs
between policy choices.
Analysing poverty and social impact is not new, but it has yet to be routinely applied to
macroeconomic and structural policy measures. In August 2000 the International
Monetary Fund (IMF) and World Bank agreed to consider the poverty and social impact
2
of major reforms in their lending programmes to developing countries. Increasingly,
developing country governments are initiating plans to undertake PSIA of key policy
measures as part of the process of refining their Poverty Reduction Strategies (PRSs).
In 2001, the UK Department for International Development (DFID) undertook to support
demonstration studies in six countries, in response to requests from governments and
other national stakeholders for ex ante analysis of the likely poverty and social impact of
particular policies or programmes. Six DFID-supported PSIA pilot studies were carried
out in Indonesia, Honduras, Armenia, Uganda, Rwanda, and Mozambique.
The World Bank also undertook to pilot PSIA in six countries. In October 2002, findings
from the DFID- and World Bank-supported pilot studies were brought together at a
workshop in Washington DC, hosted by the World Bank, the IMF and DFID. Key findings
of the workshop include that it is feasible to undertake PSIA using existing data and
knowledge in country, and that for PSIA to be effective in informing policy decisions, it
3
needs to be country-owned and embedded in the national PRS process.
The following report has been produced by independent researchers, and has been
independently peer reviewed. The analysis and views contained in the study are the
authors’ alone.
1
See World Bank, 2002, 'A User's Guide to Poverty and Social Impact Analysis,' available at
http://www.worldbank.org/psia and Robb, C, 2003, Poverty and Social Impact Analysis - Linking
Macroeconomic Policies to Poverty Outcomes. Summary of Early Experiences, Working Paper, IMF
Washington, DC.
2
See for example, IMF 'Key Features of IMF Poverty Reduction and Growth Facility', August 16, 2000.
3
See 'Poverty and Social Impact Analysis- Linking Policies to Poverty Outcomes'. Workshop Summary
Report, October 15-17, 2002. DFID/World Bank/IMF (available at http://www.worldbank.org/psia).
Disclaimer
This report is the work of independent researchers. It was commissioned by the
Government of Indonesia in collaboration with the Department for International
Development (DFID). The report does not necessarily represent either the views of
the Government of Indonesia or of the Department for International Development.
In its present form, the responsibility for any of the opinions expressed in this report rests
with the authors alone. Comments may be directed to:
Simon Hunt
PSIA Team Coordinator
Oxford Policy Management
6 St Aldates Courtyard
38 St Aldates
Oxford
OX1 1BN
Email: [email protected]
Contents
1
2
BACKGROUND TO THE INDONESIA PSIA ................................................................................. 4
1.1
PSIA in Indonesia................................................................................................... 4
1.2
Choice of Policy for Analysis .................................................................................. 4
1.3
Methodology and Report Structure ........................................................................ 6
INDONESIA CONTEXT: RICE TRADE, RICE TARIFFS AND POVERTY-LED POLICY
DECISION-MAKING .......................................................................................................................... 8
3
4
5
6
2.1
Indonesia Background: Crisis and Political Change .............................................. 8
2.2
Poverty Reduction and Macroeconomic Policy-Making......................................... 9
2.3
The Rice Trade in Indonesia ................................................................................ 10
2.4
The Debate: Rice Tariffs ...................................................................................... 11
2.5
Key Players in the Rice Tariffs Debate and What they Say ................................ 13
POVERTY IN INDONESIA.............................................................................................................. 19
3.1
Poverty Definitions ............................................................................................... 19
3.2
How Many are Poor ............................................................................................. 20
3.3
Rice Production in Indonesia ............................................................................... 24
3.4
The Importance of Rice for the Poor .................................................................... 24
3.5
Links between Rice Tariff and Poverty ................................................................ 24
CGE MODEL AND RESULTS ....................................................................................................... 26
4.1
The Case for a General Equilibrium Treatment ................................................... 26
4.2
The Wayang General Equilibrium Model ............................................................. 26
4.3
Results ................................................................................................................. 33
CRITICISM, VARIATIONS AND LIMITATIONS OF THE CGE MODEL................................... 36
5.1
Assumptions of the Model .................................................................................... 36
5.2
Timeframe of the CGE Model .............................................................................. 37
5.3
Dataset Used in the CGE ModelLing Exercise .................................................... 38
5.4
Effects of Varying Key Parameters ...................................................................... 38
5.5
Conclusions .......................................................................................................... 41
IMPLICATIONS AND RECOMMENDATIONS ............................................................................ 42
6.1
The Political Economy of Decision-Making .......................................................... 42
6.2
Methodology Process........................................................................................... 43
6.3
Implementation of PSIA—Demand ...................................................................... 44
6.4
Institutional Options for PSIA Implementation ..................................................... 46
6.5
Recommendations ............................................................................................... 46
Poverty and Social Impact Analysis: Indonesia Rice Tariff
FIGURES
Figure 6.1
Figure 6.2
Figure 6.3
Figure 6.4
Figure 6.5
Real Price of Rice, Indonesia, 1969 to 2001 ................................................. 47
World Price and Domestic Price of Rice, Indonesia, 1985 to 2002............... 48
Simulated Changes in Poverty Incidence: Varying Elasticity of Import
Supply of Rice ............................................................................................... 48
Simulated Changes in Poverty Incidence: Varying Elasticity of
Substitution in Rice Production ..................................................................... 49
Simulated Changes in Poverty Incidence: Varying Armington Elasticity of
Substitution in Rice Demand ......................................................................... 49
TABLES
Table 3.1
Table 3.2
Table 3.3
Table 3.4
Table 3.5
Table 3.6
Table 3.7
Table 4.1
Table 6.1
Table 6.2
Table 6.3
Table 6.4
Table 6.5
Table 6.6
Table 6.7
Table 6.8
Table 6.9
Table 6.10
2
The Population Classified as Poor in Indonesia (1981 – 2000) ................... 20
Changes in the Mean of Nominal and Real Expenditures by Quintile .......... 21
Human Poverty Index .................................................................................... 21
Human Development Index ........................................................................... 22
Poverty Incidence and Human Development Index ...................................... 22
Percentage Changes in Inequality Indices between 1996 and 1999
Based on Per Capita Real Expenditure ......................................................... 23
Poverty Incidence and Contribution to Total Poor by Main Sector of
Occupation, February 1996 and February 1999 (%) ..................................... 23
Poverty Rates by Household Type, Before and After Tariff Rise .................. 35
Indonesia: Rice Production, Consumption and Trade, 1990 to 2001 ........... 50
World Rice Trade by Country ........................................................................ 50
Expenditure and Poverty Incidence by Household Group ............................ 51
Factor Ownership of the Broad Household Groups ...................................... 52
Simulated Macroeconomic Effects of a Rice Tariff: Varying Rice Import
Supply Elasticity (per cent change) ............................................................... 53
Simulated Distributional Effects of a Rice Tariff: Varying Rice Import
Supply Elasticity ............................................................................................ 54
Simulated Macroeconomic Effects of a Rice Tariff: Varying Elasticity of
Substitution in Paddy Production (per cent change) ..................................... 55
Simulated Distributional Effects of a Rice Tariff: Varying Elasticity of
Substitution in Paddy Production .................................................................. 56
Simulated Macroeconomic Effects of a Rice Tariff: Varying Armington
Elasticities in Rice Demand (per cent change).............................................. 57
Simulated Distributional Effects of a Rice Tariff: Varying Armington
Elasticities in Rice Demand ........................................................................... 58
Abbreviations
BAPPENAS
BKPK
BPS
CES
CGE
CGI
DFID
HDI
HIPC
HPI
I-PRSP
IMF
KPK
NGO
PIM
PRGF
PRS
PRSC
PRSP
PSIA
Rp
SAM
Susenas
UNDP
USAID
National Planning Board
Poverty Reduction Cooperation Board
Statistics Indonesia
constant elasticity of substitution
computable general equilibrium
Consultative Group on Indonesia
Department for International Development
human development index
Highly Indebted Poor Countries
human poverty index
Interim Poverty Reduction Strategy Paper
International Monetary Fund
Poverty Reduction Committee
non-governmental organisation
policy interests matrix
Poverty Reduction and Growth Facility
poverty reduction strategy
Poverty Reduction Support Credit
Poverty Reduction Strategy Paper
Poverty and Social Impact Analysis
Rupiah
social accounting matrix
National Socio-Economic Survey
United Nations Development Programme
United States Agency for International Development
3
Poverty and Social Impact Analysis: Indonesia Rice Tariff
1 Background to the Indonesia PSIA
1.1
PSIA IN INDONESIA
1.
In August 2002 a scoping mission team, with members from SMERU (a Jakartabased economic think tank) and three staff from DFID London, spent a week in Jakarta
meeting with key stakeholders in the Indonesia policy-making process. This was to build
interest and support for PSIA and to identify a current policy issue where a PSIA might
usefully inform government on poverty outcomes of a proposed policy. The DFID Jakarta
office had earlier laid groundwork to build interest in the PSIA.
2.
In attempts to link PSIA with the emerging Indonesian Interim Poverty Reduction
Strategy Paper (I-PRSP), the scoping team met with most of the senior decision-makers
from government who have been engaged with the production of the I-PRSP and
representatives from civil society concerned with poverty.
3.
The users of the PSIA are expected to be those connected to the production of the
PRSP and those in civil society with an interest in the policy being analysed. Interest in
the PSIA as an emerging methodology for policy making so far remains donor-driven.
However, the potential for PSIA to lead to more transparent decision-making and wide
dissemination of its outputs that could assist civil society groups to monitor policy
decisions was also investigated.
1.2
CHOICE OF POLICY FOR ANALYSIS
4.
After consultation over the course of a week, the policy chosen for analysis was
rice trade policy (import tariffs / pricing issues). This issue emerged strongly in our
discussions with the various stakeholders including government, civil society and the
World Bank.
5.
The government expressed an interest in raising tariffs on imported rice. Bulog (the
state commodities logistical agency) advocates increasing the import tax on rice from the
current Rp 450/kg to Rp 750/kg, equivalent to moving from 25% to 50% ad valorem.
6.
Simplified, there are currently two main opposing arguments for and against rice
import tariffs:


some support a high rice tariff policy since higher prices are associated with
higher incomes for farmers;
others believe that poor people are net rice consumers or buyers, who will suffer
due to high rice prices.
Rice Tariff Policy Background
7.
In 1998, under structural agreements with the IMF, Indonesia began a market
liberalisation process, which removed Bulog’s monopoly position with food commodities.
Legislation in June 1998 allowed private enterprises to compete with Bulog in importing
and marketing of several commodities, including rice. On 22 September 1998 rice
imports were freed with 0% tariff. On 01 January 2000 the Ministry of Trade began
4
imposing tariff on rice imports of Rp 430/kg (equivalent then to 30% ad valorem tariff).4 In
September 2000, based on Bulog’s recommendation, the Directorate General of
Customs and Excise put red lane inspection on rice imports in place5.
8.
Recently Bulog announced its plan to implement a quota-tariff to maintain the
domestic rice price through import limitation6. The resulting controversy became even
more marked when the Tariff Team of the Ministry of Agriculture proposed a further rice
import tariff increase to the National Tariff Team of the Ministry of Finance. The Ministry
of Agriculture is also suggesting that only selected importers will be allowed to import rice
to protect Indonesian farmer incomes7.
9.
Rice policy is an emotive issue in Indonesia and rice issues are constantly headline
news. Recently, ‘on behalf of rice farmers’, the governor of South Sulawesi refused to
have 100,000 tons of imported rice unloaded in that province8. A similar action was taken
by the East Java provincial government, which issued a regulation to ban rice imports to
protect farmers’ incomes. A PSIA on rice tariff policy is timely as decisions are soon to be
made in terms of specific rice tariff policy.
Existing Data
10. Considerable data and reports on the issue of rice trade policies over the last 30
years already exists, much of which appears contradictory. Timmer’s work on rice policy
and rice self-sufficiency in Indonesia has been prominent. More recent work has focused
on the costs to the poor of higher rice prices as a result of the increase in rice tariffs to
30% in 2000 (which also occurred in the context of rapidly falling world rice prices). Both
Ikhsan at University of Indonesia and additionally the Centre for Agro-Sociological
Research (Bogor) have done work showing that the poor are net rice consumers, and so
have suffered as a result of the tariff rise.
11. This research is well summarised in a recent BAPPENAS (national planning board)
report, 'An Approach to Macro Food Policy' (March 2001, with the support of the United
States Agency for International Development (USAID)) which also argues
that Indonesia’s rice productivity is already very high (eg. above Thailand) and that there
are few linkages between rural wage rates and the rural economy with higher rice
prices. Thus, according to the BAPPENAS paper, these justifications for higher rice
prices are spurious. Indeed farm labour rates are slow to change and already there is
extensive non-rice diversification. However, HS Dillon and other Indonesian researchers
challenge this approach. They argue that rice prices are still key to the rural economy
and that high prices would increase rural agricultural wages in the long term.
12. It was expected that developing a computable general equilibrium (CGE) model of
the Indonesian economy, in order to model the impact of increasing rice import tariff from
4
M. Husen Sawit, et al., BULOG: Pergulatan dalam Pemantapan Peranan dan Penyesuaian
Kelembagaan [BULOG: a Struggle in the Role Establishment and Institutional Adjustments], (2002), p.
420.
5
'Jalur Merah Bagi Impor Beras Sudah Berlaku' [The Red Lane Inspection on Rice Import is in Effect],
www.kompas.co.id September 25, 2000
6
Kompas, July 26, 2002
7
Kompas, August 20, 2002, Media Indonesia, August 18,2002.
8
Bisnis Indonesia, July 31, 2002
5
Poverty and Social Impact Analysis: Indonesia Rice Tariff
25% to 45%, could contribute to understanding and anticipating the longer-term issues in
rice pricing and its effects on poor people.
Rationale
13. Examining the rice tariff question was well supported, despite the presence of
considerable research on the topic already, because:





existing research is contradictory and there are varying perspectives on the
impact on the poor of raising rice import tariffs and prices. No previous research
using a CGE model in Indonesia had yet looked at second-stage effects of an
increase in rice import tariffs;
the topic is timely with potentially great impacts on poor people;
it was the right size of a question in this technical exercise which could be done in
the short time-frame allowed in the pilot;
SMERU had identified an international economist who had done some work
already on a general equilibrium model of the Indonesia economy from which we
could examine the long-term effects of increasing rice tariffs;
other existing qualitative data was available on the impact of increased rice prices
on various groups of poor people during the recent Asia financial crisis.
14. These factors were felt to be important enough to choose rice tariff policy over
other policies that emerged through our discussions. Decentralisation was also raised as
a pressing issue by some but the team felt that it was too big a question to examine with
such a short period of time available.
1.3
METHODOLOGY AND REPORT STRUCTURE
15. The PSIA methodology recommended that two national consultants (an economist
and a sociologist) and two international consultants (an economist and a sociologist) be
engaged for 40 days. SMERU fielded two to six staff members for various periods
throughout the exercise. The international economist who developed the CGE model
could participate for only 10 days. DFID London provided in-house expertise with a
sociologist, who was also team leader, and an economist who could add to the modelling
work.
16. It was expected that a CGE model would allow an analysis of both first and second
round effects of a change in the tariff (ie. on prices, and on rural wages/employment),
and their relative magnitudes, to provide an assessment of the net impact of the change
on welfare and especially on the poor. Existing data sets from 1995 were used and
augmented by further published data sets acquired by SMERU. Indonesia has good data,
statistics and research capabilities, unlike many other countries engaged in PSIA.
However, data from the 2000 census was net yet available.
17. Where PSIA can make a contribution in a poor policy-making climate is to link
existing research and data with a transparent and inclusive process for policy decisions.
With the CGE model at the heart of the PSIA, a further qualitative method—the policy
interests matrix (PIM)—was developed to address the poverty impacts of the rice tariff
policy issue, focusing on mapping the policy decision-making process and the interests of
6
key stakeholders in the process. In contrast to the CGE model, the PIM reveals the
underlying politics of policy decision making.
18. Key stakeholder perspectives were gained from interviews and existing
documentation, reports and published statements and formed the basis of the policy
interests matrix. Civil society contributed their assessments of key actors’ perspectives
on the rice tariff that were included in the policy mapping. There was some recognition of
the sensitive of mapping departmental self-interest, which will be discussed later in the
report.
19. A workshop was held at the end of the PSIA exercise to provide a forum for
discussion of the results of the CGE modelling and the overall analysis of the PSIA
outputs, and also to enable others actors with perspectives on rice tariff pricing to
participate. This was well attended by academics, and donors but fewer from government
and civil society attended than originally had agreed to participate.
20. Since the workshop held in mid-October 2002, no announcement has been made
of a rice tariff increase.
