Class, power, and agrarian change land and labour in rural west java

STUDIES ON THE ECONOMIES OF EAST AND SOUTH-EAST ASIA
General Editors: Peter Nolan, Lecturer in the Faculty of Economics and
Politics, University of Cambridge, and Fellow and Director of Studies in
Economics, Jesus College, Cambridge, England; and Malcolm Falkus,
Professor of Economic History, University of New England, Armidale,
New South Wales, Australia
In the last decades of the twentieth century the small and medium-sized
nations of East and South-East Asia have begun a process of potentially
enormous political and economic transformation. Explosive growth has
occurred already in many parts of the region, and the more slowly growing countries are attempting to emulate this vanguard group. The impact
of the region upon the world economy has increased rapidly and is likely
to continue to do so in the future.
In order to understand better economic developments within this vast and
diverse region, this series aims to publish books on both contemporary
and historical issues. It includes works both by Western scholars and by
economists from countries within the region.
Published titles include:
Melanie Beresford
NATIONAL UNIFICATION AND ECONOMIC DEVELOPMENT IN
VIETNAM
John Butcher and Howard Dick (editors)
THE RISE AND FALL OF REVENUE FARMING
Mark Cleary and Shuang Yann Wong
OIL, DEVELOPMENT AND DIVERSIFICATION IN
BRUNEI DARUSSALAM
Yujiro Hayami and Toshihiko Kawagoe
THE AGRARIAN ORIGINS OF COMMERCE AND INDUSTRY
Jomo K. S.
GROWTH AND STRUCTURAL CHANGE IN THE MALAYSIAN
ECONOMY

Medhi Krongkaew (editor)
THAILAND'S INDUSTRIALIZATION AND ITS CONSEQUENCES
Lee Sheng-Yi
MONEY AND FINANCE IN THE ECONOMIC DEVELOPMENT OF
TAIWAN
Rajah Rasiah
FOREIGN CAPITAL AND INDUSTRIALIZATION IN MALAYSIA


Class Power and
Agrarian Change
Land and Labour in Rural West Java
Jonathan Pincus
Management Information Expert
Food and Agriculture Organization

&


as

First published in Great Britain 1996 by

MACMILLAN PRESS LTD

Houndmills, Basingstoke, Hampshire RG21 6XS
and London
Companies and representatives

throughout the world
A catalogue record for this book is available
from the British Library.
ISBN 0-333-64578-2

m

First published in the United States of America 1996 by

ST. MARTIN'S PRESS, INC.,
Scholarly and Reference Division,
175 Fifth Avenue,
New York, N.Y. 10010
ISBN 0-312-15827-0
Library of Congress Cataloging-in-Publication Data
Pincus, Jonathan.
Class power and agrarian change : land and labour in rural West
Java / Jonathan Pincus.
p. cm. — (Studies in the economies of East and South-East
Asia)
Includes bibliographical references and index.
ISBN 0-312-15827-0 (cloth)
I. Land reform—Indonesia—Java. 2. Social classes—Indonesia-Java. 3. Power (Social sciences)—Indonesia—Java. I. Title.
II. Series.
HD1333.I52J386 1996
305.5'23,095982—dc20
96-39114
CIP

© Jonathan Pincus 1996

All rights reserved. No reproduction, copy or transmission of
this publication may be made without written permission.
No paragraph of this publication may be reproduced, copied or
transmitted save with written permission or in accordance with
the provisions of the Copyright, Designs and Patents Act 1988,
or under the terms of any licence permitting limited copying
issued by the Copyright Licensing Agency, 90 Tottenham Court
Road, London W1P9HE.
Any person who does any unauthorised act in relation to this
publication may be liable to criminal prosecution and civil
claims for damages.
10 9 8
05 04 03

7 6 5 4
02 01 00 99

3
2 1
98 97 96

Printed in Great Britain by
The Ipswich Book Company Ltd, Ipswich. Suffolk


Contents
List of Tables
List of Figures
Preface
1

2
3
4
5
6

vii
xi
xii

Introduction
Methodological Issues
Measuring Class Differentiation
Wage Labour Relations in Agriculture
The Process of Accumulation
Conclusion

1
19
37
93
147
188

Appendices
A The Subang Rural Price Index
B Real Wage Trends in Agriculture
C Supplementary Tables

