AN EI’ALUATION 0F SOHO-ECONOMIC DEVELOPMENT AT THE VILLAGE LEVEL IN BANDUNG REGENCY, INDGNESIA: A METHODOLOGICAL EXPLORATION Bissatation for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY ARIE LASI'ARIO KUSUMADEWA 197.5 ,A ' L I BRARY Miclfigan Sm: University This is to certify that the thesis entitled An Evaluation of Socio-Economic Development at the Village Level in Bandung Regency, Indonesia: A Methodological Exploration presented by Arie La stario Kusumadewa has been accepted towards fulfillment of the requirements for Doctor of Wegree in Resourcefievelopment Major professor Date July 9L 1975 0—7 639 g amBmG av ‘3' HUM; & SUNS' 300K BIII‘IIERY INC. LIBRAR' SINGERS our ' ' ' 'ulbl- MWMRQ 3m 5 2&2 ABSTRACT AN EVALUATION OF SOCIO-ECONOMIC DEVELOPMENT AT THE VILLAGE LEVEL IN BANDUNG REGENCY, INDONESIA: A METHODOLOGICAL EXPLORATION BY Arie Lastario Kusumadewa At the present Indonesia is trying to increase the gross national product to enhance the quality of life of the people. Since most of the people live in the rural region and the dominant sector is still agri- culture, rural community resource development strategies and programs are becoming important. However, there has not been much quantitative research-work done at the village level that deals with variables which influence income. Hence, one of the reasons for this research is to fulfill the need for research results in this field. In addition, it is hoped that this study will contribute many possible ways by which to evaluate plans and programs and also to formulate development strategies. In order to make this contribution, factor analysis was applied to identify the influencial Arie Lastario Kusumadewa variables and to screen out the trivial variables in different levels of data sources-~household level, village level and district level. Conceptual variables were condensed from the familial variables, community variables and environmental variables. The combined variables that were analyzed multi-level-wise were then used to construct hypothetical statements that explained the condition of a region. Such a multi-level approach in data analysis is a way to conserve the holistic nature of variables in an existing socio- economic system. The application of multi-level analysis was demonstrated in the regency of Bandung, West Java, Indonesia. By including variables from different levels for analysis, a "vertical" and "horizontal" profile of the region can be examined. When oblique rotation technique was used in the factor analysis, a higher order factorization for Bandung data aggre- gated the hypothetical statements from ten to two groups of statements. For the Bandung Regency the prominent influential hypothetical statement with regards to family income is: demographic composition will influence intensive agri- culture. In other words, this study indicated that in those villages where there was an appropriate balance of children to total p0pulation and balance in kinds of Arie Lastario Kusumadewa professions within the population, the agricultural pro- duction intensity will be greater and therefore family income will be higher. Policy implications that stem from these analyses suggest that these variables should be considered when new rural development program formulated. For instance this study indicates that high priority should be given to road improvement, to stimulate the development of cooperative organizations, and primary and vocational education. Fortunately these variables indicate kinds of programs that can be carried out simultaneously because they can be implemented by different levels of government. However, before embarking on any pro- jects, a careful benefit cost analysis should be made of all alternatives strategies emphasizing these variables. AN EVALUATION OF SOCIO-ECONOMIC DEVELOPMENT AT THE VILLAGE LEVEL IN BANDUNG REGENCY, INDONESIA: A METHODOLOGICAL EXPLORATION BY Arie Lastario Kusumadewa A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Resource Development 1975 <¢3.u<_. (>(9 Pm‘u (p¢eo> C><fi ham; :0».le DEDICATED TO THE REPUBLIC OF INDONESIA .095:th 3% a... C0,: ( S.- .00 0 6 N 6 nag 59‘ em, 9350 Qua ”he on. _ 08 $88 coaaosn 6 a 32th QO‘QE-QI .3036 <_wmzooz_ % .nRUnmw... :3. u 395% Q «3.23m ii ACKNOWLEDGMENTS The writer wishes to express his sincere appre- ciation to Dr. Milton H. Steinmueller who, as Chairman of the Guidance Committee, gave his help, encouragement, support and friendship throughout my entire doctoral program. Gratitude is also expressed to the following members of my Committee: Dr. Raleigh Barlowe, Dr. William J. Kimball, Dr. Daniel E. Chappelle, Dr. Robert J. Marty, and Dr. Kirkpatric Lawton, all of whom gave their time and support throughout my doctoral program. In addition, the writer is indebted to first, the government of the Republic of Indonesia which sent him to Michigan State University to pursue the program-- special thanks here to Dr. I Made Sandy, Director of the Directorate Land Use, and Mr. Tojiman Sidikprawiro MPA Chief of the Bureau for Education and Training, both in the Department of Home Affairs, who continuously encouraged the completion of the study; and second, to US-AID mission to Indonesia which provided the funds that enabled the writer to stay in the United States during the course work and writing stage. iii The writer wishes to acknowledge the Agricultural Development Council (A/D/C) mission to Indonesia which gave financial support during the initial field work in Indonesia, the Bupatis of Sumedang, Cianjur and Bandung Regencies who graciously welcomed the intensive field work in their regions, and the Ford Foundation mission to Indonesia ("Program Latihan Penelitian Ilmu-Ilmu Sosial") which provided funds during the concluding phase of the field work. It would be most difficult to adequately thank all of the writers, friends and colleagues in his home country and in the United States who have made innumer- able contributions to the enjoyment and accomplishment of this endeavor. Therefore, I simply will say "thank you" and hope that each of you will feel the warmth of my sincere appreciation. Finally, to my wife Ratna Mayar and my children, Rina, Restu and Ridho, who gave and still will give unending love, encouragement, and understanding, I offer my warmest gratitude. To our parents, too, for their constant support and encouragement, we express our deepest appreciation. iv TABLE OF CONTENTS Chapter Page I. INTRODUCTION . . . . . . . . . . 1 The Framework . . . . . . . . . 1 Stage of Development. . . . . . . 3 The Role of Villages. . . . . . 11 The Role of Upper Government . . . . 13 The Objectives and Hypotheses. . . . 14 The Research . . . . . . . . . 17 II. METHODOLOGY . . . . . . . . . . 20 Selection of Study Areas . . . . . 20 Data Collection . . . . . . . . 26 Data Analysis . . . . . . . . . 28 Conclusions. . . . . . . . . . 36 III. THE IDENTIFICATION OF VARIABLES. . . . 38 Intensive Data, Household Level . . . 38 Familial Variables. . . . . . . 38 Pattern of Familial Variables . . . 45 Extensive Data, Village Level. . . . 49 Community Variables . . . . . . 49 Pattern of Community Variables. . . 50 Secondary Data, District Level . . . 59 Environmental Variables . . . . . 60 Pattern of Environmental Variables . 60 Aggregate Data, Multi-Level . . . . 65 Combined Variables. . . . . . 66 Pattern of Combined Variables . . . 66 Conclusion . . . . . . . . . . 73 Chapter Page IV. THE APPLICATION OF MULTI-LEVEL ANALYSIS . . 75 Description of Bandung Regency . . . . 75 Physical Characteristics. . . . . . 76 Demographic Characteristics. . . . . 77 TeChnOlogy O O O O I O O O O O 78 Economics . . . . . . . . . . 82 Government Activity . . . . . . . 83 The Pattern of Conceptual Variables. . . 90 Higher Order Factor Analysis . . . . . 94 Strategic Pattern for Development . . . 102 Strategic Variables Which Influence Family Income . . . . . . . . . 103 Food Cost as a Proportion of Consumption Expenditures . . . . . . . . . 112 Distance to Larger City . . . . . . 112 Membership in Cooperative Organization . 113 Education Cost as a Proportion of Con- sumption Expenditures . . . . . . 114 Asphalted Roads. . A. . . . . . . 114 Conclusion . . . . . . . . . . . 117 V. SUMMARY, CONCLUSIONS, POLICY IMPLICATIONS, AND RESEARCH RECOMMENDATIONS . . . . . 120 smary O O I O O O O O O O O O 120 Conclusions. . . . . . . . . . . 124 Policy Implications . . . . . . . . 126 Additional Research Recommendations. . . 130 Extended Analysis Based on Present Available Data . . . . . . . . 130 Household Data . . . . . . . . 130 Village Data . . . . . . . . . 131 District Data. . . . . . . . . 133 Multi-Level Aggregate Data . . . . 133 Cianjur Regency Data . . . . . . 133 Further Research . . . . . . . . 134 Research Technique and Analysis . . . 134 vi Page APPENDICES Appendix A. THE DISTRIBUTION OF DISTRICT TYPES IN REGEN- CIES OF WEST JAVA . . . . . . . . . 137 B. THE SCATTERING OF DISTRICT TYPES WHERE THE VILLAGE SAMPLES ARE LOCATED (West Java Province). . . . . . . . . . . . 138 C. IDENTIFICATION OF DISTRICT/VILLAGE SAMPLES AND THE TYPE AND NUMBER OF VILLAGE SAMPLES: EXPECTED AND ACTUAL . . . . . 