21.
The report is divided into six chapters:
22. Chapter 2 provides very brief background on the Indonesia rice trade and rice
tariffs issues and experience with poverty led policy decision-making. It identifies the
economic logic associated with both pro- and anti-tariff increases. It reviews the positions
of key actors, both individuals and institutions, on the rice tariff debate.
23. Chapter 3 highlights data on the situation of poverty in Indonesia in the last five
years with information focused on the effects of the Asia Financial Crisis on poverty.
Some focus is given to consumption issues and the provision of safety nets in the form of
cheap rice for poor people. This chapter also provides descriptive experiences of poor
people and their agricultural livelihoods as a result of the shocks associated with the
financial crisis. This is intended to provide the context for the next chapters which models
the potential effects of an increase in rice tariffs on the well being of various categories of
people in Indonesia, both rural and urban.
24. Chapter 4 explains the CGE model, its structure and assumptions, and provides the
result of the analysis. This chapter is a limited technical discussion, with more detail
provided in the appendix.
25. Chapter 5 critically discusses the model and its findings, including variations and
limitations of the CGE analysis. This includes comments raised by DFID economists,
academics and researchers both in Indonesia, and in Washington, who attended the joint
DFID–World Bank PSIA workshop in late October 2002.
26. Chapter 6 takes the results of the PSIA and puts it into the policy-making context in
Indonesia. The results indicate that raising the tariff increases poverty in the majority of
income and livelihoods categories, though the effects are small, and some rural groups
do benefit. The chapter provides thoughts and comments on how the PSIA results can be
used in pro-poor policy making. Some discussion is provided on the PSIA process itself,
and thoughts for PSIA’s use in the future.
7
2 Indonesia Context: Rice Trade, Rice Tariffs and
Poverty-Led Policy Decision-Making
2.1
INDONESIA BACKGROUND: CRISIS AND POLITICAL CHANGE
27. The last five years in Indonesia have been a tumultuous period economically and
politically. During mid-1997, Indonesia became engulfed in the regional economic crisis
that also affected other South East Asian countries. The Indonesian currency collapsed
from near Rp 3,000 per US dollar in the second half of 1997 to Rp 15,000 by mid-1998.
Inflation skyrocketed9. In the domestic political crisis that followed, massive riots in May
1998 in the capital Jakarta led to further political instability. With the murders of students
in street demonstrations protesting the effects of the economic crisis, Suharto, who had
been the president for more than 32 years since 1965, was forced to step down.
Democracy is now emerging slowly in Indonesia, with the current president, Megawati
Sukarnoputri, the third since the end of the Suharto era in 1998.
28. The financial and political crises of 1997 and 1998 led to a contraction of GDP by
13.7% in 1998. The social impact of the crisis, particularly in terms of poverty, was
immediate and substantial (see also Chapter 3). The pre-crisis poverty rates of almost
11% in the second half of 1997 (according to data from phase 2 of the Indonesian Family
Life Survey) increased more than two-fold to 27% by February 199910. Although the
poverty rate appears to have improved since February 1999, poverty is still higher than at
its pre-crisis level. Currently (September 2002) the Indonesian rupiah remains at just
under 9,000 to the US dollar, a significant devaluation from 1997 levels of Rp 2,500 to
the US dollar11.
29. Indonesia today is considered to be 'post crisis', but the current levels of
international debt remain high at US$150 billion, with debt servicing representing more
than 40% of government revenues for several years to come (Tabor 2001). Debt
payments have recently been rescheduled. The advice given to Indonesia during the
crisis by the IMF has come under considerable criticism in Indonesia and been blamed
for ongoing poor economic performance. Though further loans have been negotiated with
the World Bank and IMF, the Indonesian government has not taken many of these up.
Not considered a Highly Indebted Poor Countries (HIPC) country, Indonesia is not likely
to participate in Poverty Reduction Support Credit (PRSC) or Poverty Reduction and
Growth Facility (PRGF) mechanisms. Considerable suspicion remains towards
international financial institutions.
30. In the current complicated political climate of policy-making with a fragile
democracy and lagging reform, as well as endemic graft and corruption, there is a need
for policy transparency, with more policy-making built on solid analysis linked to data
generation. In this context the PRSP has been undertaken with guidance from the World
9
The food price index, which was 100 in the second half of 1996 increased to 261 in September 1998,
with housing, clothing and health price indices increasing to only 156, 225, and 204 respectively.
10
A. Suryahadi and S. Sumarto, The Chronic Poor, the Transient Poor, and the Vulnerable in Indonesia
Before and After the Crisis (May 2001)
11
The economic crisis resulted in sharp increases in the number of the poor (see Ch3). In order to
overcome the impact of the crisis, the government implemented a Social Safety Net Program
(JPS/SSN). This includes a) Health SSN programs, b) Education SSN programs, and C) Cheap Rice
Programs (OPK).
Indonesia Context: Rice Trade, Rice Tariffs and Poverty-Led Policy Decision-Making
Bank, as a demonstration of good governance, as an act of global citizenship but which
is not a requirement of any loan facility. An I-PRSP has just been produced, which will be
revised over the next two years before a full PRSP is completed. The high levels of
foreign debt, suspicion of international pressure for good governance, and the increasing
decentralisation of decision making make it difficult to gain support for poverty reduction
led by the PRSP.
2.2
POVERTY REDUCTION AND MACROECONOMIC POLICY-MAKING
Early New Order Era
31. The Government of Indonesia began implementing poverty reduction programmes
in the 1960s with the beginning of the New Order period of Suharto. During this early
period several departments implemented these programmes, including the Department of
Home Affairs, through experimental savings and loans projects; the Department of Social
Affairs, through projects focused on increasing the welfare of the poor and needy; and
the Department of Agriculture through programmes endeavouring to increase the income
levels of small farmers.
Late New Order Era
32. Beginning in 1994, the government implemented several new and more direct
programmes of community-managed social fund mechanisms to fight poverty, including:






Presidential Instruction on Disadvantaged Villages (IDT);
Disadvantaged Village Infrastructure Development Programme (P3DT);
Urban Poverty Reduction Programme (P2KP);
Takesra/Kukesra;
Small Farmers/Fisherman Income Expansion Project (P4K);
Kecamatan (Sub-district) Development Programme (PPK/KDP).
Post New Order Era
33. At the beginning of April 2001 during the Abdurrahman Wahid administration, the
government formed the Poverty Reduction Coordination Board (Badan Koordinasi
Penanggulangan Kemiskinan—BKPK), headed by HS Dillon. Institutionally, this board
existed under the coordination of the Vice President of the day, Megawati Sukarnoputri.
The primary objective of the BKPK was to co-ordinate the poverty reduction programs
more effectively and in an integrated manner. To carry out their function, the BKPK
undertook four main roles as co-ordinator, catalyst, mediator and facilitator for poverty
reduction.
34. In December 2001, through Presidential Decision No.124, 2001, Megawati, after
replacing Wahid as president, established the Poverty Reduction Committee (KPK)
dissolving the BKPK. The co-ordinating minister for Peoples Welfare and Poverty
Alleviation heads the KPK. Articles 4 and 5 of this decree state that the function of the
committee is to 'take concrete measures to accelerate the reduction in the number of
poor people in all regions of Indonesia. The function of the KPK is to make policy,
monitor, and report on poverty alleviation to the president'. The Secretariat of this
committee is based in BAPPENAS, with committee chair Jusuf Kalla from the Ministry for
Peoples Welfare and Poverty Alleviation. The development of a PRSP was pledged in
November 2001 at the Consultative Group on Indonesia (CGI) meetings of donors and
government. It is the KPK that is responsible for the production of the PRSP.
Commitment to the PRSP at both the political and bureaucratic levels remains elusive.
9
Poverty and Social Impact Analysis: Indonesia Rice Tariff
35. Independently of the PRSP, the Government of Indonesia must prepare a plan of
action for poverty reduction as part of its basis for accountability under the national
planning mechanisms, the Propenas in 2004 and the Repeta exercises in 2003 and
2004. This will be an interim strategy, and BAPPENAS is responsible for working out the
policy and how it will be operationalised. These difficulties between the I-PRSP's role in
leading a poverty strategy and the poverty reduction imperatives from national
government led by BAPPENAS highlights the difficulties of harmonising externally led
processes and ongoing government imperatives in an unclear and uncertain policy
environment marked by slow democratisation of government.
2.3
THE RICE TRADE IN INDONESIA
36. Indonesia is a net importer of rice, though the magnitude of its imports varies from
year to year depending on domestic production, international prices and the size of
stocks. Over the four years following the crisis of 1997–98 (1998 to 2001 inclusive) rice
imports were 9.1% of total consumption of rice (Table 1). Indonesia is the world’s largest
rice importer, accounting for 18% of the world’s total imports between 1998 and 2000
(Table 2). Thailand is the largest exporter, followed by Vietnam and the USA.
37. Prior to the 1997–98 crisis Indonesia’s rice imports were monopolised by a public
agency, Bulog. Figure 1 shows that except for the periods of the 1973 commodity price
boom and the 1997–98 exchange rate crisis, the real price of rice in Indonesia has been
relatively stable, but its post-crisis level has been above its level over the previous three
decades, even though international rice prices have declined relative to other traded
commodities.
38. From this and from Figure 2 it is apparent that the effects of Bulog’s market
interventions were to stabilise rice prices relative to international prices at a level not
significantly different from the trend level of world prices. With the exchange rate volatility
of the crisis period, local currency prices of imported rice surged. For a brief period,
Indonesian domestic prices remained well below exchange rate adjusted world prices,
but from about 2000 onwards they have stabilised at levels 40–50% above import prices.
39. The large difference between the domestic and import price arose from changes in
rice import policy that followed the 1997–98 crisis. After Bulog’s monopoly on rice imports
was abolished in 2000, agency still accounted for around 75% of total imports. Private
imports were subject to a specific tariff (rather than an ad valorem tariff) of Rp 430/kg,
which in mid-2002 was around 25% of the import price (cif). In addition, private sector
rice imports were subject to 'red lane' customs treatment, meaning stricter standards of
customs inspection than other food items, and were also subject to special import
licensing requirements.12 The tariff plus these non-tariff barriers apparently account for
the increased difference between the border price of imported rice and domestic prices.
40. There is now a proposal before the Parliament to increase the tariff by 75%, from
Rp 430/kg to Rp 750/kg, raising the ad valorem equivalent tariff from 25% to about 45%.
12
These requirements are known as NPIK: Nomor Pengonal Impor Khusus.
10
Indonesia Context: Rice Trade, Rice Tariffs and Poverty-Led Policy Decision-Making
2.4
THE DEBATE: RICE TARIFFS
41. This section first discusses the various political and economic arguments used by
proponents of both an increase and an abolition of the import tariff on rice in Indonesia.
The section that follows offers some empirical evidence that backs up the debate and the
third part discusses a methodology that can analyse the welfare effects of the tariff.
Arguments for Raising the Rice Tariff: Tariff Theory
42. The arguments for raising the tariff on rice that have been made in the Indonesian
debate thus far focus on the political value of self-sufficiency in rice, global trade issues,
and the role of rice in income generation for the rural poor. These have been extensively
debated in public13, and are briefly outlined below.
Thin World Rice Market / Self-sufficiency in Rice Production
43. Achieving self-sufficiency in rice production has been a long-standing goal of
Indonesian agricultural policy. Due to the thin and unstable international rice market (only
4% of rice production is traded), the price of rice can fluctuate and a large buyer can
drive up prices rapidly (Indonesia accounted for around 10% of total world rice imports in
1998, the peak of the economic crisis).
Unfair Competition / Dumping
44. Governments around the world subsidise and protect their domestic rice industry.
An oft-quoted argument for increasing tariffs is to protect domestic producers from the
'dumping' of cheap imports on the world market. To the extent that such price fluctuations
have a short-term effect that distorts longer-term resource allocation decisions,
proponents of an increased tariff argue that Indonesia should isolate its rice industry from
the world market.
Improving Incomes of Rural Farmers
45. Advocates of an increase in the tariff stress the importance of rice incomes to the
rural economy. The induced increase in domestic rice prices would feed through into
better incomes for rural farmers, who are widely perceived to be amongst the poor (NB.
70% of the poor in Indonesia are rural, though rice farmers are not necessarily the
poorest group (Ikhsan 2001)).
Rice Industry as an Engine of Growth in Rural Areas
46. Following on from the above point, increased returns to rice farming may act as an
incentive to increase production of rice in Indonesia, which would be a stimulus to the
rural economy through some combination of wage and employment effects, which have
not been documented in any quantitative study as yet. An increase in the price of rice
would increase the aggregate demand for unskilled labour that would produce some
combination of increased employment and increased real wages for landless labourers.
Longer-Term Incentives to Improve Rice Productivity
47. A further long-term effect of protecting the rice industry now is that increased price
generates incentives to invest and therefore improve rice productivity in the longer term14.
13
See Ikhsan, 2001, Timmer, 2000, Bappenas, 2001 for more details
Bulog has also claimed that protecting the rice industry is good for the environment, because it would
keep irrigated land in rice production that might otherwise become idle. Few observers have agreed
that the increased pesticide and fertilizer use that would follow, not to mention increased demand for
irrigation water, could have environmental benefits. In any case, there seems little possibility that
irrigated land not used for rice production would be left idle.
14
11
Poverty and Social Impact Analysis: Indonesia Rice Tariff
The 'infant industry' argument is often used in this context. However, Indonesian rice
production is amongst the most productive in the world (Robinson et al., 2000).
The Optimal Tariff Argument
48. Indonesia is a large importer of rice relative to the world market and therefore the
volume of Indonesian imports can affect the world price15. The gains in national income
from imposing a positive tariff are achieved through an improvement in the terms of
trade—the tariff induces a reduction in the price of imports relative to exports by reducing
the quantity of imports. If the elasticity of supply of imports to a country is , then the
proportional rate of tariff that maximises national income is 1/. For example, if the
elasticity of world supply was 5, the optimal tariff would be 20%. Warr (2002) considers
that a reasonable estimate of the long run elasticity of supply of rice imports to Indonesia
would be between 7% and 10%, and therefore that tariffs in the neighbourhood of 10–
14% would be the largest that could be justified through the optimal tariff argument.
Nevertheless, the true value of the long run elasticity of supply of rice imports to
Indonesia must be considered uncertain.
Arguments Against Higher Rice Tariff
49. Arguments that have been raised in the Indonesian context against raising the tariff
are diverse. In terms of welfare, the main concern is that the poor are net consumers of
rice and will necessarily be hurt by an increase in price. Some consider the tariff not to be
the source of the high domestic rice price and therefore irrelevant to the above
arguments.
An Increase in the Rice Price Hurts the Poor
50. 'High rice prices have been a large burden on consumers and have been a primary
cause of the surge in poverty in recent years. Allowing rice price to fall by Rp 430/kg
through elimination of the current rice import tariff would do more for poverty alleviation in
Indonesia than all other government programs combined'.16
51. Poor people are net consumers of rice, and therefore would lose out from an
increase in rice. There are more net rice consumers amongst the poor than net
producers. Rice consumption is inelastic with respect to price. If the price goes up, poor
people tend to protect their rice consumption and consume less of other foods (that may
have more nutritional content). Despite a rise in the relative price of rice, rice
consumption has steadily increased from 24.41 million tons in 1990 to 27.72 million tons
in 2000.
In the Long Term, Rice is a Declining Industry and Tariffs Distort Resource
Allocation
52. Economic theory suggests that in the long term, and with no distortions in markets,
zero tariffs would be the optimal rate (except possibly in the case of optimal tariff
argument). A competitive price is more efficient in terms of both productive efficiency and
allocative efficiency. In practice there are many departures from perfect competition,
especially in agricultural trade. However, higher tariffs can distort resource allocation
away from the sectors in which a country has a comparative advantage. Raising the tariff
artificially raises the returns to rice and creates incentives for farmers to produce rice,
15
16
In economic terms, the marginal cost of Indonesia’s imports exceeds the world price
Indonesian Food Policy Program, Policy Brief No. 22
12
Indonesia Context: Rice Trade, Rice Tariffs and Poverty-Led Policy Decision-Making
whereas allowing the rice price to be market-determined would create incentives for
farmers to diversify into higher-yield export crops.
Import Tariff is Not the Most Efficient Way to Protect Rural Farm Incomes
53. Rodrik (1995) argues that a tariff is the most costly way of achieving the objective
of protecting an industry. Cheaper policies include subsidising agricultural workers
directly or subsidising rural farmers directly through an employment subsidy or a
production subsidy. The reason a tariff is often chosen is that it is the only option that
increases government revenue. However, it would still be possible to introduce a subsidy,
and raise taxes from another, less discriminatory source. This introduces political
economy issues, since a tariff on rice can be considered regressive in terms of its effect
income distribution when the poor spend disproportionately more of their incomes on it.