194

198
204

Glossary of Terms and Acronyms
Bibliography
Index

222
224
245

v



List of Tables
1.1
1.2
1.3
1.4
1.5
2.1
2.2
2.3
3.1
3.2
3.3
3.4
3.5
3.6

3.7
3.8
3.9
3.10
3.11
3.12
3.13
4.1
4.2

Producer subsidies in rice in selected Asian cou ntries
6
Distribution of operated farm sizes, Java 1963-1983
10
Average annual growth of wage labour force and 11
population of self employed farmers, Java 1980-1990
Sectoral distribution of increase in wage employment, 12
Java 1980-1990
Real daily wages for male hoers in some rice- 13
cultivating regions of Asia, 1987
Population density and agrarian density in North and 23
South Subang subdistricts, 1971 -1990
Components of the possessions score
28
Possessions score summary statistics
30
Female-headed households
43
Spearman correlation coefficients for possessions 49
scores and control over irrigated rice fields

Mean possession scores by area controlled
50
Correlations of labour hiring, family labour, posses- 53
sions scores and access to land
Correlations of labour hiring, family labour, posses- 54
sions scores and access to land, small and middle
farmers
Spearman correlation coefficients for agricultural 56
labour days hired out with possessions scores and
access to land, farming households
LOGIT analysis of participation in agricultural wage 57
labour among farming households
Mean values of selected indicators for landless 61
agricultural labour households
Income from nonagricultural wage labour by 69
land-controlling category
Migrants to domestic work in the Middle East
70
Mean earnings from trade and industry by 72
land-controlling groups
Distribution of households by class
79
Post-hoc comparisons of group means for small and 80
middle farmers, South and East Subang
Location of pre-harvest wage employment, agricul- 102
tural workers from North Subang, dry season 1990
Labour days worked locally by North Subang farm 103
labourers by wage system, dry season 1990
vn

viii
4.3
4.4

List of Tables

Location of harvest employment, North Subang
105
Labour days worked locally by East Subang farm 109
labourers by wage system, dry season 1990
4.5
Harvest shares under bawon and ceblokan systems 111
by class, East Subang, dry season 1990
4.6
Crosstabulation of ceblokan arrangements by class 113
category of farmer and worker, East Subang
4.7
Correlation matrix of ceblokan earnings from large 114
farmers, East Subang, dry season 1990
4.8
Location of harvest employment, East Subang
115
4.9
Location of pre-harvest wage employment, East 116
Subang
4.10 Location of pre-harvest and harvest labour days, 118
South Subang
4.11 Labour days worked locally by South Subang farm 120
labourers by wage system, dry season 1990

4.12 Earnings per hour for one, two and three task 122
ceblokan arrangements, South Subang (combined
harvest and pre-harvest wages)
4A3 Correlation matrix of ceblokan earnings from large 123
farmers, South Subang, dry season 1990
4.14 Real wages for daily wage and contract labour tasks, 129
dry seasons 1971, 1978/79 and 1990, (Rupiah per
hour, constant 1971 prices)
4.15 Labour and Capital's share of rice production per 132
hectare for large farmers, constant 1971 prices
5.1
Mean output per hectare and capital costs in rice 150
production by class category, dry season 1990
5.2
Distribution of bank borrowing by class category (Rp. 152
'000)
5.3
Mean output, returns and interest payments for 155
sharecroppers by class category, dry season 1990
5.4
Mean income per hectare from own-cultivation and 159
leasing by class category, dry season 1990 (Rp. '000
per hectare)
5.5
Purchases and sales of rice fields by class category, 163
1985-1990 (hectares)
5.6
Size distribution of rice field ownership, South 164
Subang, 1969,1979 and 1990
5.7

Mortgage of rice fields by class category, dry season 166
1990
5.8
Trade and nonagricultural activities
168
5.9
Ownership of mechanical tractors and rice hullers, 169


List of Tables

ix

Subang District, 1970-1989
5.10 School enrolment rates for children 7 to 18 years of 171
age, West Java, Java, Indonesia and study villages
1989
5.11 Participation in salaried employment by class 172
category
5.12 Status of Village Co-operative Units and Farmer 174
Credit (Kredit Usaha TanU Subang District 19891990
A. 1
A.2
B.l
B.2
C.l
C.2
C.3
C.4
C.5