139 D. ORIGINAL RESULTS OF HOUSEHOLD LEVEL FACTOR ANALYSIS 0 C O O O O O O O O O O 145 E. ORIGINAL RESULTS OF VILLAGE LEVEL FACTOR ANALYSIS 0 O O O O O O O O O O O 147 F. ORIGINAL RESULTS OF DISTRICT LEVEL FACTOR ANALYSIS 0 C C C C O C O O C O O 149 G. ORIGINAL RESULTS OF MULTI-LEVEL FACTOR ANALYSIS 0 O O O O O O O O O O O 150 H. ORIGINAL RESULTS OF FACTORIZATION, MULTI- LEVEL ANALYSIS ON BANDUNG REGENCY. . . . 152 I. ORIGINAL RESULTS OF SECOND ORDER FACTORI- ZATION, MULTI-LEVEL ANALYSIS OF BANDUNG REGENCY O O C I C C O C O O O O l 5 4 J. ACTUAL AND ESTIMATED VALUE OF TOTAL FAMILY INCOME (Y), AND THE RESIDUALS FROM THE LINEAR MULTIPLE REGRESSION ANALYSIS (in rupiah, $1.00 = Rp. 415.00) OF BANDUNG REGENCY (1973) . . . . . . . . . . 155 SELECTED BIBLIOGRAPHY. . . . . . . . . . . 156 vii LIST OF TABLES Table Page 1. FAMILIAL VARIABLES OF CIANJUR AND BANDUNG REGENCIES (1973) . . . . . . . . . 40 2. HOUSEHOLD LEVEL FACTOR ANALYSIS. . . . . 46 3. COMMUNITY VARIABLES OF 300 VILLAGES (1973) . 51 4. VILLAGE LEVEL FACTOR ANALYSIS . . . . . 57 5. ENVIRONMENTAL VARIABLES OF CIANJUR AND BANDUNG REGENCIES (1973) . . . . . . 61 6. DISTRICT LEVEL FACTOR ANALYSIS . . . . . 64 7. COMBINED VARIABLES OF CIANJUR AND BANDUNG REGENCIES (1973) . . . . . . . . . 67 8. MULTI-LEVEL FACTOR ANALYSIS . . . . . . 72 9. THE DISTRIBUTION OF INCOME AND AMOUNT OF EXPENDITURES, BASED ON KIND OF PRO- FESSIONS (BANDUNG REGENCY, 1973). . . . 84 10. THE AVERAGE CONSUMPTION EXPENDITURES PATTERN (BANDUNG REGENCY, 1973) . . . . 84 11. THE CONTRIBUTION OF SECTORS TO TOTAL NET REGENCY PRODUCTS (NRP) OF BANDUNG REGENCY (1973) o o o o o o o o o o 85 12. THE GOVERNMENT EXPENDITURES OF 1973/1974 IN BANDUNG REGENCY . . . . . . . . 87 13. LAND TAXES AND OTHER TAXES, IN PERCENT OF GROSS PRODUCT (BANDUNG REGENCY, 1973) . . 89 14. PER CAPITA TAXES, IN PERCENT OF INCOME (BANDUNG REGENCY, 1973). . . . . . . 89 15. FACTORIZING OF MULTI-LEVEL ANALYSIS OF BANDUNG REGENCY (1973) . . . . . . . 92 viii Table ‘ Page 16. FACTOR CORRELATION MATRIX, MULTI-LEVEL ANALYSIS OF BANDUNG REGENCY (1973) . . . 95 17. SECOND ORDER FACTORIZATION OF MULTI-LEVEL ANALYSIS OF BANDUNG REGENCY (1973) . . . 97 18. SECOND ORDER FACTOR CORRELATION MATRIX, MULTI-LEVEL ANALYSIS OF BANDUNG REGENCY (1973) o o o o o o o o o o o o 99 19. THIRD ORDER FACTORIZATION, MULTI-LEVEL ANALYSIS OF BANDUNG REGENCY (1973) . . . 101 20. SOURCES FOR REGRESSION MODEL OF BANDUNG REGENCY (1973). . . . . . . . . . 105 21. SIMPLE CORRELATION MATRIX BETWEEN VARIABLE OF TOTAL FAMILY INCOME AND OTHER VARIABLES (BANDUNG REGENCY, 1973). . . . . . . 107 22. PRODUCT MOMENT CORRELATION MATRIX OF NINE- TEEN VARIABLES, MULTI-LEVEL ANALYSIS OF BANDUNG REGENCY (1973) . . . . . . . 108 23. THE REGRESSION COEFFICIENT AND ITS STANDARD ERROR IN MULTI-LEVEL ANALYSIS OF BANDUNG REGENCY (1973) o o o o o o o o o o 110 24. THE CONTRIBUTION OF VARIABLES TO THE COEF- FICIENT OF DETERMINATION (R2) IN MULTI- LEVEL ANALYSIS OF BANDUNG REGENCY (1973) . 111 25. THE CONTRIBUTION OF VARIABLES TO THE COEF- FICIENT OF DETERMINATION (R2) IN THE POLYNOMIAL EQUATION, MULTI-LEVEL ANALYSIS OF BANDUNG REGENCY (1973) . . . . . . 116 26. THE REGRESSION COEFFICIENT AND ITS STANDARD ERROR IN THE POLYNOMIAL EQUATION, MULTI- LEVEL ANALYSIS OF BANDUNG REGENCY (1973) . 118 A-l. THE DISTRIBUTION OF DISTRICT TYPES IN REGENCIES OF WEST JAVA . . . . . . . 137 B-l. THE SCATTERING OF DISTRICT TYPES WHERE THE VILLAGE SAMPLES ARE LOCATED (West Java Province) . . . . . . . . . . . 138 ix Table Page C-l. IDENTIFICATION OF DISTRICT/VILLAGE SAMPLES WEST JAVA, 1973 o o o o I o o o o o 139 C-2. THE TYPE AND NUMBER OF VILLAGE SAMPLES: EXPECTED AND ACTUAL. . . . . . . . . 144 0-1. ORIGINAL RESULTS OF HOUSEHOLD LEVEL FACTOR ANALYSIS 0 o o o o o o o o o o o 145 E-l. ORIGINAL RESULTS OF VILLAGE FACTOR ANALYSIS . 147 F-l. ORIGINAL RESULTS OF DISTRICT LEVEL FACTOR ANALYSIS 0 I O C O O O I O O O O 149 G-l. ORIGINAL RESULTS OF MULTI-LEVEL FACTOR MALYSIS C C C C O C I C O O O O 150 H-l. ORIGINAL RESULTS OF FACTORIZATION, MULTI- LEVEL ANALYSIS ON BANDUNG REGENCY (1973). . 152 I-l. ORIGINAL RESULTS OF SECOND ORDER FACTORI- ZATION, MULTI-LEVEL ANALYSIS OF BANDUNG REGENCY (1973) o o o o o o o o o o 154 J-l. ACTUAL AND ESTIMATED VALUE OF TOTAL FAMILY INCOME (Y), AND THE RESIDUALS FROM THE LINEAR MULTIPLE REGRESSION ANALYSIS (in rupiah, $1.00 = Rp. 415.00) OF BANDUNG REGENCY (1973) . . . . . . . . . . 155 LIST OF FIGURES Figure Page 1. Map of West Java . . . . . . . . . . 18 2. Generalized Land Use in West Java . . . . 22 3. The Location of District Samples. . . . . 25 4. Intensive Observation Region in West Java. . 39 5. The Location of Village Samples in Regency of Bandung . . . . . . . . . . . 79 xi CHAPTER I INTRODUCTION The Framework It is quite obvious that a nation or a community is always trying to raise its standard of living, espe- cially since doing so promotes easier communication between countries and helps remove meaningful barricades between them. Most of the time, the steps in that development process within a country are not in harmony, however, which causes the income gaps between people to become wider. The wider the gap, the more potential for conflict, since people want to be treated as equals. The success of any national development program depends on local natural endowments, human skills, favorable funds, technology and institutions. Moreover, the particular kind of development program cannot merely be the same as that for any other place, but must be fitted to local capabilities. Development projects should not be imitations, but should be designed Spe- cifically according to the local situation. Among the development planners in develOping countries, especially in southeast Asia, one question always arises: "Where should the development be started; should we start in the big cities to build industries by which we can create an economic pull, or should we start in the villages to raise the standard of living of most of the people?" If the first alternative is chosen, too much time will be spent discovering what kind of industry will fit the needs of the people; what kind of raw materials are available; what kind of skilled labor there is. Some disadvantages of this alternative are consequences of urbanization, concentration of settlements, pollution problems and urban unemployment and unrest. On the other hand, it must be recognized that this approach will probably make it possible for some peOple with special skills or advantageous associations to become relatively wealthy in the short-run. Sometimes these kinds of people in developing countries become the new entrepreneurs in the expanding economy. And, this alternative will also require a great amount of capital which will lead to dependency on other countries. If the second alternative is chosen, development beginning in the village, more time will be needed because of the large number of villages. Also, a com- prehensive preparation will be necessary so that the development plans will suit local conditions and fulfill the greatest number of local needs. Nevertheless, this second alternative will insure the expansion of develOp- ment activity and, by the same token, will widen the opportunity to increase the per capita income in rural regions so that the economic gap between the haves and the have nots can be lessened. In gearing a development program to the village level, there are innumerable variables that need to be considered. Those variables should be derived from dif- ferent levels, since considering variables from only one level will result in an incomplete analysis. In fact, policy considerations for development strategy are better if based on complete variables, which are reflected in the conceptual variables resulting from factor analysis. Stage of Development Up to the year 1945, Indonesia was under foreign domination through colonization by the Dutch, the British for a while, and then the Dutch again, and later, during the second world war, by Japanese occupation. When the Second world war ended, the Indonesians began a struggle with the Dutch for the right to control‘the country. The revolution stage lasted until 1949 when the Nether- lands withdrew its armies and conceded sovereignty over the East Indies, excluding West Irian, to the Indonesians. Since then, the Indonesians, as a new nation, have started building a government and have tried to promote national development through an "urgency program" beginning in 1951 and lasting to 1956. The first national plan was implemented during the period, 1956 to 1960, but its development was unsuccessful. In evaluating this first national plan, Higgins stated, in 1957: Now, however, Indonesia cannot afford further depletion of her reserves. Continued budget deficits, combined with continued import restrictions, could lead to a cumulative wage- price-currency-depreciation spiral, ending in the collapse of the monetary system. In a cumu- lative inflation, foreign capital is unlikely to flow in; such investments as take place will be for Speculative rather than productive purposes: and imports of materials and equipment may fall even below levels needed merely to maintain cur- rent output. Given the economic instability of this order, political instability will aggravate economic instability and so on. If the young Republic is to maintain its economic and political independence, it