Farmers May Not Receive the Benefits from the Tariff
54. The domestic rice price is already higher than imported rice price. During this
period, the average domestic rice price was Rp 2,511.66/kg while the average imported
rice price is Rp 2,337.14/kg17.
2.5
KEY PLAYERS IN THE RICE TARIFFS DEBATE AND WHAT THEY SAY
55. The section above provides an overview of the rationale used by proponents of
both a higher tariff and a lower tariff. Given the degree of research and data available in
Indonesia to justify each perspective, it is useful to try to make sense of these arguments
by situating them in the Indonesian policy context.
56. One of the challenges for the PSIA researchers was to link the politics of policy
making with the body of evidence collected on the topic of rice pricing and rice tariffs.
57. Decision-making in Indonesia appears to be rooted in politics and bureaucratic
economic interests, as our scoping mission found in interviews with key players18.
58. To link policy perspectives and policy-makers, the various policy narratives of key
Indonesian government, academic, non-governmental organisation (NGO) and donor
actors in the policy debate were identified and examined. These are presented in the
matrix below (Table 2.1). Each key actor cited was considered key because of public
pronouncements on the issue of rice tariff increases. Not all, however, have the same
degree of influence.
59. The horizontal logic of the matrix presents the policy perspectives of the institutions
or individuals, explains how the benefits and constraints of the policy perspective were
viewed by the individual or institution and assesses the relative degree of influence of
each actor. This was done through analysis of documents, through interviews, through
newspapers and through discussions with individual academics and focus group
discussions with Jakarta-based civil society (Appendix 2).
17
Thai 25% broken f.o.b plus US$ 20 per ton shipping cost from Bangkok to Jakarta wholesale market,
plus rice import tariff of Rp 430 per kg, plus wholesale-retail price margin of 10%.
18
The logic of sound argument which assumes a rational approach with a technical fix based on sound
data ignores the actual experience of policy making. Any process of policy change is inherently political.
It can include and exclude interests and perspectives of various groups of people including the poor .
13
Poverty and Social Impact Analysis: Indonesia Rice Tariff
60. The policy actor narratives can also be grouped together from the vertical logic by
columns, on the basis of pro- or anti-tariff, on the basis of the rationale or argument or by
key interests.
61. It was agreed by those interviewed that the four key Indonesian institutional actors
for policy-making in the rice tariff issue include the Ministry of Economy (low tariff);
BAPPENAS (no tariff); Bulog, which advocates a high tariff; and the Ministry of
Agriculture, also advocating a high tariff. The coordinating ministry for People’s Welfare
and Poverty Alleviation has made contradictory statements on the issue. The World Bank
is a highly influential non-state actor, though it functions outside direct decision-making.
62. Whose interest do the key actors represent and how much do their interests count?
The final chapter, chapter 6 will return to examine this in more detail.
14
Indonesia Context: Rice Trade, Rice Tariffs and Poverty-Led Policy Decision-Making
Key Actors—Policy Interest Matrix
Key players
Policy Objective
Explicit
In
formal
BULOG
high tariff,
regulate
imports
source of
finance
Dept Agric.
High tariff
Encourage
domestic
prod
of
rice
Self
sufficiency
Bappenas/DAI
(consult
ants
No tariff
to
maintain
low
rice
prices
Argument
rationale
Benefits
Constraints
(Shortterm)
(Mediumterm)
(Longterm)
(Shortterm)
(Mediumterm)
protect
local
farmers
from
import
dumping
stable
rice mkt,
domestic
production
Increase
rice
produce
increase
farmers
(profitability)
self
sufficiency
control of
rice
economy
don't have
control of
policy
smuggling
Tariff less
binding
Creation
of
black
market
Its
their
job
perform
based on
Agri
prod
Higher
returns to
rice
farmers
Maintain
high
income
of
farmers
More
rice
availability
None
DAI
represents
US interests?
Java
should
diversify
out of rice
low price
benefit to
poor
stable
economy
food
available
Transmission
channel
Interests
Degree
of
influence
high
pricehigher wages
for
labour,
benefit
to
farmers
Source of
income
high
linked to
ruling
party
fund
raising
Political
High tariff
local rice
prod
High
(less than
Bulog)
Pro
free
market,
allied with
intl
community
WB
High but
declining
Influence
is waning
None
stated.
Unsure if
would
encourage
high
wages
Efficient
resource
allocation
removes
distortion
Farmers
will plant
high yield
crop
Sustainability of
diversity
Low price for
rice
15
Poverty and Social Impact Analysis: Indonesia Rice Tariff
Key players
Policy Objective
Explicit
In
formal
Low prices
for rice
No
Min Trade/
Industry
No tariff
more open
trade
Min Peoples
Welfare
Kalla-low
tariff
Lubis-high
tarrif
Ministry
Finance
Ministry
Economy
16
&
Argument
rationale
Benefits
Constraints
(Shortterm)
(Mediumterm)
(Longterm)
(Shortterm)
(Mediumterm)
No
rice
tariff will
help poor
people
Increase
purchase
power
Flexibility to
plant high
price crops,
less
dependent
on govt
Better
resource
allocation
Farmers
will suffer
in
short
term
Employment
probs
Shortage
of rice, no
self
sufficiency
Lower
direct/
indirect
costs
to
indust
less
politics
efficient
resource
allocation
low price
ease
pressure
on wage
demands
Macro econ
benefit
None
low food
price for
poor
low price
benefit
poor
No power
to enforce
Transmission
channel
Interests
Degree
of
influence
Prices
wages
and
Stabilized
lower price
of
basic
needs
(econ
high, fin
high but
less so)
lower
than
BULOG
Low price for
rice
industry/
prices,
clean govt
medium
(less than
agri)
declining
less
burden on
them!
high
(more
than
Bapp)
some say
no
influence
on tariff
setting
Indonesia Context: Rice Trade, Rice Tariffs and Poverty-Led Policy Decision-Making
Key players
Policy Objective
Argument
rationale
Benefits
(Shortterm)
Constraints
Explicit
In
formal
(Mediumterm)
CAP (an NGO)
High tariff
None
protect
agri prod
increase
product
vity before
indust
rialise
higher
rural
wage
lowers
poverty
a. Academ
b. Research
Low tariff
Econ
principles
high price
bad
for
poor net
cons.
Low
price
helps
poor
Increased
consumption
USAID/
WB/IMF/ADB
No tariff
Trade
liberal
isation
increase
welfare
high price
leads
to
black
market
Cheap
price
Helps
poor
Stabilize
price
c. HKTI
Producers
NGO
High tariff
Business
interest
Higher
prices
protects
farmers
Benefit
farmers
(Longterm)
(Shortterm)
high
productivity
in
agri
sector
Transmission
channel
Interests
Degree
of
influence
price
increase
wages
maybe
political
ambition?
No (has
no allies)
Academic
based on
theory and
data
medium
(some
access to
minister
and
media
free trade
high (can
block)
usually
informal
(Mediumterm)
rise
Market
efficiency
Better
resource
allocation
not policy
makers
Like
BAPPENAS
Higher
productivity
not part of
govt
see Bulog
low
but
vocal
17
Poverty and Social Impact Analysis: Indonesia Rice Tariff
Key players
Governor
East Java
18
Policy Objective
of
Explicit
In
formal
Ban
Imports
Populist
Rent
seeking
Argument
rationale
Protect
farmer
interests
Benefits
Constraints
(Shortterm)
(Mediumterm)
(Longterm)
(Shortterm)
Greater
sales of
domestic
rice
Higher
incomes for
farmers
Expansion
of
rice
production
Un
enforce
able
Transmission
channel
Interests
Degree
of
influence
Blockages at
ports, customs
Election
2004
Access to
media
and
politicians
(Mediumterm)
3 Poverty in Indonesia
This chapter provides data on the situation of poverty in Indonesia in the last five years,
and focuses on the effects of the Asia financial crisis on poverty. This is the context for
poor people that the proposed rice tariff policy will affect.
3.1
POVERTY DEFINITIONS
63. Indonesia’s approaches to poverty have been characterised as welfarist, using a
narrow and traditionalist approach linking to income and consumption definitions of
poverty. The consumption-based measurement of poverty (defined as the inability of a
person to fulfil their minimum basic material needs of consumption) encompasses a
poverty line, which identifies the minimum requirements needed to live, including both
food and non-food stuffs, which are consumed by each person. Though consumptionbased poverty is useful for identifying the numbers of vulnerable in economic terms,
poverty may be defined more broadly to include other dimensions of life in which people
may be vulnerable. If we examine poverty incidence from Sen’s human capability
perspective, which includes not only income dimensions but unmet basic needs in health,
housing, education and literacy, clean water and access to infrastructure, the
consumption-based indicators of poverty and deprivation only partially capture the
magnitude and intensity of poverty in Indonesia. More recently the concept of poverty has
included dimensions of future security, and social participation, vulnerability,
powerlessness, and voicelessness of the poor.
The Consumption / Expenditure approach to Measuring Poverty
64. Beginning in 1976, based on data from the National Socio-Economic Survey
(Survey Sosial Economi Nasional—Susenas), Statistics Indonesia (Badan Pusat
Statistik—BPS) has been estimating poverty rates using this approach. The poverty line
determined by the BPS is made up of two components: the food poverty line and the
non-food poverty line. The food poverty line is based on the minimum food requirements
to live healthily, which is determined to be approximately 2,100 calories per person per
day. Up until 1990, this minimum value was obtained by directly calculating the value
(cost) of 2,100 calories in rupiah. However, the price of the calories used actually referred
to the price paid by people whose income was sufficient to purchase 2,100 calories of
food per person per day. Beginning in 1993, the food poverty line was determined by
calculating the value in rupiah of a basket of commodities (containing 2,100 calories).
The basket of food commodities (including 52 types) was chosen based on the amount of
calories consumed, the frequency with which a household consumed the calories, and
other considerations.
65. For non-food commodities, adequacy is based on a level of expenditure considered
to reflect basic non-food needs. Until 1990, 14 types of non-food commodities were
included in the calculation of the poverty line for urban areas and 12 types of non-food
commodities for rural areas. Beginning in 1993, the composition of the non-food
commodities was increased to 46 types without distinguishing between urban and rural
areas. The urban–rural differences were accounted for by the price differences for each
of the commodities used in the calculation.
Poverty and Social Impact Analysis: Indonesia Rice Tariff
3.2
HOW MANY ARE POOR
66. Using this method of calculation, Table 3.1 below indicates the developments in the
total population categorised as poor in Indonesia between 1981 and 2000. In 1981, 40.6
million people were recorded to be poor (26.9%). A number of direct and indirect
government development and poverty alleviation programmes carried out after 1981
significantly reduced the total population categorised as poor. The total poor population
dropped to 27.2 million (15.1%) in 1990, and 22.5 million (11.3%) in 1996.
67. The economic crisis, which began in mid-1997, caused a large proportion of the
population’s real incomes to experience a drop. The direct impact of this has been a
sharp increase in the poor population. Based on data from December 1998, the poor
population reached 49.5 million (24.2%), 17.6 million of them residing in urban areas and
the remaining 39.1 million in rural areas.
68. In 1999, based on these methods of calculation, the value of the poverty line for
Indonesia was Rp 94,507 or US$10.60 (for urban areas), and Rp 74,405 or US$8.40 (for
rural areas)19 per capita per month calculated at the present exchange rate. This level of
the poverty line constitutes the total value of food commodities equivalent to 2,100
calories valued at Rp 70,741 (for urban areas), and Rp 58,917 (for rural areas) per capita
per month; and non-food commodities valued at Rp 23,766 (for urban areas) and Rp
15,488 (for rural areas) per capita per month.
69. In 2000, the total poor population in Indonesia (excluding the provinces of Aceh and
Maluku) was 37.3 million (19.0%), the majority of them residing in rural areas (25.1
million), while the remainder were in urban areas (9.1 million). A large proportion of these
were living primarily in Java and Bali (59%), Sumatra (25%), as well as Kalimantan,
Nusatenggara, Maluku and Irian (25%).
70. In 2001, preliminary data from the Central Bureau of Statistics indicated that the
total poor population in Indonesia (excluding Aceh) was 18.4% (PRSP, 2001).
TABLE 3.1 THE POPULATION CLASSIFIED AS POOR IN INDONESIA (1981 – 2000)
Year
Total Poor Population (million)
Pre-crisis
1981
1990
1996
40.6
27.2
22.5
Urban
Rural
Total
Post-crisis
1998
17.6
31.9
49.5
2000
9.1
25.1
37.3
Source: Coordinating Team for the Preparatory Stages of Policy Formulation on Poverty Alleviation
(2002), ‘Rancangan Kebijakan Interim Strategi Penanggulangan Kemiskinan’.
The Mean of Real Expenditures Declined from the Crisis
71. Table 3.2 shows the mean of nominal and real expenditures by quintile in February
1996 and February 1999. Nominal expenditures show an increasing trend from 1996 to
19
The current exchange rate at present is approximately Rp8.900/USD
20
Poverty in Indonesia
1999 with the poorer quintiles showing a higher increase than the richer quintiles.
However, in real terms, the mean declined ranging from -6.47% for the poorest quintile to
-23.84% for the richest quintile.
TABLE 3.2 CHANGES IN THE MEAN OF NOMINAL AND REAL EXPENDITURES BY
QUINTILE
Quintile
st
Nominal Expenditures
Feb 1996
Feb 1999
1
27,848
nd
2
39,969
rd
3
52,400
th
4
72,459
th
5
157,192
Total
69,972
Source: SMERU calculation.
61,470
86,107
109,981
146,376
282,517
137,284
% Change
120.73
115.44
109.89
102.01
79.73
96.20
Real Expenditures
Feb 1996
Feb 1999
27,848
39,969
52,400
72,459
157,192
69,972
26,046
36,486
46,602
62,024
119,710
58,171
% Change
-6.47
-8.71
-11.06
-14.40
-23.84
-16.86
Human Poverty Index
72. Recent poverty assessment literature has increasingly focused its attention on
techniques that try to delineate the non-income dimensions of poverty by drawing
attention to such basic needs as access to safe water, education and health. Thus, a
broader measure of poverty is the United Nations Development Programme's (UNDP’s)
human poverty index (HPI), which combines indicators on life expectancy, illiteracy,
malnutrition, and access to safe water and health services. As shown in the table below,
the HPI fell from 27.6% in 1990 to 25.2% in 1995, and kept steady at this level until 1999.
Within Indonesia, HPI ranges from a high of 47.7% in the district of Jaya Wijaya in Papua
to a low of only 8.3% in North Jakarta. 20
TABLE 3.3 HUMAN POVERTY INDEX
1990
1996
1999
Human Poverty Index
27.6
25.2
25.2
People not Expected to Survive to Age 40 (%)
15.2
18.3
15.2
Adult Illiteracy rate (%)
18.5
14.4
11.6
Population without Access to Safe Water (%)
54.7
53.1
51.9
Population without Access to Health Services (%)
54.7
10.6
21.6
Under-nourished Children Under the Age of Five (%)
44.5
35.4
30.0
Source: BPS-Statistics Indonesia/Bappenas/UNDP (2001), ‘Human Development Report 2001: Toward
a New Consensus’.
Human Development Index and Poverty Incidence
73. The human development index (HDI) is a composite measure that reflects not just
income, but also life expectancy, infant mortality rate, literacy rate, and mean years of
schooling.
20
See BPS-Statistics Indonesia/Bappenas/UNDP (2001), page 8.
21
Poverty and Social Impact Analysis: Indonesia Rice Tariff
TABLE 3.4 HUMAN DEVELOPMENT INDEX
1990
1996
1999
Human Development Index
63.4
67.7
64.3
Life Expectancy (years)
63.2
66.4
66.2
Infant Mortality Rate
56.0
44.0
44.9
Literacy Rate (%)
81.5
85.5
88.4
Mean Years of Schooling
5.3
6.3
6.7
Purchasing Power Parity (Thousand Rupiah)
555.4
587.4
578.8
Source: BPS-Statistics Indonesia/Bappenas/UNDP (2001), ‘Indonesia Human Development Report
2001: Towards a New Consensus’.
74. The table below shows the relationship between poverty incidence and HDI.
Provinces with high HDI tend to have low poverty incidence. As expected, Jakarta that
has the smallest poverty incidence (2.82%) shows the highest HDI of 73. On the other
hand, Papua and East Nusa Tenggara, two provinces with the highest poverty incidence
(61.18% and 54.89%) show relatively low HDI of 60 and 59 respectively.