C.6
C.7
C.8
C.9
C. 10
C. 11
C.l2
C.l3

Weights used for Subang 9 commodity price index
195
Comparison of the Subang rural price index with 196
Subang rice prices and the Java-wide rural price
index
Percentage Increase in Real Hoeing Wages From 199
Village Studies, 1970 to 1987
Selected indicators of changes in labour demand and 203
supply, Java 1980-1990 (average growth rates per
annum)
Size distribution of paddy field ownership, area 204
operated and area controlled, dry season 1990, North
Subang
Size distribution of paddy field ownership, area 205
operated and area controlled, dry season 1990, South
Subang
Size distribution of paddy field ownership, area 206
operated and controlled, dry season 1990, East
Subang
Ownership of dry land fields and fishponds, South 207
Subang

Agrarian density, gini ratios for sawah ownership 208
and per cent landless in 17 Javanese villages
Post-hoc comparisons (Scheffe test) of possessions 209
score means for land-controlling groups
Categorical groups for logit analysis
210
Participation in nonagricultural wage labour
211
Mean wages and labour days worked, nonagricul- 212
tural and agricultural wage labour, dry season 19901
Trade and industry activities
213
Work participation rates by class
214
Pre-harvest labour use by task, North Subang 1979 215
and 1990 dry seasons (person-hours per hectare)
Pre-harvest labour use by task, East Subang 1990 216


x

List of Tables

dry season (hours per hectare)
C14 Pre-harvest labour use by task, South Subang 1978
and 1990 dry seasons (hours per hectare)
C.l5 Crosstabulation of ceblokan arrangements by class
category of farmer and worker, South Subang
C.16 Educational attainment of rural population 10 years
of age and older, West Java, Java, Indonesia and

study villages
C.17 Educational attainment levels for residents ten years
of age and older

217
218
219
220


List of Figures
3.1
3.2
3.3
5.1

Agrarian Density and Land Ownership
46
Agrarian Density and Landlessness
47
Income Sources For Landless Agricultural Labourers 63
Flow of consumption loans within and between 178
classes following partial harvest failure, East
Subang

B.l
B.2

Real Hoeing Wages in Java, 1976-1990
Real Tranplanting Wages in Java, 1976-1990

xi

200
201


Preface
This book is a revised version of my dissertation. I would
like to thank the editors of the series, Peter Nolan and Malcolm
Falkus, for the chance to present my results and views to a
wider audience.
The research on which this book is based was funded by
the Board of Graduate Studies of Cambridge University, the
Cambridge Political Economy Society and Wolfson College.
Additional financial support for fieldwork in Indonesia was
provided by the US Agency for International Development and
the Ellen McArthur Fund. I am grateful to Michael Hammig,
then with USAID, Jakarta, for his constant encouragement and
support. Thanks are also due to the Pusat Penelitian Sosial
Ekonomi Pertanian (PSE) for serving as my official host in
Indonesia, and to Mohamad Saat of the Central Bureau of
Statistics in Jakarta for his patience with my many queries.
This book could not have been written without the practical
assistance and insightful comments of John Sender. I have also
learned much from comments given by Terry Byres, Peter
Kenmore, Mushtaq Khan, David McKendrick, Peter Nolan,
Gabriel Palma and Rizal Ramli at various stages of the book's
preparation. I alone, however, am responsible for all remairing
errors of fact and interpretation.

xn


CHAPTER ONE

Introduction
THE AGRARIAN MYTH

THIS BOOK is about the impact of locally specific patterns
of class formation and the exercise of class power on agrarian
dynamics at the village level.
Based on a comparison of three villages in West Java, the
book focuses on three mutually reinforcing processes: social and
economic differentiation, the evolution of production relations
and the accumulation of capital. It is argued that the pace and
form of these changes are shaped in important ways by the
power of agrarian social classes to impose or resist economic
changes consistent with their interests.
The proposition that local patterns of agrarian change are
closely bound up with class structure and class power runs
counter to two common assumptions about Asian villages in
general, and Javanese villages in particular. The first relates to
the relevance of social class to the analysis of agricultural
development. Even in the face of profound economic and social
change, policy makers, scholars and development workers
continue to cling to an outmoded view of the countryside as
populated by homogeneous, small-scale producers farming on
the basis of family labour and supported by 'traditional' village
institutions.1 The historian Richard Hofstadter, noting a similar

bias in American political discourse, used the term 'agrarian
myth' to describe the resilience of these populist images despite
their increasing irrelevance to rural economic structure. Sadly,
his observation that 'the agrarian myth came to be believed more
widely and tenaciously as it became more fictional' applies
equally well to Java (and to many other regions of Asia) as it does
to the US (1956, 30).
Failure to consider the issues of class and class power has
also contributed to mainstream economists' preoccupation with
demographic factors as the sole or main instigator of economic
change at the village level. An important example of this ten1