must resolutely pursue policies which will bring financial equilibrium. . . . In some respects, however, it would appear that the planning authorities have lacked the time, infor- mation, assured financing, and general directives needed to produce a truly comprehensive and effec- tive plan. The second national plan was the eight-year plan, started in 1961 and continuing to 1968. However, again, the eight-year development plan also failed. According 1Benjamin Higgins, Indonesia's Economic Stabili- zation and Development (New Yofk: Institute of Pacrfic fi Relafion, 1957), pp. I21-22. to the United States' economic team, sent to Indonesia to evaluate the long-term plan: The job of the National Planning Council was to weigh the relative merits of the many thousands of proposals submitted and to select those items which they felt should receive the highest pri- ority. Priorities were not always based on the immediate and direct contribution a project might make to economic development. In addition to the aim of fostering more rapid economic growth, the Council obviously had other (and sometimes con- flicting) objectives in mind. For example, self- sufficiency in food and clothing was a paramount goal, even though in some degree it conflicts with an efficient use of resources and with develOp- ment; cultural values were to be stressed . . . the military's roughly 50 percent share of the regular government budgets, reflecting nationalistic concern over West Irian.2 Later, in April 1969, Indonesia initiated a new five-year national develOpment plan which ended in March 1974. During the first five-year plan, Indonesia's development goal was primarily to increase the output of agricultural products, especially food crops, and also to increase the export capability in raw materials. What happened within this period? Was the country facing unsuccessful development again? Mangkusuwondo stated: On the whole, the first four years of Repelita I (first five-year plan since the new regime) were quite successful. Gross domestic production increased at the average of 5.3 percent per year between 1966 and 1971. Under the first five-year plan, between 1969 and 1971, the rate of growth 2Indonesia: Perspective and Proposals for United States EconomIE‘Aid (New Haven: YaIe University Southeast ASIa Studies, 1963). PP. 84-85. was significantly higher at 7.0 percent per year. This is a respectable rate of growth by normal standards.3 During the same period, the rate of growth of p0pu1ation was about 2.6 percent per year. The success of development in that period is shown by high rates of growth in mining (11.3%), manufacturing industries and construction (21.3%), trade (11.8%) and banking (23.5%). The rate of growth in the agricultural sector was 1.4 percent per year; this growth occurred mainly in forestry and rice production. Since the share of the agricultural sector is 50 percent of the total GDP, its role in the rural regions is still important. At the present, Indonesia is in the period of the second five-year plan (1974 to 1978), in which the emphasis is on the processing of agricultural products leading to industrialization and also the fulfillment of food stuffs for the people. In the meantime, the export of minerals (especially oil) and timber is con- tinuously on the increase, and concordance with grant aid and loans from other countries will cause the development programs to move faster. The current consequence of such development is the broadening of the income gap between groups of people, a disparity 3Suhadi Mangkusuwondo, "Dilemnas in Indonesian Economic Development," Bulletin of Indonesian Economic Studies, 9 (July 1973): 30431. 7 which is dueto the fact that the changes in the role of the sectors are occurring faster than the shift in pro- fessions. The declining average of agricultural employ- ment to nonagricultural professions is one percent per year,4 but the contribution of the nonagricultural sector to the GDP is increasing--4 percent between 1968 to 1970,5 or 2 percent per year. Since the profession of the majority of Indo— nesians is farming (more than 70% are peasant farmers) and most of them live in the villages, the greatest number of the people who need to be helped are in this group.6 The Indonesian government has been taking steps to spread the increase in per capita income to the rural regions by undertaking programs that are oriented to the rural regions. Examples of these programs are: (1) Village subsidy (2) Regency subsidy (3) Intensive labor work system 4Dwight Y. King, Social Development in Indonesia (Jakarta: Biro Pusat Statistik, I973), p. I} 5H. W. Arndt, "Survey of Recent Development," Bulletin of Indonesian Economic Studies, 7 (July 1971): 3. 6Mubyarto, Pengantar Ilmu_§grtanian (Jakarta: Lembaga Penelitian, Pendidikan dan Penerangan Ekonomi & SOSial, 1973): p0 370 (4) Resettlement program (5) Farm credit program (6) Public elementary school program (7) Public health and family planning (8) Other sector-oriented programs, such as inten- sive guidance systems in paddy ricefields, nutrition programs, informal education and rural radio programming. The purpose of the village subsidy program is primarily to stimulate the village government in mobiliz- ing local funds for local development. The subsidy is as much as 200,000.00 rupiah, or about $500.00 for each village. Together with local funds (usually up to four times as much as the subsidy), mutual activities and collective actions, the village subsidy funds should be used for the development of village public projects, such as building a small earth dam for irrigation or a water reservoir and simple, small bridges on village roads, and extending school buildings. In this way, the village headman and the staff will learn, by exper- ience, how to prepare sound and appropriate plans, to mobilize funds and to manage the Operation of programs. The subsidy is given by the central government to every village once a year.7 7Hariri Hady, "Pembangunan Daerah Dalam Repelita II," Prisma, 3 (April 1974): 69. The regency subsidy is given by the central government to the regency government, based on the size of the regency population. This subsidy is also pro- vided only once a year, and the amount is 300 rupiah or $.75 per person. The objective of the program is to step up the rehabilitation and expansion of the economic infrastructure, to create larger employment oppor- tunities and to enlarge the participation of regions and sub-regions in the planning and execution of develOpment projects.8 Most of the regency subsidy funds have been used for rehabilitation of inter-district and inter-village (intraregency) infrastructure such as irrigation sys- tems, roads and bridges. The difference in development procedure between the village subsidy and the regency subsidy is the size and the way of executing the project. The regency subsidy program is for financing a project in a business way and does not include mutual activity. But the purpose is the same as that of the village sub- sidy, building construction for public use. The intensive labor work system is the same program as the food for work system. The government gives wheat and a certain amount of money (three kilo- grams wheat plus thirty-five rupiah for a day of work per person) to critical regions which annually suffer a 8Atar Sibero, "Program Bantuan Pembangunan Kabupaten/Kotamadya," Ekonomi dan Keuangan Indonesia, 21 (Juni 1973): 95-112. 10 lack of food and concentration of unemployment. In these critical regions, the surplus labor force is organized into a working group which is then employed in public construction and is paid with wheat and money according to the number of days worked.9 The resettlement program, as well as the so- called transmigration program, involve the arrangement of moving people out of the densely populated regions to less populated areas. The government drafts the people (voluntarily) who want to be transferred. Preparation of the new region is also done by the government. Then, during the first two years' stay in the new region, the government gives subsidies in the form of food, farm tools, petroleum, and other needed items to the new settlers. In the new region the government also provides arable land for cultivation (two hectares or five acres per family), a temporary house, an elementary public school and a public health 10 centre . The farm credit program is a program by which the government, which is very concerned about the 9Arie L. Kusumadewa et al., Laporan Penelitian Padat Karya di Jawa dan Madura (Jakarta: Team Penelitian Padat Karya, 1972), pp. 38-41. 