TABLE 3.5 POVERTY INCIDENCE AND HUMAN DEVELOPMENT INDEX
Poverty Incidence (%)
Feb-99
Urban
Rural
Total
HDI
2.82
9.21
9.47
11.15
12.89
13.62
15.27
20.44
20.64
21.67
22.18
22.47
22.63
23.81
26.6
26.95
28.52
30.76
32.78
33.31
36.61
36.8
41.78
48.4
54.89
61.18
73
67
66
67
65
66
67
65
62
68
65
67
64
64
65
69
63
61
65
62
63
63
54
67
59
60
Indonesia
16.34
34.1
27.13
Source: SMERU calculation and BPS-Statistics Indonesia/Bappenas/UNDP (2001),
‘Indonesia Human Development Report 2001: Towards a New Consensus’.
64
Jakarta
Riau
West Sumatera
Central Kalimantan
Aceh
Bali
North Sumatera
Bengkulu
South Kalimantan
East Kalimantan
Jambi
North Sulawesi
South Sulawesi
South Sumatera
West Java
Yogyakarta
Central Sulawesi
West Kalimantan
Central Java
East Java
Southeast Sulawesi
Lampung
West Nusa Tenggara
Maluku
Papua
East Nusa Tenggara
22
2.82
8.53
8.78
5
5.43
10.67
10.81
10.41
7.99
8.74
15.41
11.7
17.42
14.47
20.82
22.12
16.72
6.17
23.72
19.51
13.74
19.9
30.17
18.64
6.07
28.67
9.62
9.74
13.43
15.41
15.61
18.91
24.55
26.38
35.06
25.25
26.83
24.94
27.93
31.87
36.78
32.69
38.04
37.76
40.87
44.44
40.57
44.71
59.9
72.19
66.11
Poverty in Indonesia
Poverty Incidence is Deeper in Eastern Indonesia
75. In February 1999, the poverty incidence in Indonesia was 27.13%, implying around
55.8 million poor people. Indonesia’s urban poverty incidence was 16.34% and rural
poverty incidence was 34.10%. By region, the smallest urban poverty incidence of 2.82%
was found in Jakarta, meanwhile the highest urban poverty incidence of 30.17% was
found in West Nusa Tenggara. As expected, compared to urban poverty incidence, rural
poverty incidence in Indonesia showed a wider range from the smallest incidence of
9.62% in Riau, to the highest poverty incidence of 72.19% in Papua. If rural and urban
poverty incidence were combined, Riau showed the smallest poverty incidence of 9.21%,
while East Nusa Tenggara showed the highest poverty incidence of 61.18%.
Inequality Slightly Declines
76. Indonesia’s inequality indices are shown in the table below. As expected, inequality
was more marked in urban areas (indicated by a higher Gini coefficient) than in rural
areas. It also shows that both urban and rural areas experienced a decrease in
inequality.
TABLE 3.6 PERCENTAGE CHANGES IN INEQUALITY INDICES BETWEEN 1996 AND
1999 BASED ON PER CAPITA REAL EXPENDITURE
February 1996
February 1999
Percentage Changes
Source: SMERU calculation.
Urban
Rural
Total
0.37
0.35
-6.37
0.28
0.27
-6.03
0.36
0.33
-8.07
Poverty Profile in Agriculture
77. Table 3.7 shows the poverty incidence across sectors as well as the contribution of
each sector to total poverty in both February 1996 and February 1999.
TABLE 3.7 POVERTY INCIDENCE AND CONTRIBUTION TO TOTAL POOR BY MAIN
SECTOR OF OCCUPATION, FEBRUARY 1996 AND FEBRUARY 1999 (%)
February 1996
Agriculture
Trade, hotel, and restaurant
Manufacturing industry
Civil, social, and private services
Transport and communication
Construction
Receiving transfer
Mining and quarrying
Others
Finance, insurance, and leasing
Electricity, gas, and water
February 1999
Poverty
incidence
Contribution
to total poor
Poverty
incidence
Contribution
to total poor
26.29
7.96
10.69
5.73
8.85
14.04
6.58
15.34
13.29
1.24
6.10
68.54
8.10
5.71
5.72
3.32
5.42
1.86
1.01
0.10
0.06
0.16
39.69
17.63
22.92
13.13
24.02
28.97
15.57
29.81
32.00
5.23
14.48
58.38
11.13
7.71
7.36
5.58
5.52
2.65
1.00
0.27
0.23
0.17
Note: Sorted by contribution to total poor in February 1999
Source: SMERU calculation.
78. Table 3.7 indicates that all sectors uniformly experienced an increase in poverty
incidence during the period. Though in relative terms, the finance, insurance, and leasing
sector had the highest increase in poverty incidence, and other modern sectors such as
trade, manufacturing, and services were also proportionately hard hit by the crisis.
23
Poverty and Social Impact Analysis: Indonesia Rice Tariff
79. Nevertheless, the agriculture sector consistently had the highest poverty incidence
as well as the highest contribution to the total number of poor people during the period.
This reflects two things. First, people in agriculture sector have always been relatively
poor compared to other sectors. Therefore, even though this sector was not hit by the
crisis as hard as the modern sectors, the poverty incidence in this sector still has the
highest of all sectors. Second, the agriculture sector remains the largest sector in terms
of employment. In fact, during the crisis many workers who were laid off in modern
sectors returned to agriculture, so that between 1997 and 1998 the employment share of
agriculture increased from 40.8% to 45%21. The combination of these two factors
explains the persistence of the agriculture sector as the largest contributor to the number
of poor people, even though its importance has declined from 68.5% in February 1996 to
58.4% in February 1999.
3.3
RICE PRODUCTION IN INDONESIA
80. Rice production shows an increasing trend from 28,552,971 tons in 1990 to
31,725,062 tons in 2001. During the same period the volume of rice imports fluctuated.
During 1990 to 1999, the volume imported reached the lowest level of 3,093 tons in
1993, falling from 566,441 tons in 1992, and reached the highest level of 3,055,414 tons
in 1999.
81. As of January 2002, data shows that domestic medium-quality rice was traded at
Rp 2,978.47/kg in rural areas in Java, while in Jakarta it was traded at Rp 3,232/kg. In
the world market, Thai 25% broken rice were traded at around Rp 1,737—Rp
1,804.40/kg (fob). Taking shipping costs, the current import tariff of Rp 430/kg and price
margin into account, imported rice is traded at around Rp 2,612–Rp 2,686.64/kg. As of
July 2002 domestic medium quality rice was traded at approximately Rp 2819.66/kg.
3.4
THE IMPORTANCE OF RICE FOR THE POOR
82. Ikhsan (2001) shows that expenditure on rice contributes 60–65% of food
expenditure of the poor. Before the Asia crisis, rice accounted for 20% of total
expenditures for the poorest quartile of urban households. For the poorest 5%, this share
rises to 25% and was even higher at the peak of the crisis. Most urban households are
net rice consumers, while more than 55% of rural households are net rice consumers22.
The impacts of higher rice prices on the poor are dramatic. It is estimated that as many
as 10 million people could be lifted out of poverty (as of December 1998) from a 15%
decline in rice prices (Ikhsan 2001), though this analysis does not take into account any
decrease in wage income that would affect workers in the rice industry.
3.5
LINKS BETWEEN RICE TARIFF AND POVERTY
83. A change in the tariff on rice will affect people’s lives through three main channels:
price effects which affect consumers of rice; income effects which relate to producer’s
profits; also wages and employment in the rice industry. There are also government
revenue effects of the increased income from the tariff. In terms of poverty in Indonesia, it
21
Feridhanusetyawan, Tubagus (1999), The Impact of the Crisis on the Labor Market in Indonesia,
Report prepared for the Asian Development Bank, Center for Strategic and International Studies,
Jakarta in Pradhan et al (2000).
22
In addition Susenas 2001 (National Socio-Economic Survey) shows that of 51,372,653 households,
only 25.36 % of households plant paddy
24
Poverty in Indonesia
has been shown that poverty is highest in the rural areas, but also that the poor are net
consumers of rice.
84. Overall, the net effect of a change in rice tariff will depend on the relative
magnitudes of these three effects, and people in different livelihood groups will be
affected in different ways.
85.
In the next section we consider these questions by using a CGE model.
25
4 CGE Model and Results
This chapter provides a technical discussion of the CGE model, and the results of the
simulated increase in rice tariff.
4.1
THE CASE FOR A GENERAL EQUILIBRIUM TREATMENT
86. An adequate analysis of the distributional effects of a tariff on rice imports needs to
take account of its effects both on households’ expenditures, disaggregated by
household group, and its effects on their incomes. This requires taking account of its
effects on the labour market as well as the returns to land. In doing this, the rice industry
cannot be considered in isolation. An increase in unskilled wages would affect profitability
in other industries, with effects on outputs and prices in those industries as well. These
effects would have repercussions on household incomes. These effects would then have
to be balanced against the effects on consumers of an increase in the price of rice. But
the consumption of rice could not be considered in isolation either. An increase in the
price of rice would have implications for the demand for other staple foods, such as those
based on corn and wheat flour, another significant import.
87. The following section describes the Wayang general equilibrium model of the
Indonesian economy that was used in this analysis. It is a fairly technical discussion,
which non-specialists can skip if necessary, as the following two sections have been
designed to read using only the background provided in the Executive Summary. The
results section simulates an increase in Indonesia’s rice tariff, in particular noting its
effects on poverty incidence. The fourth section discusses the validity of the CGE results
in terms of assumptions used, dataset used, and sensitivity analysis around the assumed
values of key parameters. The fifth and final section concludes.
4.2
THE WAYANG GENERAL EQUILIBRIUM MODEL
88. This study uses the Wayang general equilibrium model of the Indonesian economy
(Warr et al. 1998; Wittwer 1999; Warr and Wittwer 2003), which identifies ten different
types of households, defined by socio-economic groups. The advantage of working with
a general equilibrium model with a disaggregated household sector is that it becomes
possible to conduct controlled experiments, which focus on the consequences for
household incomes, expenditures, poverty and inequality that arise from different
economic shocks, taken one at a time. Wayang is a conventional, real, micro-theoretic
general equilibrium model of the Indonesian economy. Its features are designed primarily
to enable it to address micro-economic policy issues relevant for Indonesia.23 As well as
disaggregating households, it also has a disaggregated industry and commodity
structure. The microeconomic behaviour assumed within Wayang is competitive profit
maximisation on the part of all firms and competitive utility maximisation on the part of
consumers. In the simulations reported in this paper, the markets for final outputs,
intermediate goods and factors of production are all assumed to clear at prices that are
23
A detailed paper describing the technical features of the full model is available (Wittwer 2000). The
present summary is intended to be as non-technical as possible to enable non-specialist readers to
grasp the essential features of the model.
CGE Model and Results
determined endogenously within the model.24 The nominal exchange rate between the
rupiah and the US dollar can be thought of as being fixed exogenously. The role within
the model of the exogenous nominal exchange rate is to determine, along with
international prices, the nominal domestic price level. Given that prices adjust flexibly to
clear markets, a 1% increase in the rupiah/dollar exchange rate will result in a 1%
increase in all nominal domestic prices, leaving all real variables unchanged.
89. This section briefly describes the major elements of the Wayang model (section
4.2.1). The household sector of the model is crucial for analysis of poverty incidence and
its most important features are summarised in this overview. The theoretical structure of
the model and its data base are described in sections 4.2.2 and 4.2.3. Important features
of the Wayang parameter estimates are described in Section 4.2.4.
4.2.1 Overview
90. The structure of the model itself is relatively conventional. Wayang belongs to the
class of general equilibrium models which are linear in proportional changes, sometimes
referred to as Johansen models, after the seminal work of Johansen (1964), which also
used this approach. Wayang shares many structural features with the highly influential
ORANI general equilibrium model of the Australian economy (Dixon, et al. 1982), which
also belongs to this Johansen category, but these features have been adapted in light of
the realities of the Indonesian economy. The principal features of the model are
summarised below.
Industries
91. The national model contains 65 producer goods and services produced by 65
corresponding industries—18 agricultural industries and 47 other industries. Each
industry produces a single output, so the set of commodities coincides with the set of
industries. The various industries of the model are classified as either ‘export-oriented’ or
‘import-competing’. The level of exports of an export-oriented industry are treated as
being endogenous, while the exports of an import-competing industry are treated as
being exogenous.25 The criterion used to classify these industries is the ratio of an
industry's imports to its exports. If this ratio exceeds 1.5, then the industry is regarded as
producing an importable. If the import/export ratio is less than 0.5, then the industry is
deemed to be export-oriented. For ratios between 0.5 and 1.5, additional relevant
information is used in classifying the industry.
Commodities
92. Wayang contains two types of commodities—producer goods and consumer
goods. Producer goods come from two sources: domestically-produced and imported. All
65 producer goods are in principle capable of being imported, although some have zero
levels of imports in the data base, services and utilities representing most of the
examples. The 20 consumer goods identified in the model are each transformed from the
producer goods, where the proportions of domestically produced and imported producer
24
Variations to this assumption are possible. For example, the possibility of unemployment can be
introduced by varying the closure to make either real or nominal wages exogenous, thereby allowing the
level of employment to be endogenously determined by demand.
25
Given that the exported and domestically sold good are treated as being identical, this assumption is
necessary to make it possible to separate the domestic price of the import competing good from the
price of the exported good. Otherwise, the Armington structure we have described above would be
redundant.
27
Poverty and Social Impact Analysis: Indonesia Rice Tariff
goods of each kind used in this transformation is sensitive to their (Armington) elasticities
of substitution and to changes in their relative prices.
Factors of production
93. The mobility of factors of production is a critical feature of any general equilibrium
system. 'Mobility' is used here to mean mobility across economic activities (industries),
rather than geographical mobility. The greater the factor mobility that is built into the
model, the greater is the economy's simulated capacity to respond to changes in the
economic environment. It is clearly essential that assumptions about the mobility of
factors of production be consistent with the length of run that the model is intended to
represent.
94. Two types of labour are identified: 'unskilled labour' and ‘skilled labour’. They are
distinguished by the educational characteristics of the workforce: skilled labour is defined
as those workers with lower secondary education or more. Indonesian labour force data
indicate that very little educated labour is used in agriculture. We therefore assume that
no skilled labour is employed in agriculture, but that skilled labour is fully mobile across
all non-agricultural sectors. However, unskilled labour is assumed to be mobile across
the entire economy. These assumptions imply that unskilled wages must be equal in all
sectors and that skilled wages must be equal in all non-agricultural sectors.
95. There are two kinds of mobile capital: one that is mobile among agricultural sectors,
and another that is mobile among non-agricultural industries. It is assumed that mobile
agricultural capital cannot be used outside agriculture and mobile non-agricultural capital
cannot be used in agriculture. In this treatment, agricultural capital is thought of as
machinery such as tractors of various kinds, which can be used in a variety of agricultural
activities. Non-agricultural mobile capital is thought of as industrial machinery and
buildings.
96. In every sector, it is assumed that there is constant elasticity of substitution (CES)
production technology with diminishing returns to scale to variable factors alone.
However, we introduce a sector specific fixed factor in every sector to assure that there
are constant returns to scale in production to all factors. We refer to the set of specific
factors in the agricultural sectors as ‘land’, and to the set of those in the non-agricultural
sectors as ‘fixed capital’. The assumption of constant returns means that all factor
demand functions are homogeneous of degree one in output. In each sector, there is a
zero profit condition, which equates the price of output to the minimum unit cost of
production. This condition can be thought of determining the price of the fixed factor in
that sector.
Households
97. The model contains ten household types, seven rural and three urban,
differentiated by socio-economic group. The sources of income of each of these
household types depend on their ownership of factors of production. These differ among
households and are estimated from the 1995 BPS Social Accounting Matrix (SAM). The
parameters of the consumption demand equations for the various household types also
differ. An approximate disaggregation to the level of individual households makes it
possible to derive estimates of poverty and inequality from data on the incomes and
expenditures of the 10 broad household types.