2

Class Power and Agrarian Change

dency is the neoclassical theory of 'induced innovation'. This
approach attributes changes in production relations and income
distribution to movements of relative prices resulting from
population growth (Hayami and Ruttan 1985; Ruttan 1978). In
their influential book Asian Village Economy at the Crossroads
(1981), Hayami and Kikuchi found support for induced
innovation in their study of two villages in the Subang district of
West Java. Both of these villages have been resurveyed for this
book. The evidence collected ten years after Hayami and Kikuchi
completed their survey indicates that the induced innovation
hypothesis represents a vast oversimplification of agrarian
dynamics in the study villages. More specifically, the authors'
focus on population pressure on land resources can only be sustained if it is assumed that the Subang villages are essentially

closed systems, in which both labour and capital are immobile.
These two assumptions about agrarian change have proven
particularly durable in the case of Java. This is partly explained
by the island's extremely high population density and relatively
small average farm sizes.2 Social scientists have also played a
part in perpetuating the agrarian myth in Java. In his celebrated
theory of 'agricultural involution', Clifford Geertz dismissed the
relevance of class formation in the Javanese context in favour of
his notion of 'shared poverty', in which villagers adopt increasingly elaborate tenancy and work-spreading arrangements in
response to mounting population pressure.3 In answering his
numerous critics (see White 1983; Alexander and Alexander
1982; Kano 1980), Geertz later claimed that the concept of
shared poverty has been broadly misinterpreted, and that it was
meant solely 'as a sensitising, heuristic concept designed to
elucidate situations to which it could be applied to the degree
that it could, and, contrastively, those to which is could not'
(1984, 527). Whatever his original pedagogic intent, the notion of
shared poverty has gained wide currency among Indonesian
academics and policy makers, and continues to represent a
substantial obstacle to the acceptance of a more realistic
interpretation of agrarian change.
These views still find support in the work of contemporary
economists of a more technocratic bent. Tabor, for example,
concludes:
Although access to land is far from completely equal, landholdings remain far more evenly distributed in Indonesia
than in many other developing countries...According to the
1983 Agricultural Census, only 4.4 per cent of Javanese and


Introduction

3

1.2 per cent of non-Javanese agricultural households are
landless. This, coupled with the relatively low rural to urban
migration rates [sic] and the rise in agricultural employment
levels, provides empirical support for the argument that
green revolution advances have increased labour demand and
improved returns in agriculture without marginalising the
small farmers (1992, 167).
Yet, as Breman notes, 'The way in which such statistical material
is interpreted is strongly dependent not only on the reliability of
methods for data collection but also on its style of presentation'
(1983,127). Nowhere in the above quotation, nor in the following
text, is there mention of the fact that the term 'agricultural
households' as used in the census refers only to households in
which at least one individual is directly engaged in crop
production, fisheries or raising livestock. This definition excludes
an estimated 40 per cent of rural Javanese households which
neither own nor operate land (see White 1991, 57).4
The political ideology of the Suharto regime has also worked
to sustain the agrarian myth in Indonesia. A central pillar of New
Order agrarian policy has been the rigid control of political
activity in rural areas. In the years following the violent
destruction of the Indonesian Communist Party in 1965-66, the
government took immediate steps to remove any further threat
from class-based movements in the countryside. Under the socalled 'floating mass' policy all political activity is banned below
the district level except for the brief (and closely monitored)
campaigns prior to general elections. Independent organisations
of farmers and farm workers are forbidden, and the regime has

employed an effective strategy of bureaucratisation and
militarisation of local government to restrain all forms of
political initiative and resistance in the countryside (Husken and
White 1989, 249-251). Into the ideological vacuum created by
these policies, the government has injected a strong rhetorical
emphasis on populist themes such as 'mutual self-help' (gotong
royong) and the role of state-sponsored village co-operatives. In
this environment it is hardly surprising that discussion of class
formation and rural class conflict is met with disbelief and in
some cases hostility among Indonesians both within and outside
of government circles.
Much of what is contained in the following pages represents
a challenge to the set of beliefs which constitute the Javanese
agrarian myth. Some readers, including some who consider
themselves to be reasonably familiar with the patterns of Java-


4

Class Power and Agrarian Change

nese village life, will not recognise the image of the island
presented here.5 However, it should be stressed from the outset
that this book is not an attempt to introduce an alternative
uniform view of rural Java. It is obvious that substantial variation
exists between villages in terms of rural social structure, class
relationships and dynamics of economic change. Instead, the aim
here is to demonstrate the centrality of class relations for a fuller
understanding of these differences, and the locally-specific
factors which give rise to them.