10Subroto, "Kebijaksanaan di Bidang Kesempatan Kerja & Transmigrasi dalam Repelita II," Prisma, 3 (April 1974): 18-29. ll availability of food for the people, gives soft loans (credit) to farmers who need funds to bring the "new input" to their crops. This soft loan is limited only to paddy rice production, but will be extended to animal husbandry as well as production for export commodities. If the yield fails, the payment can be delayed to the next harvest time. The public elementary school building_and public health and family planning programs are both new programs. Every regency is assigned a quota of a number of buildings that can be constructed each year. The regency level decides the region/village that gets the new building for a certain year. In this way, within five years all districts will have public health and family planning centers, and all villages will also have school buildings. Several other sector-oriented programs are also conducted at the rural level, such as intensive guidance systems in the paddy ricefields (since 1963), nutrition programs, adult education and evening schools, technical vocational education, and rural radio broadcasting system. The Role of Villages The government body nearest the people is located in the village where the community lives together with its formal institutions and within the boundaries of the region. The people who live in the same village feel that they are bound together and need to work 12 hand in hand to solve their problems.11 The village life attitude is completely opposite to that of the urban life where the people are more individualistic. The "desa" is seen as the "third order" autonomy level of government, but in Java, generally, it is doing the job without any legislative body. The village headman's status is that of an elected leader by popular election; he is appointed to this post by the regency head on behalf of the "government." The staff of a village head is appointed by the village head himself. Being no government employee, these village officials have no salaries.1 The village people are easier to organize since most of them want to follow the local leaders, who have a big role in directing development activities. At the present, the formal local leaders are the village headman and his staff, who run the village government. There is— also a group of village people who join in a formal organization called The Self-Defense Group. Most of the time, this group has been also a pioneer in develOp- ment activities. School teachers also act as innovators in the villages. They bring new ideas from the "outside world" and introduce them to the community. In many cases, the 11 Arthur F. Wileden, Communit Develgpment (Totawa, N.J.: The Bedmister Press, E970), ppi'z-lo. 12 Sajogyo,"Modernization Without Development in Rural Java"(Bogor, Indonesia: Bogor Agricultural Uni- versity, 1973), p. 65. A mimeograph paper contributed to the study on changes in agrarian structure, organized by F.A.O. of the U.N., 1972-1973. A "desa" is a village. 13 teachers have the role as key persons in motivating development programs; later the village assembly decides what kind of program and project needs to be undertaken every year. The Role of Upper Government The central government has to decide the policy guidelines for the development programs in the rural regions, the unit of which is the villages. After all, the village is the only place where the developmental department can conduct its program and activities. For that reason, development guidelines are very important in order to avoid conflict and overlap of activities which would confuse the local peOple. The province government is responsible for the planning coordination among developmental department offices within the province to meet the central govern- ment policy guidelines for development programs. The province government also has the responsibility of inspecting whether the programs and projects are con- ducted according to the regulations that have been designed uniformly for the entire province. The regency government has to decide the priority of development projects for a certain region/district. They have to look to whether or not the projects can work in harmony or need to be cancelled. The coordination 14 of project and program execution is directed by the regency government according to the availability of project funds. The district supervisor is the lowest central government employee who has the responsibility of reporting to the higher level that all programs of development in his region have been conducted as they are supposed to be. He also has the right to postpone or to channel the development proposals from the village level to the regency government. And the district supervisor has the coordination function among the technicians who are stationed in the same district. The Objectives and Hypotheses There are several research studies that have been done on the village level with anthropological as well as sociological approaches. However, those studies were a qualitative foundation for further findings. As Koentjaraningrat stated: A more intensive knowledge of these phenomena, in specific sociocultural settings, will enable us to formulate with greater accuracy the problem and hypothesis on the social system of villages in Indonesia in general . . . the validity of which has to be tested by more sOphisticated quantitative methods.13 13Koentjaraningrat, ed., Villa es in Indonesia (Ithaca, N.Y.: Cornell University Press, I967), p. 586. 15 As a matter of fact, the research in this dissertation represents a quantitative approach and an exploration in multi-level analysis which needs further intensive observation for special cases. There are many variables that influence the increase and decrease of the per capita income in a community. Those variables could be familial variables that vary in relation to the daily life style of the people; they could also be community variables that can be changed by the society members who live together; or they could be environmental variables that the people as well as the society cannot change since they are given variables. These three distinct levels of variables will influence the per capita income of the society dif- ferently. Yet, in reality, all three levels of variables will not influence the per capita income separately, but will combine and be interrelated to each other in inducing the level of per capita income. Identification of so many variables in every level can be simplified by grouping those variables which are of about the same nature in a certain group separate from another group of variables which are not similar. Such a categorization of variables is made just to simplify the location of those variables which are the most influential with regards to the per capita 16 income according to the hierarchy of responsible levels of group variables. By analyzing multi-level-wise, group variables can be found which correlate with each other vertically. To accomplish the objectives of the research, the hypotheses of the study have been set as follows: Hypothesis I: There are certain variables and group variables that influence per capita income generation. The knowledge of these variables can be used to determine investment priorities in development programs designed to raise the income level of the society. A. Some of these variables reside at the household level, i.e., sources of income, expenditure pat- tern, membership in organizations, participation in development programs, and variables for food crop production. B. Other variables originate at the village level, i.e., physical condition of the village, the economic situation, demographic variables, public construction and the institutional variables. C. Several variables can be found at the district level, i.e., soil types, land use classes, trans- portation system, population and professionals variables. Hypothesis II: The three levels of variables can be analyzed and some can be discarded from further analysis. Those that remain can be grouped into what can be called conceptual variables at each level as described in A, B, and C. 17 Hypothesis III: If the three different levels of variables are aggregated, the combined variables can be built into extended conceptual variables and hypothetical statements. These hypothetical statements explain the development process and develOpment stage of the region. The Research The research was conducted in West Java province, Indonesia. The pre-test of village questionnaires was done in 1972, and later, in 1973, extensive research in 300 villages was undertaken. The pre-test of household questionnaires was done in 1973 in the regency of Sumedang. Later, in 1974, intensive observation was made of 699 households, 88 villages, and 44 districts, as well as regency data collected in the regencies of Cianjur and Bandung (see Figure 1). This extensive research was undertaken with the cooperation of the Directorate of Land Use, Agrarian Affairs and the Department of Home Affairs. The inten- sive observation research was done with the cooperation of the regency governments of Cianjur and Bandung (a special report has been delivered to both regencies). Funds were provided by the Directorate of Land Use, as well as the regency government of Cianjur and Bandung. The research was also beneficial for training purposes in the preparation of questionnaires, collection of data, data handling and the analysis of data recorded 18 95h. Pmoz no em: . H 6.5th at? 3:3 .33.ng 383mm Jfiaoosoa 13m Hwfimdo .mououmm «0 53.33 Season gamma .I >I.’ 9.30s...“ 0 . ./..U;..oc8=3 . / 8. O . 1.68.5 .s .I. —.\l.|./ .. 1 «iii; \. . €239.23“. . . 009.com; .— <>3. was; 5 on: “Ema vouHHMhoCoo . m ohsmfm .36 a con sag» whoa as mafivdpuaam d unease .3.» a com I 8H «d can." has _ . .... a E m 1.111.... com . 