28
CGE Model and Results
98. Since our focus is on income distribution, the households of the model are of
particular interest. The source of the factor ownership matrix is the BPS SAM. The
document exists only in the Indonesian language. The households are described as
follows. The original Indonesian language descriptions are in square brackets:
1. Agricultural employees—Agricultural workers who do not own land
[Rumahtangga buruh tani]
2. Small farmers—Agricultural workers with land < 0.5 ha
[Rumahtangga petani gurem (yang memiliki lahan pertanian < 0.5 ha)]
3. Medium farmers—Agricultural workers with land 0.5 ~ 1 ha
[Rumahtangga pengusaha pertanian (yang memiliki lahan 0.5 ~ 1 ha)]
4. Large farmers—Agricultural workers with land >1 ha
[Rumahtangga pengusaha pertanian (yang memiliki lahan >1 ha)]
5. Rural low income—non-agricultural households, consisting of small retail
store owners, small entrepreneurs, small personal service providers, and
clerical and manual workers in rural areas
[Rumahtangga bukan pertanian golongan rendah di desa]
6. Rural non-labour households, consisting of non-labour force and unclassified
households in rural areas
[Rumahtangga bukan angkatan kerja di desa]
7. Rural high income—non-agricultural households consisting of managers
technicians, professionals, military officers, teachers, large entrepreneurs,
large retail store owners, large personal service providers, and skilled clerical
workers in rural areas
[Rumahtangga bukan pertanian golongan atas di desa]
8. Urban low income households, consisting of small retail store owners, small
entrepreneurs, small personal service providers, and clerical and manual
workers in urban areas
[Rumahtangga bukan pertanian golongan rendah di kota]
9. Urban non-labour households, consisting of non-labour force and unclassified
househods in urban areas
[Rumahtangga bukan angkatan kerja di kota]
10. Urban high income households, consisting of managers, technicians,
professionals, military officers, teachers, large entrepreneurs, large personal
service providers, and skilled clerical workers in urban areas
[Rumahtangga bukan pertanian gol. Atas di kota].
99. In the social accounting matrix each household's sources of income are classified
into several sources. A summary of the sources and disposal of income appearing in the
social accounting matrix is:
1. Wages and salaries [Upah dan gaji]
29
Poverty and Social Impact Analysis: Indonesia Rice Tariff
2.
3.
4.
5.
6.
7.
8.
9.
Rent from capital [Pendapatan kapital]
Incoming transfer [Penerimaan transfer]
Total above [Jumlah pendapatan]
Income tax [Pembayaran pajak lansung]
Net income [Pendapatan rumahtangga setelah pajak]
Final consumption [Pengeluaran konsumsi akhir rumatangga]
Outgoing transfer [Penbyaran transfer]
Saving [Tabungan]
100. The categories 'wages and salaries' and 'rent from capital' are each subdivided into
various sub-categories. These categories did not corresponded exactly to those of the
model. In agriculture, returns to land and capital were not separated in the SAM, but
returns to owner-provided labour were separated. A previous study on the cost structure
of paddy production was used to allocate returns among the land and capital categories
and the various farming households received the same proportionate breakdown of this
total. For agriculture the principle used was that machinery was considered 'mobile'
capital. Of course, mobile here means mobile across crops—tractors are the best
example. This involves error in so far as some machinery is crop-specific. Land was
considered immobile. It is best to think of what is called ‘land’ here as all immobile forms
of agricultural capital, which includes much true land in the short run. In non-agriculture
the principle used was that plant and buildings were classified as ‘mobile’. A factory
building can be used for many purposes. Machinery was considered ‘immobile’, because
most of it is more industry-specific than tractors are in agriculture.
4.2.2 Theoretical Structure
101. The analytical structure of the model includes the following major components:







30
Household consumption demands, of each of the 10 broad household types, for
20 categories of consumer goods, one of which is rice. These are derived from
the linear expenditure system.
The household supplies of skilled and unskilled are assumed to be exogenous.
A factor demand system, based on the assumption of CES production
technology, that relates the demand for each primary factor to industry outputs
and prices of each of the primary factors. This reflects the assumption that factors
of production may be substituted for one another in ways that depend on factor
prices and on the elasticities of substitution between the factors.
The distinction between skilled and unskilled labour, which are ‘nested’ within the
sectoral production functions. In each non-agricultural sector, skilled and
unskilled labour enter a CES production function to produce ‘effective labour’.
Effective labour, variable capital and fixed capital then enter the production
functions for domestic output.
Leontief assumptions for the demand for intermediate goods. Each intermediate
good in each sector is assumed to be demanded in fixed proportion to the gross
output of the sector.
Demands for imported and domestically produced versions of each good,
incorporating Armington elasticities of substitution between the two.
A set of equations determining the incomes of the 10 household types from their
(exogenous) ownership of factors of production, reflecting data derived from the
official 1995 SAM, the (endogenous) rates of return to these factors, and any net
transfers from elsewhere in the system.
CGE Model and Results


Rates of import tariffs and excise taxes across commodities, rates of business
taxes, value added taxes and corporate income taxes across industries, and
rates of personal income taxes across household types which reflect the structure
of the Indonesian tax system, using data from the Indonesian Ministry of Finance.
A set of macroeconomic identities which ensures that standard macroeconomic
accounting conventions are observed.
4.2.3 Data Base
102. This section provides a description of INDOSAM: a disaggregated SAM
Indonesia, with a 1995 base. This SAM is intended to serve as the data base
Wayang, but it has other potential uses as well. The year 1995 is currently the latest
which it is possible to assemble the information required for construction of a SAM
Indonesia.
for
for
for
for
103. Three principal data sources, all compiled by the government's principal statistical
agency, BPS, were used to construct INDOSAM-95:



the input–output tables for 1995 (subsequently referred to as IO 95);
the updated input–output table for 1995 (subsequently IO 95);
the 1995 SAM (subsequently SAM 95).
104. The table specifies 66 sectors. Other, supplementary, data sources were also used
in the construction of specific tables, as described below. Abbreviations are used for
these supplementary sources in the text and full references are provided at the end of the
paper.
The principal data sources
105. The 1995 SAM produced by BPS (SAM 95) provided the starting point for the data
base but substantial additions to the information in SAM 95 were required. SAM 95
contains 22 production sectors, which is insufficient for the purposes of this study. In
addition, the SAM 95 does not include the detail of tax payments and household sources
of income that are required. The 1995 input–output table specifies 66 production sectors.
For the purposes of the present study, modifications to the data contained in IO 95 were
needed for the following reasons.



The table specifies only total intermediate goods and services transactions for
each pair of producing and purchasing industries, at producer prices. Unlike the
1990 table, these transactions are not divided into goods and services from
domestic and imported sources.
The table includes a sector (number 66, labelled 'unspecified sector'), which is
included as a balancing item. Sector 66 does not describe a true sector of the
economy and in any case the data for this sector indicates negative final demand,
an economic impossibility.
The updated table (IO 95) derived from BPS was not fully balanced. The major
imbalances were that: (i) for most industries defined in the table, the industryspecific elements of row 210 (total input) were not equal to those of row 600 (total
output) and (ii) the elements of row 200 (total imports) plus row 600 (total output)
were not equal to those of row 700 (total supply).
106. These problems were overcome as follows
31
Poverty and Social Impact Analysis: Indonesia Rice Tariff



The shares of imported intermediate goods and domestically produced
intermediate goods for each cell of the table, as implied by the published 1990 IO
table, were used to divide intermediate goods transactions into domestic and
imported components.
Sector 66 was aggregated with the much larger sector 65 (labelled 'other
services'). This eliminated the problem of negative final demands. The resulting
table thus has 65 sectors.
The revised table was balanced using the RAS adjustment method to ensure that
all required accounting identities were observed.
Elasticity estimates
107. The elasticity estimates used in Wayang for the consumer demand system and the
factor demand system were taken from empirical estimates derived econometrically for a
similar model of the Thai economy, known as PARA. These parameters were amended
to match the differences between the data bases for Wayang and PARA so as to ensure
the homogeneity properties required by economic theory.
108. The Armington elasticities of substitution between imports and domestically
produced goods were set equal to 2, except for rice, where the assumed value was 6
(this parameter is varied in the results below). All export demand elasticities were set
equal to 20. The elasticity of supply of imports to Indonesia were assumed to be infinite
(import prices were set exogenously) except for rice, where the assumed elasticity was
10 (a parameter which is also varied in the discussion below).
4.2.4 Measurement of poverty
109. This section explains the way changes in poverty and inequality, at the level of
individual households, are estimated from changes in income and expenditure for broad
household groups. 26 The measures of Indonesian poverty and inequality used here are
based on household expenditure because this is the way the official Indonesian data on
poverty are derived. However, since the model closure assumes consumption to be
directly proportional to after-tax income, the choice between income and expenditure
based poverty measures is unimportant.
110. The Gini coefficient is used to measure inequality. Two familiar measures of
poverty are reported: the ‘headcount’ rate, defined as the proportion of the population
below the poverty line and the ‘poverty gap’, defined as the proportion of total national
consumption that would just suffice (if provided by an external donor at unchanged prices
and given other sources and uses of income) to raise the consumption of those below the
poverty line to the poverty line.
111. It is assumed that, while households belonging to different groups may own factors
in different proportions, those belonging to any one group all own the various factors of
production in the same proportions. For each broad household group h, income is
assumed to be log normally distributed over households, j with mean h and standard
deviation h. Only two separate values of h were used, rather than 10, because one
common value of the standard deviation was imposed on all rural groups, and another on
all urban groups. It can be shown that although factor prices affect the mean of the
26
A fuller discussion of these issues is provided in George Fane and Peter Warr, ‘How Economic
Growth Reduces Poverty: A General Equilibrium Analysis for Indonesia’, in A. Shorrocks and R. Van
der Hoeven (eds.), Growth and Poverty, United Nations University Press, forthcoming.
32
CGE Model and Results
logarithm of household expenditures, they do not affect the standard deviation of the
logarithm of individual household expenditures. It is therefore appropriate to assume that
h remains constant throughout all the simulations.
112. The parameter h was estimated separately for urban and rural households by
searching for the values that replicate official estimates of the national headcount poverty
rate and Gini coefficient.27 Using the standard properties of the log-normal distribution it
is then possible to derive h from h. The values of household expenditures before and
after various shocks are given by the Wayang model. It is then possible to derive the
distributions of individual household expenditures within each of the 10 broad groups. It is
then straightforward to estimate how the shocks affect poverty and inequality at the
national level.28
4.3
RESULTS
Raising the rice tariff will increase poverty, but only slightly. Urban sector
unambiguously suffers, some rural households will benefit through farm
profits and higher unskilled agricultural wages.
113. The results in this section show the medium-term effects (6–24 months later) on
household welfare when the import tariff on rice increases from 25% to 45%. The first
part presents results for the preferred specification of the model or 'base' scenario29, and
subsequent discussion focuses on the validity of the assumptions in the base scenario,
and effects of varying key parameters of the model. In summary, for all scenarios, overall
poverty increases when the tariff is raised, though sometimes marginally. The effect on
different households varies however, depending mainly on sources of income of the
household.
Income poverty (as measured by the headcount index) increases in both urban
and rural areas by 0.06% and 0.04% respectively.
114. The results of the model show insignificant overall effects on poverty. The depth of
poverty also increases, again slightly more in urban than rural areas (poverty gap rises
by 0.07% and 0.05% respectively). Overall inequality (measured by the Gini coefficient)
declines slightly. Again, the effects are not significant, but rural inequality slightly
increases (by 0.06%) while urban inequality slightly declines (by 0.01%). The increase in
rural inequality is a consequence of the increase in the returns to land, which rises more
than the return to unskilled labour. The decline in urban inequality arises from the
27
A spreadsheet was used to approximate the lognormal distribution by dividing households in each
broad group into over 200 sub-groups defined in terms of narrow income bands. Initially, arbitrary values
of the rural and urban standard deviations, h, were imposed and the proportion of households in each
broad group within each narrow income band was estimated, to de!rive h for each broad group from
the imposed value of h and the data base values of the actual arithmetic mean of household income for
the broad group. Given the estimated proportions of each household group in each narrow income
band, it was straightforward to derive the implied values of the national headcount poverty rate and Gini
coefficient. A search was then conducted over the values of the rural and urban standard deviations to
find the ones that reproduced the actual national headcount poverty rate and Gini coefficient.
28
Using the same spreadsheet approach and the same estimated standard deviations that are
described above, it was straightforward to estimate the changes in the headcount poverty rate, the
poverty gap and the Gini coefficient due to changes in the arithmetic means of the real incomes of each
household group that were implied by the simulation results for each shock analyzed.
29
Which assumes import supply elasticity of 10, elasticity of substitution between domestic and
imported rice of 6, and a domestic supply elasticity of 0.3
33
Poverty and Social Impact Analysis: Indonesia Rice Tariff
increase in unskilled wages relative to skilled wages and from a decline in the return to
capital.
How does the increase in tariff affect welfare?
115. The increased tariff affects producer prices, consumer prices, and therefore
incomes (through returns to farming, and increased wage incomes in the rural area).
116. Firstly, an increase in the tariff will raise the import price of rice. This will
subsequently decline as the volume of imports contracts, but not enough to prevent the
domestic price—both the producer price and the consumer price—from increasing. The
power of the tariff, given by T = (1 + t), where t is the proportional rate of the tariff, rises
from 1.25 to 1.45, a proportional increase of 16%. If import prices did not change, the
landed price of imports would thus increase by 16%, but the c.i.f. import price declines by
4.3% as a result of the large decline in the volume of imports. As a result, the increase
in the domestic price of imported rice is 11.7%.
117. The consumer price of rice increases, but by much less than the 11.7% increase in
the price of imports because imports and domestically produced rice are imperfect
(though relatively close) substitutes. The price of domestically produced rice increases by
1.88%. The rice entering the consumption basket is a composite of this and imported
rice, where the share of imports in consumption in the data base of the model is just over
7%. The consumer price of rice thus increases by 2.5% [=0.93 * (1.88) + 0.07 *
(11.7)].
118. Since the share of rice in the average consumer’s budget is 7.5% (higher for poor
groups), the increase in the cost of living for the average household is 0.188%
(=0.075 * (2.5)). Note that empirical estimates for the poorest groups find that rice
represents around a third of their total expenditure, and therefore the increase in cost of
living would be around 0.8%. The final increase in the consumer price index induced by
the rice tariff rising is 0.25%.
119. The increase in the domestic price stimulates rice production (by 0.53%) and
reduces consumption. The increase in rice production induces an increase in unskilled
wages relative to skilled wages. The reason for this is that rice production uses large
quantities of unskilled labour, but very little skilled labour. Nominal skilled wages rise
by 0.09% whereas unskilled wages increase by 0.21%.
The effects differ considerably among poor households.
120. Poor households are not affected uniformly. For some households the income
effects resulting from a tariff increase outweigh the increase in their cost of living.
Inspection of the factor ownership data in Table 4.1 indicates that these results are
strongly related to the share of the household’s income that is derived from the returns to
unskilled labour. It will be recalled that the returns to this factor increase with the tariff.
Those households with high shares of income deriving from unskilled labour gain from
the tariff increase. Those with lower shares lose from it, but the losers outnumber the
gainers. Moreover, the fact that the size of the poverty gap increases indicates that
among the poor the magnitude of the losses outweighs the magnitude of the gains.
121. All of the urban households are poorer after the tariff increases, though the
differences are small (see table below). In rural areas, the results are mixed, though for
the two categories with the highest poverty rates, agricultural employees (landless
34
CGE Model and Results
households) and non-farm labourers, there is a slight drop in poverty (which is due to the
increase in unskilled wage rates).
TABLE 4.1 POVERTY RATES BY HOUSEHOLD TYPE, BEFORE AND AFTER TARIFF
RISE
Household type
Agricultural Employees
Small Farmers
Medium Farmers
Large Farmers
Rural low income (non-farm)
Rural non-labour
Rural high income
Urban low income
Urban non-labour
Urban high income
Poverty rate at current
tariff level (%)
Poverty rate after
tariff rise (%)
38.87
15.14
10.53
1.52
16.53
0.46
0.35
21.30
15.41
0.78
38.86
15.19
10.49
1.51
16.48
0.46
0.34
21.31
15.41
0.79
122. The effects of a tariff are primarily redistributive, causing redistribution from
consumers to producers. Its effects on national aggregate income or GDP are very small
indeed, and for example, do not indicate a slump in the economy.
123. In summary, although real wages rise in response to the expansion of rice
production, this effect on the incomes of the poor is quantitatively less important than
their increased living costs, resulting in a reduction in the real expenditures of the poor
and an increase in poverty incidence.
124. Poverty incidence increases in both rural areas and urban areas, but the increase
in urban areas is larger. The urban poor benefit less from the increase in unskilled wages
than do the rural poor and receive almost no benefit from the increased return to land.
125. Whilst the macroeconomic impacts of the tariff change from the model appear to be
insignificant, the model results can add value, by offering up areas for further
investigation. For example, to expand the analysis to a regional level in a further PSIA
exercise, research can focus on the key factors affecting welfare: rice consumption
behaviour, proportion of household income from unskilled labour, likely supply response
of rice farmers. This should be combined with an institutional analysis of factors that may
constrain the benefits or costs accruing to households (for example, if the tariff revenue
does not enter government accounts, or if the structure of the rice market prevents
wages /income from rice rising).
126. The following chapter outlines some of the extensions that could be made to the
model, some technical criticisms and testing of assumptions.