Organisation of the Book
The book is organised as follows. The remainder of this
chapter provides a brief overview of the macro level evidence
concerning trends in Javanese agrarian change. Chapter 2 describes the methodology used in the field studies, and addresses
some common problems encountered by researchers studying
agrarian change at the village level. Two methodological issues of
immediate relevance are discussed in detail. These are the use of
the 'village' as a unit of analysis, and problems associated with
the use of resurvey evidence in studying processes of change.
The chapter also introduces the 'possessions score', a measure
of household well-being based on the ownership of durable
goods.
Chapter 3 presents an analysis of class structures in the
three study villages. It is argued that landholding is at best a
partial measure of class differentiation, since the group of small
farm households—which is often presented in the literature as
an economically homogeneous group-4n fact encompasses both
extremely poor and well-off villagers. This group is more
accurately differentiated in terms of the use of hired labour as
opposed to reliance on family labour for cultivation, and the
participation of household members in the labour market as
sellers of wage labour.
Wage relations in agriculture are examined in Chapter 4. The
proposition that production relations adjust to restore equilibrium to the labour market is rejected in favour of a less
deterministic framework focusing on class structure and the
relative bargaining power of classes at the village level.
Chapter 5 examines strategies of capital accumulation employed in the three villages. These encompass own cultivation,
land leasing, land acquisition, non-agricultural activities, salaried
employment and 'political accumulation' pursued through

Introduction

5

positions in village government and co-operatives. The relatiaiship between agrarian crisis and accumulation is analysed
through an account of a partial harvest failure in one of the
study villages in the 1989-1990 wet season.
THE MACRO SETTING: EVIDENCE OF AGRARIAN CHANGE IN
JAVA

Studies conducted in a range of Asian regions have emphasised several common themes in the development of postcolonial agriculture.6 These include the heightened role of the
state in the agrarian economy, the commercialisation of peasant
agriculture, accelerated processes of capital accumulation and
proletarianisation and changes in relations of production, most
notably the casualisation of the wage labour force and the
rationalisation of labour use. All of these trends are discernible
in the Javanese context, although, as noted above, the quality
and coverage of the macro-level evidence is uneven and hence
often subject to misinterpretation.
State Penetration
One area in which there is agreement among observers of
Indonesian agriculture is the heightened involvement of the state
in the agrarian economy and in village social and economic life
since the rise of the New Order regime. This is particularly true
of Java's fertile, lowland rice-growing regions. Mindful of the
importance of sufficient supplies of rice for economic and
pohtical stability, the government launched a series of initiatives
in the late 1960s to increase rice production. Most prominent
among these was the rehabilitation and expansion of irrigation

systems in Java. Between 1969 and 1980, extension or rehabilitation of irrigation works covered more than half of all paddy
field area in Java, and about 20 per cent of area outside Java
(Booth 1988, 144). Related investments included the development of the domestic fertiliser industry, seed production
facilities and a national network for transporting and storing rice
and production inputs (Fox 1991, 61-63).
These infrastructural improvements facilitated the rapid
adoption of new technologies based on modern rice varieties and
more intensive use of inorganic fertilisers. Another policy
instrumental to the increase in fertiliser use was government
intervention in markets for both fertilisers and rice. From 1970
to 1989, domestic prices for urea and phosphatic fertilisers


Class Power and Agrarian Change

6

averaged approximately 40 to 50 per cent of world prices, a level
of subsidisation which effectively shifted part of the burden of
risk associated with technological learning from farmers to the
government (Tabor 1992, 180; Timmer 1985, 70). In the rice
market, the State Logistics Board (Bulog) was assigned the task of
defending official 'floor' prices for rice purchased from farmers,
prices which during the 1980s were consistently above world
market levels (World Bank 1992, 90).7 The combination of
fertiliser subsidies and state intervention in the rice market
resulted in a modest level of protection for farmers, amounting
to a strategy of import substitution in rice with the aim of
generating employment in rural areas and saving scarce foreign
exchange (Table l.l).8