03 pa 303358 a Hence 48 a 8” .. o a... 33 a? «83383 a $835 8m 8% Boa 8 36382 pang 43 a 8m 53 43 Son 8 82 .5 ...z... a 8m .. 8H s. 3038“.” Hui . O O y .\ I D U D > .19. . o o n o. r , : .... o c u u . . '4 . 1 IIIII \ . run V O I llllll . o ... . o o . IIIIIIIIIII -r as. o o s so I . a o o o . . o . I. I . llllllldllIlDl I w. A . . eon - o . o . II IHIIIIIIIIIIIIIH . . . o I III IIIIIIIIII IIIII . . , .H . .. IIHIIIIIHIHMIHIHIHIHIHI a. .... . ”wuunnmuunuwnuunuuuc , I I IIWIIHIHIHIM. . . .n. I I I I l I I l .. . ... . . , k . . .. .. . . .H. a . . .. . .p A” 000.com?" . H H. o % <><_. emu; . ... 9 ; 23 diagonal line running from the northwest to the southeast of West Java and on the distribution of district types in the regencies of West Java. The distribution of district types can be found in Appendix A and the scattering of district samples in Appendix B. Since employing a district as a research unit would be impractical due to the large size of a district and the low reliability of available data at the district level, two villages were selected from every district to represent the district sample. In every district, a list of villages in rank order based on their develop- ment stage was made. Then, the listed villages were categorized into five groups. The first group repre- sented the most developed villages in the district and the fifth group, the least developed villages. The third group contained the medium developed villages. Two village samples for the districts were taken, one village each from the second and fourth groups. In other words, from every district there were two village samples, one representing above medium development and another one representing below medium development. When the two villages are averaged, a median village of the district is obtained. From the whole of West Java, 150 districts were taken (out of 361 districts, 41.5%), and in every dis- trict two villages were taken or 300 villages from the 24 whole province (out of 3,835 villages, 7.8%). The location of the district samples can be seen in Figure 3, and the list of the district village samples can be found in Appendix C-1 and C-2. The intensive observation research was conducted in Cianjur and Bandung regencies. Both regencies have only seven strata out of the nine strata that were identified. The absent categories are: (l) Paddy rice field at 0 to 100 meters asl. and (2) Forest and plan- tation at 0 to 100 meters asl. See Appendix B. Two villages in all of the districts of the Cianjur and Bandung regencies were selected as village samples. The procedure for drawing these village samples was the same as that for the extensive research, one village representing above medium development and one village representing below medium development. In every village, seven to eight households were chosen randomly. In drawing sixteen household samples in every district, an attempt was made to include dif- ferent kinds of existing professions such as, food crop farmer, animal husbandryman, fisherman, merchant/trader, home industryman, excavation worker, and service employee. In the district in which intensive observation at the household level was made, secondary data at the district level were also gathered. These data were checked against and augmented by data available at higher levels (regency, province, and national). moaneum voahpmfin mo :oflpwoon one . n mhswdh 333...? 3. . .3. a com 8a.» 38. 8% < m ”Assumed 3 ends handguns '09” m o m . . ma I .300va .8 Augean » Eu. we 3 H a 8m 03 . flotsam. R . 43 a o3 . o D :ofivm>mam hp wonky Poahvmao ho mcfiuopvaom one 3>§<_. hmmi ‘0 26 Data Collection When the enumerator visited the districts that were selected at random, the district supervisor was asked to list the villages in the district in rank order according to level of development. The enumerator then checked the list with several local key persons to make sure that the rank order was accurate. According to the list, the villages were divided into five groups, and one village was selected randomly from every group numbered two and four. This meant two village samples for every district; one above and one below average development. An inventory of the kinds of professions within the district was also made. For the purpose of village-extensive research, a questionnaire had been made up beforehand. The first group of questions concerned such items as village name, majority of land use and elevation type, size of popu- lation, and other general attributes of the village. The second group of questions concerned variables found in five sub-groups: physical variables, human variables, fund variables, technological variables, and institu- tional variables. Finally, the third set of questions was addressed to the nine major village sectors: food crop sector, plantations sector, forestry sector, animal husbandry sector, fishery sector, industrial sector, trade sector, transport and communication sector, and 27 services sector. The value of production in these sectors can be summed up as Village Gross Product (VGP). In every village, there are neighborhood associ- ations (in Indonesia called "rukun tetangga," or RT, and ”rukun kampung,“ or RK). In the regencies of Cianjur and Bandung where the intensive research was conducted, the village headman was asked for a list of the neighborhood associations from which a random sample of households was made. From every household association one sample was taken (out of about twenty to forty households). The kinds of samples in each village consisted of several types of professions for the purpose of meeting the representativeness of professions according to the prior list that was collected at the district level. The maximum number was seven to eight household samples for each village, and thus a fourteen to sixteen house- hold sample for each district. A prepared questionnaire was also used at the household level. Questions were again grouped into five major categories regarding: (l) The household income (sources and amounts); (2) The production function (inputs and outputs); (3) The expenditure pattern (consumption, production, and investment); 28 (4) Membership in any organization that exists in the village; and (S) The attitude of the sample family toward develop- ment issues. In the regencies of Cianjur and Bandung, at the district level, secondary data concerning different kinds of sectors were gathered from several sources in order to attain a general idea as to what kind of sector domi- nates the region and defines the environmental situation of the two villages. At the regency level, secondary data were obtained about budgetary allocations which were also considered as environmental conditions. Data Analysis There were numerous variables that were collected from each level, and an even greater number if all the variables that came originally from the three different levels are combined. To approach this problem, factor analysis was used. The single most distinctive characteristic of factor analysis is its data reduction capability. Given an array of correlation coefficients for sets of variables, factor analytic techniques enable us to see whether some underlying pattern of relationship exists such that the data may be "arranged" or "reduced" to a smaller number of factors or components that may be taken as source variables accgunting for the observed interrelation in the data.1 15Norman Nie, ed., SPSS: Statistical Packa e for the Social Science (New York: Mchaw-Hill Inc., 19 , p. 209. 29 Factor analysis, then, is a technique by means of which a large number of variables may be clustered on the basis of their intercorrelations, each set of which is presumed to reflect a single dimension which is causing the association within the set of variables.16 Factor analysis is both a hypothesis creating and a hypothesis testing method. It can be applied across different facets of the basic data relation matrix in a variety of techniques, with and with- out time sequence arrangement, or manipulative control of variables.1 The technique of factor analysis is still being improved, especially with the advancement of the com- puting system. However, it is not the purpose of this dissertation to argue the technique of factor analysis; rather factorization is merely treated as a tool for the data analysis. After the variables had been reduced by using this technique, multiple regression analysis was employed to ascertain the role of significant variables and factors which could explain the income variable, which was the dependent variable. In employing factor analysis and regression analysis, a package program called SPSS (Statistical 16Phillip M. Gregg and Arthur S. Banks, "Dimension of Political System: Factor Analysis of a Gross Polity Survey," American Political Science Review, 59 (September 1965): 602-14. 17Raymond B. Cattel, "Factor Analysis: An Intro- duction to Essentials. II. The Role of Factor Analysis in Research," Biometrics, 21 (June 1965): 405-35. 