35
5 Criticism, Variations and Limitations of the CGE
Model
This chapter offers a discussion of the limitations of the model and a critique of the
results. CGE models offer theoretical rigour, consistency of assumptions and analysis of
complex interactions and 'knock-on' effects between sectors of the economy. In this PSIA
exercise, it is very useful to analyse such effects and their impact on poverty. However,
any CGE model must include some problematic assumptions, and is limited in that it can
only offer insight into the medium-term effects on the economy.
5.1
ASSUMPTIONS OF THE MODEL
127. The technical structure of the CGE model used in this analysis was described in
section 4.1. As discussed above, the value-added of CGE models is that they can
explicitly take into account the fact that changes to policy in one sector of the economy
will have a 'second-round' effect both on that sector and on other sectors. For example in
the context of the rice import tariff, an increase in the price of rice will affect both
consumer spending on rice and other goods, and also on levels of rice production and
thus wages in the rural sector. This is both a strength of CGE models and a weakness.
The complicated nature of the modelling process means that a series of assumptions are
imposed on the model, such as: equilibrium in goods and factor markets that always
clear, constant elasticity of substitution production technology, and a fixed fiscal and
trade balance.
Closure Conditions
128. Knock-on effects of the economy continue into infinity. In order to stop the CGE
model iterating infinitely, the modeller must impose closure conditions on the model.
Since household consumption within the single-period horizon of the model is chosen as
the welfare indicator, and is the basis for the calculation of poverty incidence, the
macroeconomic closure must be made compatible with this measure. This is done by
ensuring that the full economic effects of the shocks to be introduced are channelled into
current-period household consumption and do not 'leak' into other directions, with realworld intertemporal welfare implications not captured by the welfare measure. The choice
of macroeconomic closure may thus be seen in part as a mechanism for minimising
inconsistencies between the use of a single-period model to analyse welfare results and
the multi-period reality that the model represents.
129. To prevent intertemporal and other welfare leakages from occurring, the
simulations are conducted with balanced trade (exogenous balance on current account).
This ensures that the potential benefits from the export tax do not flow to foreigners,
through a current account surplus, or that increases in domestic consumption are not
achieved at the expense of borrowing from abroad in the case of a current account
deficit. For the same reason, real government spending and real investment demand for
each good are each held fixed exogenously. The government budget deficit is held fixed
in nominal terms. This is achieved by endogenous across-the-board adjustments to
personal income tax rates so as to restore the base level of the budgetary deficit.
130. The combined effect of these features of the closure is that the full effects of the
tariff increase are channelled into household consumption and not into effects which are
not captured within the single period focus of the model. However this does not reflect
Criticism, Variations and Limitations of the CGE Model
reality in important ways. Firstly, the tariff would raise a certain amount of government
revenue. The way this revenue was spent by the government would influence the net
distributional effects of the tariff. In the model, increases in revenue from the tariff are
directly offset by reductions in income tax rates. Similarly, it is unlikely that a change in
tariff policy would not affect the trade balance. These two important effects should be
borne in mind when interpreting the results.
131. In the various simulations of the model, results showed that government revenue
could increase by around 0.2% (through a combination of increased tariff revenue from
the rice import duty, and lower export duty revenue). To balance the budget to presimulation levels, the total income tax revenue is reduced by 0.2% through a cut across
the board in tax rates.
Behavioural equations
132. Further aspects of the CGE model are also unrealistic, though are frequently used
in both CGE and other econometric models, for example the structure of production
functions, efficiency in markets, flexible prices and markets that clear. Economists are
often divided on the validity of making such assumptions. The important question is
whether the assumptions can approximate the reality of an economy enough to draw
inference on what the likely impact of a policy change will be. A particular criticism of the
model used in this analysis is that a number of the parameters in key behavioural
relationships are taken from econometric analysis of the Thai economy (Warr, 2000). In
the time allowed for the PSIA exercise, it has not been possible to examine in detail how
valid this procedure is.
Poverty assumptions
133. As with all economic models, the welfare indicator is household consumption. This
does not take into account other indicators of poverty as discussed in Chapter 3 such as
access to basic social services, concepts of capability, security or empowerment. The
advantage of the CGE model is that it is able to disaggregate welfare changes between
heterogeneous household types, in particular related to the way that households make
their livelihoods, as shown above in the Results section. The poverty calculations are
based on assumptions that can be challenged. In this model, 10 representative
household types are used. For each of the 10 types, an income distribution for all
households of the type is assumed. In particular, the distribution of income is assumed to
be lognormal, and also that the standard deviation of rural households is the same in all
groups, likewise for urban households. This is not an unusual assumption in a CGE
model, but clearly does not offer insights into intra-group welfare levels. However, there
has been progress in assessing welfare at the household level without imposing such
rigid assumptions on income distribution, by integrating the database of the CGE model
(usually the SAM, or Input-Output table for the macro-economy) with information on
household income/consumption at the individual level (usually from a national household
survey such as the World Bank Living Standards Measurement Survey) 30. Again, in the
time available, it has not been possible to conduct a more detailed analysis of poverty.
5.2
TIMEFRAME OF THE CGE MODEL
134. As discussed in the introduction, the CGE model results offer a static simulation of
the second-round effects on the economy of an increase in the rice import tariff. The final
result assumes that the impact of the price change is felt by consumers, and that
30
An excellent reference is Bourgignon et al 1999.
37
Poverty and Social Impact Analysis: Indonesia Rice Tariff
producers of rice (and other goods affected by the change in relative prices) have
responded by increasing production to a certain extent, and that unskilled wages have
also risen by a certain amount. The model does not specify how long this transition may
take (though an educated guess would be 6–18 months), nor the intermediate stages
that the economy (and Indonesian society) will go through in order to reach the endpoint.
This section briefly outlines the impact on the economy in the short run, and in the longer
term.
135. Short run: In the short run producers have not yet responded to the increase in
rice prices induced by the tariff increase. Therefore the welfare impact will be the effect of
the increase on consumers price of rice, and the revenue effects to farmers. Essentially,
the net consumers of rice will suffer, and net producers will benefit, as outlined in Ikhsan
(1999). As discussed in Ikhsan, net consumers of rice outnumber net producers of rice by
a ratio of approximately 3:1. Rice comprises a higher share of expenditure of the poor
and therefore will have a significant impact on poverty. Policy makers should consider
options to deal with the transition period after the tariff increase when potentially
significant numbers of people will fall below the poverty line.
136. Long run: Debate on the long-run implications of the tariff increase go beyond the
scope of this exercise. Economic theory implies that it is not efficient to permanently and
increase domestic wages and unskilled employment by protecting labour-intensive
industries in which a country has no comparative advantage. The argument is that
resources are devoted into sectors of the economy that are uncompetitive, this slows
growth prospects, and consumers suffer through having to pay higher prices. If Indonesia
has a comparative advantage in labour intensive production it should deploy its labour in
those products in which it has such an advantage rather than in the rice industry.
5.3
DATASET USED IN THE CGE MODELLING EXERCISE
137. As described in section 2 of this chapter, the latest complete dataset available is
the 1995 INDOSAM SAM with some modifications. Indonesia has experienced extreme
social and economic change in the period since 1995, and Chapter 3 of this report
attempts to quantify the impacts of the crisis on current poverty and livelihoods data. If
further time were available, more explicit analysis detailing whether key structural
parameters in the economy had changed since 1995 would strengthen the results of the
CGE analysis.
5.4
EFFECTS OF VARYING KEY PARAMETERS
138. Key point: Varying three key parameters in the CGE model within a plausible
range does not significantly affect the results of the model.
139. To what extent do the results summarised above depend on the assumed values of
key parameters? This question is important, because the above discussion indicates that
there is considerable uncertainty surrounding the true values of several parameters that
seem particularly relevant for the results. These include: the elasticity of supply of rice
imports to Indonesia; the elasticity of supply response of paddy with respect to its price;
and the elasticity of substitution in demand between domestically produced and imported
rice.
38
Criticism, Variations and Limitations of the CGE Model
The Elasticity of Supply of Rice Imports
140. Simulation A assumes that imports of rice are available to Indonesia with an
elasticity of supply of 10. This means that a 10% increase in the volume of Indonesia’s
imports induces a 1% increase in the international price. Simulations using values of 2.5,
5 and 20 for this parameter are considered and elaborated in Appendix A. The
implications for poverty incidence at the national, rural and urban levels are summarised
in Figure 3. For values of this elasticity in excess of about 3, poverty incidence rises and
for values less than this, it falls. The reason is that the lower the elasticity of supply of
imports the greater is the terms of trade gain from a given tariff.
141. Econometric estimates of the supply of imported rice to Indonesia have apparently
not been undertaken, but a closely related question has been studied at length. This is
the elasticity of demand for rice on the world market for the world’s largest exporter,
Thailand. The direct connection between these two matters arises as follows. Suppose
first that Thailand exported one million tons additional rice onto the world market. The
world price would fall, somewhat. Now suppose that Indonesia imported one million tons
less rice from the world market. Again, the world price would fall, and the effect would be
virtually identical to that resulting from the increase in Thailand’s exports. Indeed,
because Indonesia’s rice imports come primarily from Thailand, the types of rice involved
are essentially the same.
142. Studies of the elasticity of demand for Thailand’s rice exports have produced
estimates ranging from -2.5 to -5.31 If the volume of Indonesia’s imports was the same as
the volume of Thailand’s exports, the elasticity of supply of rice imports to Indonesia
would be the same as this but with the opposite sign. Over the three years 1998 to 2000,
Indonesia’s rice imports have been about 70% of the level of Thailand’s rice exports,
implying elasticities of supply of 3.6 to 7.2.
143. The central problem with this analysis, however, and with the econometric studies
on which it are based, is that the estimated elasticities almost certainly understate the
true long run elasticities of supply. The reason is that if the world price were to rice, say
because a major importer like Indonesia restricted its imports, relative to the level they
would otherwise have taken, new suppliers would almost certainly enter the world
market. But because these suppliers are not exporters at current world prices, their
supply behavior is not reflected in available statistical data. It would seem likely that a
reasonable estimate of the long run elasticity of supply of rice imports to Indonesia would
be between 7 and 10. Nevertheless, the true value of the long run elasticity of supply of
rice imports to Indonesia must be considered uncertain.
The elasticity of supply response of paddy with respect to its price
144. The results of our baseline model use an elasticity of supply response of 0.31,
which means that for every ten% increase in the price of rice, farmers will increase
production of rice by three%. Simulations in Annex A illustrate the effect of varying the
supply response on the results. Poverty incidence increases throughout the range of
variation. As expected, lower values of supply response imply larger increases in poverty
incidence, but implausibly large elasticities of supply response (well outside the range
considered here) would be required to turn the increase in poverty incidence into a
reduction.
31
This literature is reviewed in detail in P. G. Warr 'Welfare Effects of an Export Tax: Thailand’s Rice
Premium', American Journal of Agricultural Economics, vol. 83 (4), (November 2001), 903-920.
39
Poverty and Social Impact Analysis: Indonesia Rice Tariff
145. An important point to note is that in the very short run, the supply response to an
increase in price will be zero, and therefore the effect on poverty will be directly related to
the price rise impact on consumers, as summarized in Ikhsan (1999). The nature of crop
production is that supply response generally occurs only with some delay—say, six
months to two years. So long as it remains in place, a tariff increases the domestic price
permanently. How would Indonesian producers respond? It seems likely that the long run
supply response in the Indonesian rice industry would be highly inelastic, but this does
not mean that it would be zero.
146. Several empirical studies have looked at the issue of supply response in the
Indonesian context and their results vary widely. An early study by Mubyarto (1975)
estimated the long run elasticity of planting area with respect to price on Java to be very
low, at 0.03. Tabor (1988) estimated that in Java the elasticity of planting area with
respect to price was 0.22 in wet land rice production and 0.45 in dry land production. A
study by Hutauruk (1996) estimated the planting area response elasticity on Java to be
0.04 and off Java to be 0.78. Since the overall elasticity of supply includes the response
of yield to price as well as the response of planted area, the implied output supply
elasticities with respect to price will be larger than these estimates.
147. Finally, a recent paper by Irawan (2002) estimates short and long term elasticities
of supply response for several regions and for both wet and dry land rice production. The
short-term estimates for wet land rice are: Java 0.11, Sumatra 0.12, Sulawesi 0.45 and
Kalimantan 0.02. His long-run estimates are: Java 0.13, Sumatra 0.52, Sulawesi 1.25
and Kalimantan 0.21. His estimates for dry land rice supply response are generally about
50% larger than the above estimates. For example, the long run estimate for dry land rice
supply response for Java is 0.21 and for Sulawesi it is above 2.
148. In summary, the available econometric evidence supports the view that in
Indonesia the overall elasticity of supply response of rice is low, but not zero. The
estimates are higher in the long run than the short run, higher in dry land conditions than
wet land conditions and generally higher off-Java than on-Java. Estimates of the long run
elasticity of output with respect to price in the range of 0.2 to 0.4 would be consistent with
the available evidence. Nevertheless, it must be recognised that considerable uncertainty
remains as to the true value of this key parameter.
The Armington Elasticity of Substitution in Rice Demand
149. The small effect that a rice tariff has on the domestic price derives in part from the
size of the Armington elasticity that we have assumed. Although the assumed value of 6
is quite high, its value could certainly be questioned. Simulations in Appendix A show the
effects of varying this parameter across the range 2 to 10, since empirical work for the
Philippines found that this was a plausible range of variation32. The results on poverty
incidence are summarised in Figure 5. The higher the assumed Armingtion elasticity, the
larger the effect of the tariff on the domestic price. This in turn magnifies the poverty
increasing effect of the tariff. Variations in the assumed Armington elasticities will not turn
the simulated poverty increase into a reduction in poverty, nor will they turn a ‘small’
increase in poverty incidence into a ‘large’ one. This is mainly because imported rice is
only approximately 7% of total rice consumed in Indonesia (see results section).
32
Kapuscinski, C and P Warr (1999) 'Estimation of Armington Elasticities: An Application to the
Philippines', Economic Modelling, vol. 16.
40
Criticism, Variations and Limitations of the CGE Model
5.5
CONCLUSIONS
150. This section has presented the results of running the CGE model to simulate an ad
valorem increase in the rice import tariff from 25% to 45%. The model showed that
through the channels of consumer price increase, increased rice production, and
unskilled wage increases, poverty would increase very slightly in the medium-term (6–18
months). The policy implication of this is that increasing the rice tariff is not an effective
way to reduce poverty in the medium-term.
151. The CGE model offers information on the medium-run effects of a policy change
only. In the short term, it must be borne in mind that an increase in the rice price would
definitely increase poverty, since the consumption effect would dominate. In the long
term, it can also be argued that tariffs have a distorting effect on resource allocation in
the domestic economy, though an alternative view supported by important groups in the
Indonesian policy context is that tariffs can protect domestic sectors whilst investing in
productivity improvements. The CGE model cannot offer insights into these dynamic
processes behind longer term growth strategies.
152. With these limitations in mind, the results presented here can be considered
relatively robust, since varying the key parameters of the model do not change the results
significantly. With the time available to complete the pilot PSIA, this has been the scope
of the analysis. However, with more time or in a follow-up study it is suggested that the
results from the CGE should be combined with more qualitative analysis on
disaggregated livelihoods and further political/institutional analysis, building on the PIM.
This would strengthen the model’s conclusions and place it more in context. Further
extensions to the analysis given more time could also include incorporating government
services (human development outcomes), dealing with uncertainty in the agricultural
sector, and a deeper treatment of the institutional aspects of rice markets in Indonesia.
153. The final chapter of this study concludes, and makes some initial connections
between the results of the CGE analysis and the political economy of Indonesia, and
suggestions for the future of PSIA in the Indonesian context.
41
6 Implications and Recommendations
6.1
THE POLITICAL ECONOMY OF DECISION-MAKING
154. This chapter links the results of the PSIA findings with the policy making
environment in Indonesia. As Chapters 4 and 5 have shown, though the overall net effect
of a rice tariff increase is small, in the short term it will significantly increase poverty at a
time that the Indonesia government is under pressure to reduce poverty to pre Asia crisis
levels quickly.
155. The CGE modelling suggests that raising tariffs will not protect poor farmers and
landless agricultural labourers’ economic interests, in either the short term or medium
term that some policy makers have suggested. They are all net rice consumers who will
not benefit from higher tariffs on cheaper imported rice. Those in urban areas, both the
poor and well off—who are also rice consumers—will also not benefit. The results of the
CGE model shows that an increased tariff, intended to protect the rice industry, will deny
certain groups of poor people cheaper and more affordable foodstuff. Those who benefit
in the short and medium term are the medium and larger farmers in rural areas. These
are also likely to include those who are rice traders and importers who are most
influential in the formulation of rice trading policy.