Table 1.1: Producer subsidies in rice in selected Asian
countries, average 1982-1987.
Country
Japan
South Korea
Taiwan
Indonesia
India
Bangladesh
Thailand
Pakistan
China

Producers' Subsidy
Equivalent1
89.33%
74.00%
31.33%
10.50%
-1.17%
-2.33%
-3.83%
-56.70%
-62.67%

Share of world
production
3.61%
1.97%
0.73%

9.09%
20.50%
5.22%
4.46%
1.15%
45.85%

1

The Producers' Subsidy Equivalent (PSE) is the value of price supports plus
indirect subsidies as a share of the domestic price of the commodity per unit of
output.
Source: Gulati and Sharma 1992, Table 4b.

However, not all components of Indonesia's rice intensification programme have been equally successful. The fixed input
packages provided as part of the BIMAS ('Mass Guidance') rice
intensification programme were often inappropriate to local
agro-ecological conditions, and in some areas farmers were
coerced by local officials into taking loans and inputs they did
not want. In response to mounting debt arrears, and the withdrawal of smaller farmers from the programme, BIMAS was
discontinued in 1984 (Robinson and Snodgrass 1987). Until they


Introduction

7

were abolished in 1988, subsidies on pesticides encouraged
inchsaiminate use of broad spectrum insecticides which induced
the highly disruptive brown planthopper outbreaks of the 1970s

and mid-1980s (Kenmore et al., 1994; Fox 1991, 75). Finally,
state-sponsored village co-operatives have not lived up to
expectations in terms of the provision of farm inputs or as a
marketing channel for rice output, and their administration has
been prone to manipulation by village elites (Booth 1988, 256).
Although rice intensification has received the most attention
in the literature, other aspects of state penetration into the
agrarian economy have also played a large role in shaping
patterns of change. Not least among these has been the state's
tight control over village pohtical life. Selection of village heads is
screened by district officials, and previously elected village
councils have been replaced by 'village community security
bodies' (LKMD) appointed by the village head. The centralisation
of local power in the hands of village, sub-district and district
heads, in tandem with the suppression of autonomous pohtical
activity, has deprived the rural labouring classes of protection
from illegal levies and other abuses of power on the part of
village and supra-village elites. The absence of social control on
the behaviour of village elites has naturally widened the scope
for various forms of pohtical accumulation. The expansion of
government employment, particularly during the 1970s and
early 1980s, also fit neatly into the diversified accumulation
strategies pursued by village elites. Meanwhile, failure to
implement the 1960 land reform laws removed the last legal
threat to the power of large landowners, who have benefited
hugely from the economic and pohtical patronage of the New
Order state (Gunawan Wiradi 1984; Rajagukguk 1988, 64; Hart
1986a, 40ff).
Accumulation and Proletarianisation
Owing to the accelerated pace of change associated with the

rice intensification effort, it is perhaps understandable that
observers of the agrarian scene in the 1970s tended to interpret
the increasingly commercial orientation of Javanese rice farmers,
particularly the rationalisation of labour relations, as a direct
effect of the adoption of new production technologies (see, for
example, Collier et al. 1978). Subsequent work, however, has
placed greater emphasis on the historical roots of contemporary
developments. (Booth 1988; Hiisken and White 1989, 237).9
Although this renewed sensitivity to historical factors is to be


8

Class Power and Agrarian Change

welcomed, there is still little agreement as to the content and
direction of change, particularly with reference to the structure
of agrarian society.
Debate in the literature has centred around conflicting interpretations of empirical evidence pertaining to the structure of
landholdings, the degree of landlessness and growth of the wage
labour force. Booth, for example, argues that:
A process of levelling down has been taking place in Java
until the 1970s, whereby a growing population and agricultural labour force have been absorbed on limited land
through a gradual reduction in the average holding size, with
the dispersion of holdings remaining roughly the same (1988,
55).
This conclusion, however, is based on a comparison of data from
the colonial administration's 1903 welfare survey on land
ownership with data from the 1973 agricultural census on area
operated, and is therefore not justified (see Hasselman 1914,