30 Package for the Social Sciences) prepared by Norman Nie, et al. was used. This package program was developed at Northwestern University, and available in the Computer Center at Michigan State University in East Lansing, Michigan. Most of the variables in this study were para- metric variables; even part of the household data which originally was nonparametric data was converted to para- metric data. R~factoring was used for the preparation of the correlation matrix, that is, correlation between 18 The extraction of the initial factor variables. employed principal factoring without iteration (in SPSS this is called the PAl technique). In the rotation to terminal factors the oblique factors were used because, as Nie mentioned, " . . . the oblique factors are empiri- cally more realistic."19 The first step was to construct a product moment correlation coefficient matrix of the variables and the next step was to factor analyze. The matrix of cor- relation coefficients was then collapsed into the smallest possible number of columns; each column repre- sented a reference factor or component which contained 18Benjamin Fructer, Introduction to Factor Anal sis (Princeton, N.J.: D. Van NostrandT Co., Inc., I p. 2020 ' 19Nie, SPSS, p. 212. 31 cell values across factors. These factor loadings, which make up the cell values of the reference factors, can be interpreted as partial correlation coefficients that partition the variation associated with any one variable among the factors. The basic factor postulates: F + aj2 F2 + . . . + ajm Fm + dj Uj Z = variable in standardized form, known data a; ll hypothetical common factors 3 = l, 2 . . . n, are variables i = l, 2 . . . m, are common factors standardized multiple-regression coefficient of variable j on factor i, and is called factor loading 01 11 U. = the unique factor for variable j d. = the standardized regression coefficient of 3 variable j on unique factor j The basic problem of factor analysis is to determine common factor loading, that is the ajm's. The a2jm is the proportion of variance of variables explained by factor m. The total amount of variance accounted for by a factor is calculated by adding the square of the loadings in each column: 32 2 Variance accounted for by Factor 1 = a jl "tab j 1 (j = 1' 2’ o o s n) The respective value is also called eigenvalue. The proportion of total variance explained by a factor is: Proportion of total variance accounted for by n 2 Factor 1 = 2 a.1 e n j-lj The total variance of variables accounted for by the combination of all common factors, designated hi, is usually referred to as the communality of the variables: Only loadings of the common factors are utilized in computing communality. The proportion of common variance (the variance accounted for by all the common factors) accounted for by each factor is calculated as: Proportion of common variance accounted for by n Factor 1 = 2 3' 33 Common factors involve more than one variable; general factors involve almost all variables which load highly on one factor; and group factors involve more than one variable, but not all variables are loaded on the factor. Unique factors involve a single variable. Common factors account for the intercorrelation of variables, and unique factors represent that portion of a variable not accounted for by its correlations with other variables in the set. As has been mentioned before, oblique rotation was used. The idea was to minimize the cross products of the factor loadings on reference axes in order to simplify the primary factor loadings. The generic name of the rotational method based on this idea is indirect "oblimin," and can be expressed as:20 m n d n n z (z a2 a2 —— z a2 2 a2) p < q = l j = 1 39 jq n j = 1 jp J = 1 is where: p < q = common factors factor pattern loadings m m ll d = an arbitrary value which can be used to control the obliqueness of the solution. This analysis used d = 0, that is a fairly oblique (correlated) solution. 20Ibid., p. 225. 34 By using oblique rotation, a factor pattern matrix and a factor structure matrix can be derived. The difference between both matrices is that the pattern loadings distinctly display the patterns. The structure loadings, however, do not display them well; instead, they measure the correlation of variables with the patterns.21 Factor scores of component variables and set of variables that originated from the product moment cor- relation matrix were analyzed by using multiple regression analysis. Those factors and variables were then treated as independent variables and correlated to income as the dependent variable. A stepwise method was practiced. . . . the partial F criterion for each variable in the regression at any stage of calculation is evaluated and compared with a preselected per- centage point of the appropriate F distribution. This provides a judgment on the contribution made by each variable as though it had been the most recent variable entered, irrespective of its actual point of entry into the model. Any variable which provides a nonsignificant contribution is removed from the model. This process is continued until no more variables will be admitted to the equation and no more are rejected.22 21R. J. Rummel, "Understanding Factor Analysis," Conflict Resolution, 11 (December 1967): 444-80. 22N. R. Draper and H. Smith, Applied Rggression Anal sis (New York: John Wiley 5 Sons, Inc.,31366), p. s 35 The classical model of multiple regression is: Y=a+b1Xl+b2X2+....+ann where: Y = the dependent variable X = independent variables a = the constant b = coefficients which explain the individual variables n = the number of variables: 1, 2 . . . . n. By having the b coefficient, the dependent variable (Y) can be predicted. Based on the factor loadings of each factor, the factors can be translated to conceptual variables. In the multi-level analysis these can then be translated to hypothetical statements. The oblique factorization will produce factor correlation matrix, and this cor- relation matrix can be analyzed further for higher order factorization. Bandung Regency will be used as the region where the application of multi-level analysis will be demon- strated: first, the factor analysis to construct hypo- thetical statements; later, the multiple regression analysis to find out strategic variables for development. 36 Conclusions The province of West Java in Indonesia was selected as the location of the research. The whole region is divided by nine strata: paddy rice fields at 0 to 100 meters asl, paddy rice fields at 100 to 500 meters asl, paddy rice fields at more than 500 meters asl, dry land farming at 0 to 100 meters asl, dry land farming at 100 to 500 meters asl, dry land farming at more than 500 meters asl, forest/plantations at 0 to 100 meters asl, forest/plantations at 100 to 500 meters asl, and forest/plantations at more than 500 meters asl. Extensive village data were collected in the whole region representing the nine strata. For the purpose of multi-level analysis, intensive household data and secondary data at the district level were also collected in the regencies of Cianjur and Bandung. For the household level, the village level, and the district level, a number of variables were identified through factor analysis in order to get the component variables or conceptual variables. In the multi-level analysis, this leads to the construction of hypothetical statements. The application of multi-level analysis, by which the holistic nature of variables in certain regions 37 will be conserved, will be demonstrated in Bandung Regency as will higher order factorization and multiple regression analysis. CHAPTER III THE IDENTIFICATION OF VARIABLES Intensive Data, Household Level Intensive data for the household level were col- lected in two regencies, the regency of Cianjur and the regency of Bandung, both in West Java. Information regarding household conditions in 1973 was collected in 1974 and comprised 699 families. For a clearer idea of the location of the regencies, see Figure 4. Familial Variables The variables for the household level are called familial variables. Most of the data dealt with sources of income, expenditure patterns, memberships in organi- zations, participation in development programs, variables for food crops production, and other household activi- ties. A complete list of these variables can be found in Table 1. There are ninety-one familial variables that were originally selected and manipulated from more than 110 raw variables on the questionnaire sheets. 'However, not all of those ninety-one variables were analyzed due to the capacity of the package program 38 39 w>wh Hum: :fi cofimom ccfipm>hmwoo w>wmcmpsa . : onswwm was. 5 £83 3388a 33305 .33 3.673 a 2.3 fl manganese $3..ch m sofipm>pmmno .\ . . . s 352.20... \. s. .. 1.! .I s . \ s .0 .- .‘n- s s \ ...su.......Ii.: ’ I ’. 25:89.83... . ...-32v. . ... . r .l. 4.\1 I I \ asl-oust \ \ .\ o . s I I \ .\. ‘ O .0 o o .‘n’ o ‘- \ I I. -0’.‘ I I .. . 35.... .o . .I 38.0. 9 ..