156. The model cannot predict when the second round effects may come into play in the
future that is expected to have a knock on effect to raise wages in the rural economy,
though the assumption is that, ceteris paribus, the medium-term is 6–24 months after the
tariff hike. However, in the shifting economic environment in Indonesia, the projected
medium term may never happen as modelled, for example if the increase in rice prices
leads to political instability. Though the model cannot provide a disaggregated analysis of
the regional effects of an increase in rice tariff, those regions likely to gain are the large
rice producing regions of South Sulawesi and East Java. In a further study, it would be
helpful to investigate further the regional implications of the tariff, if the proposal remains
a live issue.
157. Given that the model results indicate who is likely to benefit from an increase in
tariffs, is it useful to return to the policy interests matrix of Chapter 2. This matrix, based
on published reports and interviews, was intended to show the perspectives of the key
players and institutions on the rice tariffs debate. The two most powerful government
departments advocating an increase in tariffs are the Ministry of Agriculture and BULOG,
the commodities logistic agency responsible for rice. The two departments advocating
low or no tariff are the Ministry of the Economy (low tariff) and BAPPENAS (no tariff). The
more powerful forces in the government—identified through interviews—were seen to be
those ministries advocating a higher tariff.
158. This PSIA has shown the challenges of linking existing sound data and research
with government policy-making. Currently in Indonesia there appears to be no
mechanism to systematically link together sectoral approaches to poverty into an overall
framework of macro-economic policy analysis. The CGE model makes a contribution to
the debate on rice tariffs, and has broadened the received wisdom of some agencies
(Department of Agriculture for example) by relating rice tariffs to general poverty impacts
for both rice producers and consumers.
Implications and Recommendations
159. The review of policy pronouncements by key stakeholders in the policy process
shows the politicised nature of policy making in Indonesia, as in many countries. Despite
research data provided by academics and independent think tanks, politics and political
and economic interests tend to dominate the policy making process.
6.2
METHODOLOGY PROCESS
160. The challenge of the PSIA is to provide a process for both government and civil
society to make sound decisions based on evidence, and to increase ownership and
transparency around a choice of policy options. The PSIA, using both the CGE modelling
and PIM, provides data and analysis that will help in policy decision-making. But it does
not provide a way to make decisions that balances political and economic interests.
161. The process of undertaking the PSIA in Indonesia included extensive consultation
with civil society, which increased transparency, their own voice and commitment to
participation in decision-making on this particular policy issue. The government officials
involved in this study, on the other hand, tended not to consider poverty as part of their
central remit when making policy decisions. Like a number of other social issues, poverty
has become a ‘popular’ issue in political rhetoric, yet it is still difficult to integrate practical
solutions based on sound analysis and evidence into the policy cycle. An obvious lesson
learned from this PSIA is that the incentives are not in place to enable policy-makers to
prioritise poverty concerns.
162. The Indonesia PSIA had limited success with the 'reference group' as suggested by
the World Bank methodology, made up of civil society and government to champion the
PSIA and ensure the results are taken up. Membership and commitment was sought
from the members of the KPK, responsible for the PRSP, with agreements for
involvement. The group members were invited to the final presentation of results, but few
from outside academia and the donor community attended the October meeting.
Interdepartmental forums held routinely under this task force to discuss the impact of
reform on poverty ex post will help the uptake and application of PSIA and ex ante work.
163. Given that policy decisions are highly political, and politicians—especially those in
Parliament—are now the key players, they should be participating at committee levels in
policy analysis discussions. In the past, decision-making was in the hands of the ruling
executive body. As an emerging democracy, the role of parliament or the legislative body
has become very significant, and therefore no single crucial or strategic policy can be
implemented without their agreement. In many instances, lobbying the legislators
becomes important. Without their involvement and agreement, there is little chance for
research to effectively influence policy choices. A necessary next step in Indonesia is to
understand how to translate the results into informed policy-making, and to understand
the Indonesian 'culture of decision-making' at this particular moment.
Tools and triangulation
164. Because of time constraints the PSIA was led by the CGE modelling, to quickly
assess how a projected increases in rice tariffs would affect the various groups of poor
people. This was only possible because the CGE model already existed. The benefit of
the model was to focus on both consumption and production of rice issues over time.
However, in this CGE framework there was no possibility of assessing the gender, age
and regional dimensions of a rise in rice tariffs. Though there was considerable
qualitative data available on the impact of rapid rises in rice prices from the Asia financial
crisis (1997–99) there was little opportunity to use in-depth case studies resulting from
43
Poverty and Social Impact Analysis: Indonesia Rice Tariff
the analysis, which may have provided more of a descriptive feel for the various
household types used in the CGE analysis.
165. Because CGE was the starting point, there was a 'tool box' approach to the PSIA.
By piloting the use of the PIM, and attempting to identify the politics of policy
perspectives as key issues, this PSIA went further to augment and enrich the outcomes
of a CGE-led PSIA. However, in the time available, this approach had to run in parallel
with the CGE model. With more time, it would have been possible for a sequenced
approach, with lessons learned from the CGE feeding into the political analysis, and
likewise, the political institutional analysis. More qualitative work could complement and
influence further development of the CGE model. This would be the suggested starting
point to extend the scope of the results, for example, to analyse regional or age or
gender dimensions of the policy change.
6.3
IMPLEMENTATION OF PSIA—DEMAND
166. The weak demand for the PSIA study at the moment reflects little incentive within
Indonesian government decision makers to use PSIA results for macro-economic policy
making. The timing of the PSIA so soon after the appearance of the I-PRSP resulted in
some 'poverty' fatigue among the Poverty Task Force members who were the core of the
reference group, and who possibly would be more receptive at a later stage in the PRSP
process. With greater involvement in both the substance and the process, the PSIA
would have more persuasive advocates within powerful areas of government.
167. There is some optimism that PSIA will be taken up by academics, NGOs and other
civil society groups to enable and increase their influence in policy decision-making. The
policy mapping done with civil society groups that contributed to the matrix led to greater
'in house' understanding of the various interests involved in decision making and a
recognition of where natural coalitions lay for more effective influencing. The process
increased transparency and functioned as a learning tool—and should enable civil
society organisations in the future to become better advocates on specific policy issues.
Academics and think tanks are well placed to take up PSIA as a method that will get their
results out into the policy arenas. Traditionally these groups have not been advocates,
letting their work speak for itself. However, PSIA could provide a mechanism for
academia to be more vocal in policy decisions.
Timing
168. Time pressure was the most serious constraint to achieving the ambitious aims of
the pilot study. The quick and dirty approach to the PSIA allowed researchers to analyse
the main drivers of welfare changes from the tariff, through a critical analysis of the CGE
model results, and to do a rapid stock-take of political interests using the policy interest
matrix. However, the speed at which the research had to be undertaken did constrain the
depth to which both qualitative and quantitative aspects of the work could be pursued,
and did not allow enough time to properly embed the PSIA into the PRSP process.
However, it is recognised that this is a pilot study, and the difficulties outlined here should
be used to inform the rollout of PSIA practice in future.
169. In keeping with the agreed PSIA approach, the choice of a policy was done with a
week’s consultation involving a range of decision makers who focused on the 'hot ' topic
at hand. The various stakeholders could, with more time, have been more involved with
the PSIA, but finally were key informants rather than key players.
44
Implications and Recommendations
170. Before the completion of the PSIA there were newspaper reports that a new policy
on rice tariffs was about to be announced, which would have lessened the potential
impact of the PSIA. In any event, no announcement to raise the tariffs was ever made. A
slower approach and broader consultation would have increased commitment to the
PSIA process and results.
171. Short time frames led to an emphasis on an analytical tool (CGE modelling) that
was at hand, available from an international consultant. His availability at short notice
was limited, and little opportunity arose to teach the model to other Indonesian
colleagues who would have benefited from a closer working partnership.
172. The CGE report was completed only hours before it was to be presented in the final
workshop in order to meet the timing for the World Bank meeting in October. A longer
time horizon and emphasis on process is needed, and in this case a greater working
relationship between the Task Force on Poverty, the consultants, think tank and
academics to increase the understanding and relevance of the use of PSIA. Given the
population size and geographical spread in Indonesia (220 million people spread over
17,000 islands in three time zones) consultation was with national representatives only.
Skill base
173. PSIA provides a unique opportunity to bring together researchers and policy
makers around a topic of policy relevance where applying research is necessary. Much
policy relevant research is left on the shelf and many academic researchers are reluctant
to become advocates. Government is impatient to get on with it with little attention to the
impacts their policies may have on poverty. The PSIA process can bring these two
groups to work together in ways that were not done in Indonesia because of time, but
could be.
174. The national institution SMERU was an excellent institutional partner. They
suggested the international consultant who had some of the skills in modelling they
themselves lacked. However, the development of the CGE model by the international
consultant meant that the role of the national intuition was more limited to provision of
data as requested and discussion of results. Capacity to work with the model or to use
CGE for other purposes was very limited. As a method it has considerable data
requirements, is time-consuming and complex, and the results are sensitive to the
assumption in the model. CGE models are a seductive way to make poverty and social
outcomes of macroeconomic policy more explicit. However, as Chapter 5 indicated, there
were a number of constraints embedded in the modelling.
175. Given timing constraints, the work of the international consultant was augmented
by a DFID economist. No international sociologist was hired, and this role, as well as
team leading, was undertaken by a DFID sociologist. In the interests of time, this was a
good solution, since there was clarity on the objectives of the PSIA, but again
constrained by the limited amount of DFID staff time that was available.
45
Poverty and Social Impact Analysis: Indonesia Rice Tariff
6.4
INSTITUTIONAL OPTIONS FOR PSIA IMPLEMENTATION
Who will conduct PSIAs?
176. Much careful thought needs to be given to institutionalising PSIA. Currently it is
donor-driven, but has the potential to be taken up if appropriately situated. The pilot is
only a demonstration of what is possible. There appear to be few incentives or
conditionalities attached to either the PRSP or PSIA, and in the highly political policy
making environment little incentive to use PSIA at this point. More work needs to be done
to promote the PSIA approach with government, which has not been possible yet given
the short period of time for the PSIA. The PSIA from the outset has been linked to the
PRSP process, involving key players, part of the Poverty Eradication Task Force, which
became the reference group for the PSIA.
177. In addition to academic studies such as the CGE, decision-makers also need to be
lobbied. Such lobbying can only occur if there are organisations and individuals who have
a strong commitment to the issues being studied. Similarly, they need to have access to
the elite group of decision-makers to have any real effect.
178. The formation of a reference group or committee made up of various stakeholders
(as mentioned above) could be the first strategic step carried out in the process which will
help to institutionalise the PSIA study within these elite groups. If this group is committed
with time and resources and thus well involved in the PSIA from the beginning, they will
feel some sense of ownership and pride in the process and results of the study and
therefore are more like to be successful in influencing policy decisions. If, for example,
they consider the PSIA process and results to be alien and not belonging to them they
will have no obligation to fight for their input to be included in policy design.
179. Will stakeholders take on the CGE or PIM? There is some doubt about the
capabilities of any government department or think tank to undertake other CGE analysis
in Indonesia. Without further skills development CGE will remain in the hands of
international consultants and thus have limited applicability in the future as a general tool
for PSIA. A PIM is a very user-friendly inclusive and participatory tool that will be able to
be used by stakeholders in most PSIAs.
6.5
RECOMMENDATIONS




46
Involve a wider range of stakeholders, more than were involved in the rice trade
policy PSIA. The stakeholders should be involved in the process right from the
beginning, and they should be used as more than just resource people.
Implementation could be improved by forming a committee that is able to provide
both input and direction from the outset. This needs more time and more
emphasis on process.
Representatives of this committee should have the seniority that would allow
them to at least influence or become involved in decision making processes in
their respective organisations or departments.
There needs to be intensive working collaboration between consultants and local
partners. In addition, there needs to be a transfer of knowledge between
consultants and local partners and the reverse.
A number of different kinds of training for government and advocacy groups
involved in the consultation process is required in order to apply PSIA effectively.
Implications and Recommendations
No matter how reliable the analysis, it will be very difficult to convince policy
makers to utilise the results without this training.
Donors and civil society should be more involved in the process so PSIA is given
more attention.
There are many consulting firms, independent consultants and universities that
have carried out studies on the impact of macro policies on poverty. The results
of their studies could be improved if a peer review system could be established
by the forum under KPK (as mentioned above) and financed by the donors who
commission the study. Thus, PSIA can be applied more effectively.
PSIA methods relying heavily on the existing studies could be modified to include
original studies if time allows, as indicated by initial findings.
The PSIA process should allow more time for its completion than was allowed in
Indonesia. This will improve the quality of analysis, and allow for triangulation and
lesson learning. Also, for the intended objectives to be achieved, more time
needs to be spent institutionalising PSIA into the PRSP process (as stated
above) and the PSIA should be followed through the whole of the policy
implementation phase.




FIGURE 6.1 REAL PRICE OF RICE, INDONESIA, 1969 TO 2001
Price of rice/ CPI
1600.0
1400.0
1200.0
1000.0
800.0
600.0
400.0
Jan-01
Jan-99
Jan-97
Jan-95
Jan-93
Jan-91
Jan-89
Jan-87
Jan-85
Jan-83
Jan-81
Jan-79
Jan-77
Jan-75
Jan-73
Jan-69
0.0
Jan-71
200.0
Source: Bulog (rice prices) and BPS (CPI).
47
Poverty and Social Impact Analysis: Indonesia Rice Tariff
FIGURE 6.2 WORLD PRICE AND DOMESTIC PRICE OF RICE, INDONESIA, 1985 TO
2002
Price of rice Rp./kg
6000
5000
4000
3000
2000
World Price
Jan-02
Jan-01
Jan-00
Jan-99
Jan-98
Jan-97
Jan-96
Jan-95
Jan-94
Jan-93
Jan-92
Jan-91
Jan-90
Jan-89
Jan-88
Jan-87
Jan-86
0
Jan-85
1000
Domestic Price
Source: Bulog (rice prices) and BPS (exchange rates).
FIGURE 6.3 SIMULATED CHANGES IN POVERTY INCIDENCE: VARYING ELASTICITY
OF IMPORT SUPPLY OF RICE
Change in pverty incidence (%)
0.1
0.08
0.06
Rural
Nation
Urban
0.04
0.02
0
-0.02 0
2.5
5
10
Elasticity of import supply of rice
Source: Author’s computations.
48
20
Implications and Recommendations
FIGURE 6.4 SIMULATED CHANGES IN POVERTY INCIDENCE: VARYING ELASTICITY
OF SUBSTITUTION IN RICE PRODUCTION
0.09
Change in poverty incidence (%)
0.08
0.07
0.06
Rural
Nation
Urban
0.05
0.04
0.03
0.02
0.01
0.00
0
0.15
0.20
0.25
0.30
0.40
Elasticity of substitution in rice production
Source: Author’s computations.
FIGURE 6.5 SIMULATED CHANGES IN POVERTY INCIDENCE: VARYING ARMINGTON
ELASTICITY OF SUBSTITUTION IN RICE DEMAND
Change in poverty incidence (%)
0.08
0.07
0.06
0.05
Rural
Nation
Urban
0.04
0.03
0.02
0.01
0.00
0
2
4
6
8
10
Armington elasticity of rice
Source: Author’s computations.
49
Poverty and Social Impact Analysis: Indonesia Rice Tariff
TABLE 6.1 INDONESIA: RICE PRODUCTION, CONSUMPTION AND TRADE, 1990 TO
2001
Harvested
Gabah
Rice
Rice
area
production
production
consumption
Volume
Value
(Ha)
(GKG) - (Ton)
(Ton)
(Million Ton)
(Ton)
(000 US$)
1990
10,502,357
45,178,751
28,552,971
24.41
6,378
2,907
1991
10,281,519
44,688,247
28,242,972
24.70
168,933
52,476
1992
11,103,317
48,240,009
30,487,686
24.99
566,441
159,049
1993
11,012,776
48,181,087
30,450,447
25.42
3,093
1,269
1994
10,733,830
46,641,524
29,477,443
25.85
268,802
68,736
1995
11,438,764
49,744,140
31,438,296
26.28
1,306,218
374,101
1996
11,569,729
51,101,506
32,296,152
26.16
2,040,203
731,055
1997
11,140,594
49,377,054
31,206,298
26.55
1,095
5,349
1998
11,716,499
49,199,844
31,094,301
26.86
2,793,907
831,763
1999
11,963,204
50,866,387
32,147,557
27.29
3,055,414
817,591
2000
11,793,475
51,898,852
32,800,074
27.72
3,502,090
925,562
11,419,935*)
50,197,883*)
31,725,062*)
27.97
644,732
134,912
2001
Note:
Import
Assumed rate of gabah conversion to rice = 63.2%
Consumption is estimated based on the total population and estimates of per capita
consumption.
Source: Dept. of Agriculture, Jakarta.