app.R). Moreover, this view is inconsistent with evidence presented in large-scale colonial surveys as well as historical studies
of changes in specific regions of Java. Although conditions varied
markedly from region to region, these sources point to several
salient features of Javanese agrarian structure during the late
colonial period. According to the aforementioned colonial
welfare survey, at the turn of the century one-third of the
agrarian population owned no land, while at the other extreme
the nine per cent of landowners holding more than 1.4 hectares
accounted for more than one third of all land (Hiisken and White
1989, 240).
This already pronounced degree of inequality increased over
the ensuing years, particularly during the boom years following
the First World War (Ibid., 242ff.; Elson 1984, 218ff.; Breman
1983, 74ff;). Rapid commercialisation and differentiation during
the late 19th and early 20rh centuries was also accompanied by the
decline of communal land tenure and its gradual replacement
with individual ownership rights (Elson 1984, 219; Burger 1984,
90-91; Kroef 1984,155).
These trends were stalled with the onset of the Great Depression of the 1930s, and did not regain momentum until after
the Japanese occupation and the ensuing struggle for independence (White 1989b, 69; Elson 1984, 233ff). Thus, rather than a
gradual process of 'involution', the first half of this century
appears to have been characterised by a period of nascent
capitahst development, followed by stagnation and reversal


Introduction

9

during the difficult years of 1930s and 1940s. Despite the

disruptions of depression and war, however, pervasive inequalities in access to land as well as high levels of landlesaiess
persisted, forming the social basis of the agrarian conflicts of the
early 1960s (Lyon 1970, 20ff.; Ten Dam 1966, 349; Mortimer
1972; 34ff).
Attempts to trace changes in landholding patterns since the
1960s have focused on the three decennial agricultural censuses,
the first of which was carried out in 1963. The censuses suggest
a pattern of holdings dominated by small farms, a structure
which has remained remarkably stable over the 20 year period
covered by the surveys (Table 1.2). Yet as Hiisken and White
note, 'these data in fact serve mainly to highlight the limitations
of farm-size statistics in telling us anything about agrarian
change, especially in contexts of population and productivity
growth' (1989, 255). To begin with, as noted above, the agricultural censuses provide no information on land ownership or
landlessness, and thus are of limited value in addressing
questions of agrarian structure.10 Second, the reliability of the
census data is open to question, as total land area recorded in
these surveys differs widely from other sources, such as land tax
data (Booth and Sundrum 1976, 102). Moreover, land leased in
greatly exceeds land leased out as recorded in the 1983 census,
an indication that landlords systematically underreport farm
sizes. Village resurveys, although still few in number, also point
to a process of land concentration which remains undetected at
the macro-level. For example, surveys carried out in nine lowland
rice-producing villages in 1971 and 1981 reveal a dramatic
increase in the share of rice fields controlled by farmers
cultivating more than one hectare, accompanied by a sharp rise
in landlessness (White and Gunawan Wiradi 1989, Table 13.4).n
Most importantly, farm size statistics tell us nothing about
the degree to which and processes by which small farmers and

landless households have become increasingly dependent on the
sale of wage labour for their subsistence. As Karl Kautsky
remarked nearly a century ago,
We now see that both tendencies toward centralisation and
towards fragmentation can work alongside each other. The
number of small farms whose owners appear on the market
for commodities as proletarians and sellers of labour-power is
increasing. But their landholdings are only relevant outside
the sphere of commodity-production, in the sphere of production for the household...Once this stage is reached, any in-


10

Class Power and Agrarian Change
crease in small rural enterprises simply becomes one particular form in which the number of proletarian households increases — a process which runs hand in hand with the multiplication of large-scale capitalist enterprises (1988, 179, emphasis in original).12

This aspect of proletarianisation is particularly relevant in Java
where farms of less than one-half hectare predominate numerically (although not in terms of land area) and where, as we have
seen, landless households have historically accounted for a large
share of the total rural population. From this perspective, the
process of proletarianisation in Java is not restricted to the
separation of small farmers from their land, but also encompasses the processes through which petty producers within and
outside of agriculture are transformed into sellers of wage
labour.
Table 1.2: Distribution of operated farm sizes, Java
1963-1983.1
2

1963
1973

1983
Farm size Total farms %area Total farms % area Total farms % area
-0.01-0.04
967,706
0
-393,619
0
923,523
1
0.05-0.09
3,712,012
25
4,571,702
22 5,392,117
21
0.10-0.49
1,319,396
27 2,146,053
27 2,486,168
26
0.50-0.99
605,564
23 1,124,834
27
27 1,280,929
1.00-1.99
231,509
25
407,724
24

495,555
25
2.00 +
5,868,490
100 8,643,932
100 11,546,013
100
Total
1
The 1963 census did not collect information on farms of less than 0.10 ha. In
the 1973 census, the minimum sizes were 0.05 ha. for wet rice fields and 0.10 ha.
for dryland fields.
2
Rice farms only.
Sources: 1963 Agricultural Census, Final Report; CBS 1975; 1985.