¢ uses-Dir . .f. O O. I ‘o’ .0 0’ s I s . . :5 8.. as _ . . ... .. (>(q hum; o>fimcovcH pom mowocmmom mo coflpmooq one 40 TABLE 1 FAMILIAL VARIABLES OF CIANJUR AND BANDUNG REGENCIES (1973) S-l S-2 S-3 8-4 8-5 8-6 8-7 S-8 8-9 8-10 8-11 8-12 S-13 8-14 8-15 8-16 8-17 S-18 8-19 8-20 8-21 8-22 8-23 *Primary profession of the head of family *Secondary profession of the head of family Number of family dependents, in persons Land ownership, in hectares Land operation, in hectares Total family income, in rupiah *Food expenditures, in rupiah *Clothing expenditures, in rupiah *Medicine expenditures, in rupiah *Education expenditures, in rupiah *Durable goods expenditures, in rupiah *Gift expenditures, in rupiah *Recreational expenditures, in rupiah *Other expenditures, in rupiah Total consumption expenditures, in rupiah Total house construction expenditures, in rupiah *Land investment expenditures, in rupiah *Farm tool expenditures, in rupiah *Vehicle investment expenditures, in rupiah *Building investment expenditures, in rupiah *Miscellaneous expenditures, in rupiah Totalgproduction investment expenditures, in rupiah Total money savings, in rupiah 41 TABLE l--Continued 8-24 8-25 8-26 S-27 8-28 8-29 8-30 8-31 8-32 8-33 8-35 8-36 8-37 8-38 8-39 8-40 S-41 8-42 8-43 8-44 8-45 8-46 8-47 Total expenditures, in rupiah Membership in a religious organization Membership in a Boy Scout organization Membership in a self-defense organization Membership in c00perative organizations Membership in sport organizations Membership in cultural organizations Membership in a professional organization Membership in loan unions Number of organizations to which head of family belongs Number of organizations to which members of family belong Presence during the most recent election day Election participation of head of family Intensive guidance system understanding Intensive guidance system participation Family planning understanding Family planning_p§rticipation Multi-cooperative program understanding Multi-cooperative program participation Source of information about the recent programs Reading newspaper frequency Radio listening frequency Kinds of radio programs listened to Memory of last decade's land reform program 42 TABLE l--Continued 8-48 8-49 8-50 8-51 S-52 8-53 8-54 8-55 8-56 8-57 8-58 8-59 8-60 8-61 8-62 S-63 8-64 8-65 8-66 8-67 S-68 S-69 Involvement in last land reform program Stage of involvement in last land reform program Total food crop production, in rupiah *Seed cost, in rupiah *Fertilizer cost, in rupiah *Insecticide cost, in rupiah *Real labor cost, in rupiah *Farm tool cost, in rupiah *Land rent cost, in rupiah *Land taxes, in rupiah *Miscellaneous cost, in rupiah Total_production cost, in rupiah Per capita income, in rupiah Consumption expenditures as a proportion of income, in percent House construction expenditures as a proportion of income, in percent Production investment as a proportion of income, in percent Money savings as a proportion of income, in percent Per cgpita expenditures, in rupiah Food crOp production cost per hectare, in rupiah Seed cost as a proportion of production cost, in percent Fertilizer cost as a proportion of production cost, in percent Insecticide cost as a proportion of production cost, in percent 43 TABLE l--Continued 8-74 8-75 8-76 8-77 S-78 Labor cost as prOportion of production cost, in percent7 Farm tool cost as a proportion of production cost, in percent Land rent as a prOportion of production cost, in percent’ Land taxes as a proportion of production cost, in percent Miscellaneous cost as a proportion of production cost, in percent Value of food crop production per hectare, in rupiah Income from nonagricultural sources, in rupiah *Per capita income from nonagricultural sources, in rupiah *Per capita income from agricultural sources, in rupiah Food cost as a prOportion of consumption expen- ditures, in percent Clothing cost as a proportion of consumption expenditures, in percent Medical cost as a proportion of consumption expen- ditures, in percent Education cost as a proportion of consumption expenditures, in percent Durable goods purchases as a proportion of con- sumption expenditures, in percent Gifts as a proportion of consumption expenditures, in percent Recreational spending as a proportion of con- sumption expenditures, in percent Other expenses as a proportion of consumption expenditures, in percent 44 TABLE l--Continued Production investment in land as a proportion of total investment, in percent Production investment in farm tools as a proportion of total investment, in percent Production investment in vehicles as a proportion of total investment, In percent Production investment in buildings as a proportion of total investment, in percent Production investment in miscellaneous items as a proportion 5? total investment, in percent NOTE: 8 is hou"S”ehold variable code number * Variables excluded from factor analysis 45 and the limitation of the computing system. Those variables having about the same nature were excluded. Pattern of Familial Variables Sixty-six selected variables were analyzed by using factor analysis. From the unrotated matrix, the result found by free factoring with a minimum eigenvalue of 1.0 was in twenty-two factors with a cumulative total variance of 66.6 percent. By reducing the number of factors to thirteen, there was still a cumulative total variance of 51.0 per- cent and a minimum eigenvalue of 1.4. The original results of the rotated matrix of the household factor analysis can be found in Appendix D. The main result of this factorization was an aggregation of sixty-six variables in thirteen patterns. If all variables with factor loadings lower than .40 are eliminated, forty-nine variables remain. If the patterns and variables are arrayed in rank order, the final results can be identified as those found in Table 2. By observing the data in Table 2, the household patterns can be named: Pattern One: Organizational Participation Out of the ten organizational variables that were analyzed, seven emerged with high factor loadings and came out together in the same pattern. This first 46 TABLE 2 HOUSEHOLD LEVEL FACTOR ANALYSIS Factor Number Variables 7 3 5 4 9 6 l 10 2 8 13 12 11 8-64 Money savings .95 8-23 Total money savings .94 8-61 Consumption .92 S-79 Food cost .87 8-82 Education cost .67 8-80 Clothing cost .64 8-22 Total prod. investment .91 8-63 Production investment .87 8-87 Investments in land .75 8-62 House construction .90 8-16 Total house construction .87 8-65 Per capita expenditures .71 8-24 Total expenditures .57 .57 8-81 Medical cost .51 8-5 Land Operation .90 8-4 Land ownership .90 8-59 Total prod. cost .45 .45 8-76 Income from nonagric. .89 8-75 Value of food crop .87 8-50 Total food crop prod. .85 8-33 Org. head of family .86 8-34 Org. members of family .83 8-29 Sport org. .72 8-30 Cultural org. .69 8-26 Boy Scout org. .61 8-28 Cooperative org. .59 8-32 Loan union .57 8-48 Involvement land reform .81 8-47 Land reform program .66 8-49 Stage involvement .58 8-6 Total family income .79 8-15 Total consumption .79 8-3 Family dependents .49 8-60 Per capita income .47 8-38 Intensive guidance part. .65 8-37 Intensive guidance under. .62 8-68 Fertilizer cost .51 8-41 Multi-coop. under. .50 8-70 Labor cost .48 8-45 Radio listening .62 S-46 Radio programs .55 8-39 Pam. planning under. .54 8-43 Source of information .50 8-44 Reading newspaper .45 8-72 Land rent .59 8-66 Food crop prod. cost .54 8-71 Farm tool cost .46 8-83 Durable goods .43 8-36 Election participation .42 47 pattern has the highest total variance compared to the other patterns; variation among all the variables involved in Pattern One is 6.35 percent. Pattern Two: Family Income and Expenses Four variables emerged in this pattern, variables which were related to each other. Those variables were total family income, total consumption expenditures, family dependents and per capita income. Pattern Three: "Primary" Expenditures Variables for consumption expenditures for food, clothing, and education emerged in Pattern Three. These expenditures can be considered the most essential expen- ditures for daily life. Pattern Four: "Secondary? Expenditures Expenditure variables for house construction and medical cost came together with per capita expen- ditures. These kinds of expenditures rank second in importance after the essentials expenditures. Pattern Five: Basic Investment Production investments, especially investments in land property, are considered basic investments for rural people who depend mainly on agricultural production. Pattern Six: Sources of Income In this pattern the only two variables that emerged show the sources of income: Sources from food crop production and nonagricultural income. 