TABLE 6.2 WORLD RICE TRADE BY COUNTRY
Country
Export
Taiwan
Myanmar
EU
Argentina
Uruguay
Australia
Japan
Pakistan
USA
Vietnam
China
India
Thailand
Others
Total export
50
1998
000 ton
55
98
346
500
625
641
642
1,841
3,211
3,774
3,965
5,305
6,389
633
28,025
%
0.2
0.3
1.2
1.8
2.2
2.3
2.3
6.6
11.5
13.5
14.1
18.9
22.8
2.3
100.0
1999
000 ton
135
57
300
650
675
675
225
1,641
2,665
4,537
2,920
2,855
6,677
1,138
25,150
%
0.5
0.2
1.2
2.6
2.7
2.7
0.9
6.5
10.6
18.0
11.6
11.4
26.5
4.5
100.0
2000
000 ton
125
200
300
550
650
600
400
1,850
2,500
4,200
2,400
1,500
5,900
1,175
22,350
%
0.6
0.9
1.3
2.5
2.9
2.7
1.8
8.3
11.2
18.8
10.7
6.7
26.4
5.3
100.0
Implications and Recommendations
Country
1998
000 ton
%
1999
000 ton
2000
000 ton
%
Import
Sri Lanka
168
0.6
160
0.6
Peru
220
0.8
150
0.6
China
244
0.9
175
0.7
Mexico
295
1.1
340
1.4
Japan
484
1.7
700
2.8
Iran
537
1.9
1,084
4.3
Senegal
559
2.0
871
3.5
Ivory Coast
641
2.3
465
1.8
EU 2)
787
2.8
750
3.0
Saudi Arabia
906
3.2
865
3.4
Brazil
1,438
5.1
925
3.7
Philippines
2,086
7.4
915
3.6
Bangladesh
2,545
9.1
1,475
5.9
Indonesia
6,080
21.7
4,144
16.5
Others
11,035
39.4
12,131
48.2
Total import
28,025
100.0
25,150
100.0
Note: 2000 data are estimates. EU data excludes trade among EU countries.
Source: The Rice Trader, 1998-2002.
150
150
200
350
700
900
600
500
750
800
1,000
500
500
3,500
11,750
22,350
%
0.7
0.7
0.9
1.6
3.1
4.0
2.7
2.2
3.4
3.6
4.5
2.2
2.2
15.7
52.6
100.0
TABLE 6.3 EXPENDITURE AND POVERTY INCIDENCE BY HOUSEHOLD GROUP
Household group:
Rural 1
Rural 2
Rural 3
Rural 4
Rural 5
Rural 6
Rural 7
Urban 1
Urban 2
Urban 3
Indonesia
% of total
population in
this group
Mean per capita
expenditure (Rp.
000/yr.)
% of this group
in poverty
% of all poor
people in this
group
10.0
27.3
6.2
6.4
8.8
1.5
13.0
12.4
2.6
11.8
100.0
456
625
687
1011
610
1219
1268
789
916
2336
957
38.9
15.1
10.5
1.5
16.5
0.5
0.3
21.3
15.4
0.8
13.4
28.9
30.9
4.9
0.7
10.9
0.1
0.3
19.7
2.9
0.7
100.0
Memo items:
Poverty line (Rp 000 per year)
Headcount poverty rate (%)
Poverty gap (%)
Gini coefficient (%)
Source: database of WAYANG model.
369.5
13.4
1.1
39.6
51
Poverty and Social Impact Analysis: Indonesia Rice Tariff
TABLE 6.4 FACTOR OWNERSHIP OF THE BROAD HOUSEHOLD GROUPS
Shares in household income
(%):
Unskilled
labour
Skilled
labour
Mobile
agric.
capital
Mobile
nonagric.
capital
Fixed
capital
Land
Rural 1
83.7
1.9
3.5
5.1
3.3
3.6
Rural 2
30.4
5.5
6.3
11.0
39.6
5.0
Rural 3
49.7
4.9
1.4 5
8.0
27.0
17.7
Rural 4
56.7
5.8
0.9
6.9
16.4
11.9
Rural 5
40.0
7.7
1.2
8.8
20.8
8.5
Rural 6
12.2
5.6
2.9
21.6
51.1
4.2
Rural 7
38.7
34.0
1.1
9.1
24.2
5.8
Urban 1
10.4
22.2
2.0
16.3
53.3
4.2
Urban 2
17.0
15.0
2.2
18.3
47.7
6.6
Urban 3
13.2
38.3
1.3
10.8
38.2
1.9
All poor households
45.0
10.2
2.4
10.0
26.1
10.2
All households
27.8
24.6
1.6
11.0
33.4
7.0
Ratio, poor households to all
1.62
0.41
1.50
0.92
0.78
1.45
Source: database of WAYANG model.
Notes: For each household, the shares do not add to 100, because households also pay, or receive,
transfers from other households, the government and the rest of the world.
52
Implications and Recommendations
TABLE 6.5 SIMULATED MACROECONOMIC EFFECTS OF A RICE TARIFF: VARYING
RICE IMPORT SUPPLY ELASTICITY (PER CENT CHANGE)
Shock:
Increase tariff from 25 to 45%
A
B
C
D
10
2.5
5
20
0.209
0.141
0.177
0.232
-0.011
-0.006
-0.009
-0.013
Consumer Price Index
0.251
0.014
0.207
0.283
GDP Deflator
0.220
0.147
0.186
0.245
Skilled
0.088
0.069
0.079
0.095
Unskilled
0.213
0.135
0.176
0.239
Consumer price of rice (Rp.)
1.797
1.042
1.441
2.056
Producer price of paddy (Rp.)
2.305
1.333
1.847
2.316
-4.332
-8.846
-6.557
-2.585
0.527
0.304
0.422
0.602
-0.028
-0.019
-0.024
-0.032
-0.021
-0.014
-0.018
-0.024
0.190
0.183
0.187
0.193
Tariff
0.543
1.241
0.873
0.302
Nominal (local currency)
0.130
0.183
0.111
0.193
0.251
0.173
0.215
0.281
-0.002
0.014
0.008
-0.002
Simulation:
Parameter varied:
Import supply elasticity
Overall economy
Gross Domestic Product
Nominal (local currency)
Real
Wage (nominal)
Import price of rice ($US)
Paddy production
External sector (foreign currency)
Export Revenue
Import Bill
Government budget (local currency)
Revenue (local currency)
Expenditure
Real
Household sector
Consumption
Nominal (local currency)
Real
Source: Author's computations.
53
Poverty and Social Impact Analysis: Indonesia Rice Tariff
TABLE 6.6 SIMULATED DISTRIBUTIONAL EFFECTS OF A RICE TARIFF: VARYING
RICE IMPORT SUPPLY ELASTICITY
Shock:
Increase tariff from 25 to 45%
A
B
C
D
10
2.5
5
20
rural1
0.015
0.021
0.018
0.014
rural2
-0.082
-0.035
-0.06
-0.098
rural3
0.072
0.054
0.064
0.078
rural4
0.061
0.048
0.055
0.066
rural5
0.076
0.057
0.067
0.083
rural6
-0.062
-0.022
-0.043
-0.075
rural7
0.125
0.086
0.107
0.139
urban1
-0.029
-0.004
-0.017
-0.037
urban2
-0.006
0.01
0.002
-0.011
urban3
-0.065
-0.024
-0.046
-0.079
0.048
0.044
0.059
0.054
0.046
0.073
-0.003
0.060
-0.012
-0.010
-0.015
0.007
-0.028
-0.037
-0.005
-0.001
0.035
-0.007
0.020
0.016
0.034
0.015
0.007
0.035
-0.002
0.048
-0.010
0.066
0.063
0.076
0.080
0.073
0.098
-0.003
0.069
-0.014
Simulation:
Parameter varied:
Import supply elasticity
Real consumption expenditures (deflated
by household-specific CPI) % change
Rural
Urban
Changes in poverty and inequality (% change)
Indon h-count poverty %
Rural h-count poverty %
Urban h-count poverty %
Indon-poverty gap %
Rural-poverty gap %
Urban-poverty gap %
Gini Indonesia (%)
Gini rural (%)
Gini urban (%)
Levels of poverty incidence (%)
Rural
rural1
rural2
rural3
rural4
rural5
rural6
rural7
Urban
urban1
urban2
urban3
Source: Author's computations.
54
Base level
38.874
15.142
10.525
1.520
16.528
0.457
0.347
21.299
15.409
0.784
Post-simulation levels
38.861
38.855
38.858
15.187
15.161
15.175
10.494
10.501
10.497
1.514
1.515
1.515
16.484
16.495
16.489
0.459
0.458
0.458
0.344
0.345
0.344
21.311
21.301
21.306
15.411
15.405
15.408
0.786
0.785
0.786
38.862
15.196
10.491
1.514
16.480
0.459
0.343
21.315
15.412
0.787
Implications and Recommendations
TABLE 6.7 SIMULATED MACROECONOMIC EFFECTS OF A RICE TARIFF: VARYING
ELASTICITY OF SUBSTITUTION IN PADDY PRODUCTION (PER CENT CHANGE)
Shock:
Increase tariff from 25 to 45%
E
F
G
H
0.15
0.2
0.3
0.35
0.227
0.217
0.203
0.198
-0.011
-0.011
-0.012
-0.012
Consumer Price Index
0.278
0.263
0.242
0.235
GDP Deflator
0.238
0.228
0.214
0.209
Skilled
0.080
0.084
0.091
0.094
Unskilled
0.178
0.198
0.225
0.235
Consumer price of rice (Rp.)
2.164
1.955
1.674
1.575
Producer price of paddy (Rp.)
2.439
2.202
1.885
1.773
-4.207
-4.279
-4.374
-4.408
0.420
0.481
0.562
0.591
-0.024
-0.026
-0.030
-0.031
-0.016
-0.019
-0.023
-0.025
00.206
0.197
0.185
0.181
0.132
0.131
0.130
0.129
0.281
0.265
0.244
0.236
0.003
0.003
0.002
0.002
Simulation:
Parameter varied:
Elasticity of Substitution
Overall economy
Gross Domestic Product
Nominal (local currency)
Real
Wage (nominal)
Import price of rice ($US)
Paddy production
External sector (foreign currency)
Export Revenue
Import Bill
Government budget (local currency)
Revenue (local currency)
Tariff
Expenditure
Nominal (local currency)
Real
Household sector
Consumption
Nominal (local currency)
Real
Source: Author's computations.
55
Poverty and Social Impact Analysis: Indonesia Rice Tariff
TABLE 6.8 SIMULATED DISTRIBUTIONAL EFFECTS OF A RICE TARIFF: VARYING
ELASTICITY OF SUBSTITUTION IN PADDY PRODUCTION
Shock:
Increase tariff from 25 to 45%
E
F
G
H
0.15
0.2
0.3
0.35
rural1
-0.002
0.008
0.022
0.027
rural2
-0.103 -0.091 -0.075 -0.069
Simulation:
Parameter varied:
Elasticity of Substitution
Real consumption expenditures (deflated by householdspecific CPI) % change
Rural
rural3
0.091
0.080
0.066
0.060
rural4
0.082
0.070
0.054
0.048
rural5
rural6
0.076
0.076
0.077
0.077
-0.075 -0.068 -0.058 -0.054
0.158
rural7
Urban
0.140
0.114
0.105
urban1
-0.038 -0.033 -0.025 -0.022
urban2
-0.022 -0.013
urban3
-0.068 -0.066 -0.064 -0.063
0.000
0.005
Changes in poverty and inequality (% change)
Indon h-count poverty %
0.082
0.062
0.034
0.024
Rural h-count poverty %
0.083
0.060
0.029
0.018
Urban h-count poverty %
0.080
0.068
0.051
0.044
Indon-poverty gap %
0.100
0.073
0.036
0.023
Rural-poverty gap %
0.100
0.069
0.026
0.010
Urban-poverty gap %
0.098
0.084
0.061
0.053
-0.042 -0.042 -0.044 -0.044
Gini Indonesia (%)
0.139
Gini rural (%)
Levels of poverty incidence (%)
Urban
Source: Author's computations.
56
0.098
0.089
-0.069 -0.068 -0.066 -0.065
Gini urban (%)
Rural
0.122
Base level
Post-simulation levels
rural1
38.874
38.876 38.867 38.854 38.850
rural2
15.142
15.199 15.192 15.183 15.180
rural3
10.525
10.485 10.490 10.496 10.499
rural4
1.520
rural5
16.528
rural6
0.457
0.459
0.459
0.459
0.459
rural7
0.347
0.343
0.343
0.344
0.344
urban1
21.299
21.315 21.313 21.310 21.308
urban2
15.409
15.416 15.413 15.409 15.407
urban3
0.784
1.512
1.513
1.515
1.515
16.484 16.484 16.483 16.483
0.787
0.786
0.786
0.786
Implications and Recommendations
TABLE 6.9 SIMULATED MACROECONOMIC EFFECTS OF A RICE TARIFF: VARYING
ARMINGTON ELASTICITIES IN RICE DEMAND (PER CENT CHANGE)
Shock:
Increase tariff from 25 to 45%
Simulation:
I
J
K
L
Armington elasticitity in rice demand:
2
4
8
10
0.122
0.175
0.233
0.250
-0.006
-0.009
-0.013
-0.014
Consumer Price Index
0.150
0.211
0.279
0.300
GDP Deflator
0.129
0.184
0.245
0.264
0.050
0.073
0.099
0.106
0.115
0.174
0.239
0.259
Consumer price of rice (Rp.)
0.942
1.460
2.034
2.210
Producer price of paddy (Rp.)
2.305
1.061
1.645
2.291
-2.226
-3.449
-5.006
-5.540
0.278
0.429
0.595
0.645
-0.015
-0.023
-0.032
-0.035
-0.011
-0.017
-0.024
-0.026
0.189
0.190
0.191
0.191
1.613
0.963
0.247
0.027
0.076
0.109
0.145
0.157
0.151
0.213
0.282
0.303
0.001
0.002
0.003
0.003
Parameter varied:
Overall economy
Gross Domestic Product
Nominal (local currency)
Real
Wage (nominal)
Skilled
Unskilled
Import price of rice ($US)
Paddy production
External sector (foreign currency)
Export Revenue
Import Bill
Government budget (local currency)
Total revenue (local currency)
Tariff revenue
Total expenditure (local currency)
Household sector
Consumption
Nominal (local currency)
Real
Source: Author's computations.
57
Poverty and Social Impact Analysis: Indonesia Rice Tariff
TABLE 6.10 SIMULATED DISTRIBUTIONAL EFFECTS OF A RICE TARIFF: VARYING
ARMINGTON ELASTICITIES IN RICE DEMAND
Shock:
Increase tariff from 25 to 45%
Simulation:
I
J
K
L
Armington elasticitity in rice demand
2
4
8
10
rural1
0.009
0.013
0.017
0.018
rural2
-0.043
-0.066
-0.093
-0.101
rural3
0.038
0.059
0.082
0.089
rural4
0.032
0.05
0.069
0.075
rural5
0.041
0.062
0.086
0.094
rural6
-0.032
-0.050
-0.070
-0.077
rural7
0.066
0.102
0.142
0.154
urban1
-0.015
-0.023
-0.032
-0.035
urban2
0.000
-0.004
-0.007
-0.008
urban3
-0.034
-0.053
-0.073
-0.079
Indon h-count poverty %
0.023
0.037
0.054
0.058
Rural h-count poverty %
0.021
0.034
0.050
0.054
Urban h-count poverty %
0.030
0.047
0.066
0.072
Indon-poverty gap %
0.026
0.041
0.061
0.066
Rural-poverty gap %
0.021
0.035
0.052
0.057
Urban-poverty gap %
0.036
0.057
0.080
0.088
-0.002
-0.002
-0.003
-0.003
Gini rural (%)
0.031
0.049
0.068
0.074
Gini urban (%)
-0.007
-0.010
-0.014
-0.015
Parameter varied:
Real consumption expenditures (deflated by householdspecific CPI) % change
Rural
Urban
Changes in poverty and inequality (% change)
Gini Indonesia (%)
Levels of poverty incidence (%)
Rural
Urban
Source: Author's computations.
58
Base level
Post-simulation levels
rural1
38.874
38.866
38.863
38.859
38.858
rural2
15.142
15.165
15.178
15.193
15.198
rural3
10.525
10.508
10.499
10.489
10.486
rural4
1.520
1.517
1.515
1.514
1.513
rural5
16.528
16.504
16.492
16.478
16.473
rural6
0.457
0.458
0.458
0.459
0.459
rural7
0.347
0.345
0.344
0.343
0.343
urban1
21.299
21.305
21.309
21.313
21.314
urban2
15.409
15.409
15.410
15.411
15.411
urban3
0.784
0.785
0.786
0.787
0.787