Table 1.3 compares the average annual growth in the wage
labour force with growth of the total population and selfemployed farmers for the period 1980 to 1990. Unfortunately,
because of a change in the definition of 'employee' beginning
with the 1980 population census, labour force data collected
before this date are not comparable with post-1980 data.13 Data
from the 1971 population census are also incompatible as a
more stringent definition of employment was apphed in that


Introduction

11

year.14 As shown in the table, the census figures indicate that

average annual growth of the wage labour force in Java (6.25 per
cent) far exceeded the rate of growth of the working age
population (2.43 per cent) during the 1980s. Meanwhile, the
number of self-employed farmers was nearly constant, growing
at only 0.65 per cent per annum. Even these figures do not tell
the full story, since they only cover the employment status of the
individual's primary occupation during the reference period15 as
reported by the respondent. They therefore prcvide no indication
of the extent to which small-scale producers, particularly small
farmers, are dependent on the sale of wage labour for their
subsistence. This latter issue is considered in greater depth in
the context of the three Subang villages in Chapter 3.
Table 1.3: Average annual growth of total wage labour force
and population of self-employed farmers, Java 1980-1990.
West
Java

Central
Java

Yogyakarta

East
Java

Total

Men
Women
Total

Wage labour force, ten years of age and older
9.64%
4.09%
5.11%
4.20%
5.62%
12.86%
5.67%
8.72%
4.64%
10.42%
4.60%
6.22%

5.79%
7.36%
6.25%

Men
Women
Total

Self-employed farmers
0.85%
1.32%
-0.16%
-1.01%
0.04%
-1.16%

0.47%
1.08%
-0.43%

0.78%
-0.41%
0.54%

0.94%
-0.49%
0.65%

Men
Women
Total

Total population, ten year of age and older
1.84%
3.93%
1.96%
1.38%
1.74%
3.68%
1.83%
1.26%
3.80%
1.89%
1.32%
1.79%

2.52%
2.35%
2.43%

Source: CBS 1983, CBS 1992a.

As shown in Table 1.4, an important feature of the growth of
wage employment during the 1980s was the large role played by
the manufacturing sector, which accounted for 40 per cent of the
total increment in wage employment, and nearly half of the
increment among women. As a share of total employment in
Java, manufacturing increased from 11 to 15 per cent during the


12

Class Power and Agrarian Change

same period. Construction has also been a particularly dynamic
sector, accounting for 18 per cent of the increment in male wage
employment. At the same time, however, the capacity of
agriculture to absorb new entrants into the labour force has
declined, as opportunities for new investments in irrigation
diminish and as mechanisation of rice agriculture begins to
gather pace (World Bank 1992, 32; Naylor 1992). Rapid employment growth in sectors outside of agriculture is closely related to
the growing importance of migration, particularly circular
migration, as noted in village studies (Hugo 1985; Manning
1988b, 61).16
Table 1.4: Sectoral distribution ofincrease in wage employment, Java 198(hl990.1
Men Women

Total
Agriculture
Nonagriculture
Manufacturing
Mining and quarrying
Construction
Trade
Transport and communications
Other2
Total nonagriculture
Total

5.3%

2.7%

4.4%

34.3%
5.4%
17.8%
8.7%
8.5%
20.0%
94.7%
100.0%

47.4%
1.4%
0.5%

9.6%
0.4%
37.9%
97.3%
100.0%

38.9%
4.0%
11.8%
9.0%
5.7%
26.2%
95.6%
100.0%

1
Percentage of total increment in wage employment accounted for by each of the
sectors listed.
"' Includes utilities, financial and business services, real estate government
employment and other services.
Source: CBS 1983, CBS 1992a.

Despite the rapid growth of wage employment, wages in
agriculture remain exceeding low by international standards.
This is demonstrated in Table 1.5, which presents figures for real
wages (as measured in rice equivalents), rice yields per hectare
and rural population density for a range of Asian rice-producing
countries and regions for the year 1987. As shown in the table,
daily wages for male hoers in Java ranked among the lowest in
the sample despite the fact that Java recorded the highest average paddy yields.17