48 Pattern Seven: Saviags variables for total money savings and saving as a proportion of income emerged in Pattern Seven with the highest factor loadings in this analysis. Pattern Eight: Agricultural Modernization Variables for intensive guidance system, multi- cooperative understanding, and fertilizer and labor used for food crop production all emerged in the same pattern. These variables can be considered the basis for trans- ferring peasant farming to agricultural modernization. Pattern Nine: Land Tenure Variables for land ownership, land operation, and total production cost intercorrelated with each other and emerged in the same pattern. Pattern Ten: Last Decade Program (in Land Reform) It seems that most of the household respondents still remembered their involvement in the last decade's land reform program as either the land owners or the land receivers. Pattern Eleven: "Luxury" Needs This pattern is peculiar. Variables for durable goods consumption and election participation emerged in the same pattern, though with small pattern loadings. It seems, then, that election participation can be con- sidered a luxury need. 49 Pattern Twelve: Production "Tools" Production costs for food crops, percent pro- duction costs for land rent and farm tools existed in the same pattern. Those variables are very important for agricultural production. Pattern Thirteen: Access to Development Program Several variables that can be considered as communication media emerged in the last pattern. Those were variables for knowledge about recent programs, such as family planning, through the media of radio and news- paper. This final pattern had only a total variance of 5.43 percent. Pattern One was the most common factor most influenced by the shared determinant. The larger the number of the pattern, the less common the factor, which had a smaller total variance percentage. Pattern Thir- teen, then, was the least common factor. Extensive Data, Village Level Extensive data for the village level were col- lected in West Java, among eighteen regencies. Infor- mation regarding village conditions for 1972 was col- lected in 1973 and comprised 300 villages. See Figure 3. Community Variables The variables for the village level are called community variables and these variables are more easily 50 changed compared to the familial variables for the household level. Most of the data concerned the physical conditions of the village, the economic situ- ation, the human variables, public construction and the institutional variables. A complete list of these variables can be found in Table 3. There are 87 dynamic variables that were originally manipulated from 107 raw variables from the village questionnaire sheets. How- ever, not all of those 87 variables were analyzed. Those variables having about the same nature were excluded. Pattern of Community Variables Sixty-six selected and combined variables were analyzed by using factor analysis. From the unrotated matrix, the result found by free factoring with a minimum eigenvalue of 1.0 was twenty-three factors with a cumu- lative total variance of 69.3 percent. By reducing the number of factors to thirteen, there was still a cumulative total variance of 50.7 percent and a minimum eigenvalue of 1.5. The original results of the rotated matrix of the village factor analysis can be found in Appendix E. The main result of this factorization was an aggregation of sixty-six variables in thirteen patterns. If all variables with factor loading lower than .40 are 51 TABLE 3 COMMUNITY VARIABLES OF 300 VILLAGES (1973) L-12 L-13 L-14 L-15 Per capita income, in rupiah Irrigated ricefield, as a percent of total village land Eroded and flooded land, as a percent of total village landi Ricefield intensi§y_index, maximum 200 percent Dryland intensi§y_index, maximum 200 percent Ricefield hectarage with intensive uidance system, as a percent of total village ricefiel land *Land for food crops, as a percent of total village land *Rainfed ricefield hectarage as a proportion of total village dryland, in percent Rainfed ricefield hectarage as a proportion of total village ricefield, in percent Forest and plantation hectarage, as a percent of total village landi Land for dryland farming, as a percent of total village land Average hectarage of cultivated land per household *Productive people as a percent of all adults Adult portion of total population, in percent Portion of adults with elementary education, in percent Portion of adults with more than elementary edu- cation, in percent Portion of adults with vocational education, in percent 52 TABLE 3--Continued L-18 L-l9 L-20 L-21 L-22 L-23 L-24 L-25 L-31 L-32 L-33 Pupil/teacher ratio Pupil/children ratio Farmers using credit, as a percent of all farmers *Children/household, in percent Number of people in the family, in persons *Male/female ratio Population density, in persons per hectare Credit usea/total irrigated land, in rupiah per hectare Production cost of principal crop, in rupiah per hectare *Draft animals used per hectare of ricefield, in horsepower *Draft animals used per hectare of dryland, in horsepower Tractor(s) used per hectare of cultivated land, in’horsepower Outside labor used per hectare of cultivated land in man days Large animal(s) per household *Large animal density, in animals per hectare of total land Village funds for develOpment per household, in rupiah Local contributions for development per household, in rupiah Government subsidies for development per household, in rupiah Credit used on each farm, in rupiah 53 TABLE 3--Continued L-40 L-4l L-42 L-43 L-44 L-49 L-SO L-51 *Asphalted (paved) road density, in kilometers per hectare of total village land *Graveled road density, in kilometers per hectare of total village land *Soft (dirt) road density, in kilometers per hectare of total village land Permanent houses/family, ratio Temporary houses/family, ratio *Motorable transportation tonnage capacity/bulk product sales tonnage, ratio *Nonmotorable transportation tonnage capacity/bulk product sales tonnage, ratio Fertilizer used for food crops, in quintals per hectare Insecticide used for food crops, in liters per hectare High yielding variety seed used, in quintals per hectare ’ New inputs (fertilizers, insecticide, high yielding variety seed) for rice production, in rupiah per hectare New inputs (fertilizers, insecticide, high yielding variety seed) for rice production under intensive guidance system ricefield, in rupiah per hectare *Television/household, ratio *Radio/household, ratio Distance from district city to larger city, in kiIOmeters Distance from district city to regency capital, in kiIometers Number of mutual activities, in events per year 54 TABLE 3--Continued L-54 L-SS L-56 L-57 L-58 L-59 L-60 L-61 L-62 L-63 L-64 Mutual activity participants/household per year Number of cooperatives Cooperative members/household Village staff/household *Village staff/population Households/neighborhood association Village defense member/population *Agriculture products trader/population Value of agriculture products per trader, in rupiah per hectare *Industrial product retailers/pOpulation Value of industrial and agriculture products per merchant, in rupiah per person Food crop sector product, as a percent of gross village product Animal husbandry sector product, as a percent of gross village product Fishery sector product, as a percent of gross viIlage product Forest and plantation sector product, as a percent 3? gross Village product Excavation sector product, as a percent of gross viIIage product Home industry sector product, as a percent of gross village product Transportation sector, as a percent of gross village prOduct Credit sector, as a percent of gross village product 55 TABLE 3--Continued L-81 L-82 L-83 L-84 L-85 L-86 L-87 *Trade activity volume, as a percent of gross village product *Retail activity volume, as a percent of gross village product Trade sector volume, as a percent of gross village product Labor salaries and wages, as a percent of gross viIlage product Public service sector salaries and wages, as a percent of gross village product Public construction, as a percent of gross village product Private construction, as a percent of gross village product Housing sector, as a percent of gross village product *Bicycles/household Draft animals used per hectare of cultivated land, in Horse power Motorable road density, in kilometers per hectare Radio and television set/household Traders and retailers/population Bulk product/transportation capacity, ratio Land productiviay: total gross village product/ total village land, in rupiah per hectare NOTE: L is vil"L”age variable code number *Variables excluded from factor analysis 56 eliminated, forty-five variables remain. If the patterns and variables are constructed in rank order, the final results can be identified as those found in Table 4. By observing the data in Table 4, the village patterns can be named: Pattern One: Development Activity Variables for village staff, mutual activity, local contributions and government subsidies for develOp- ment, and public construction emerged in the same pattern. These variables intercorrelated with each other and indi- cated the development activity within the village. Pattern Two: Agricultural Develgpment Activities of farmers using credit facilities for new inputs, such as fertilizer and insecticides for ricefields under the intensive guidance system program, indicated the development of agriculture. Those variables above correlated with each other. Pattern Three: Develgpment Faciliry Motorable road density is part of the public service sector and eases the provision of new input for the intensive guidance system in the ricefield. This condition intercorrelated with p0pu1ation density and permanent-house family ratio. Pattern Four: Industrial Activiry» Variables for high yielding variety seed, land productivity, and per capita income intercorrelated VILLAGE LEVEL FACTOR ANALYSIS 57 TABLE 4 Variables Factor Numbers 1 12 9 3 11 7 13 10 L-4 L-3 L-17 L-65 L-75 L-87 L-64 L-l L-46 L-7O L-82 L-41 L-31 L-6 L-ZO L-44 L-45 L-47 L-57 L-54 L-34 L-35 L-78 L-14 L-19 L-67 L-S L-71 L-83 L-48 L-77 L-40 L-24 L-Sl L-60 L-69 L-SS L-56 L-22 L-84 L-66 L-18 L-3 L-36 L-25 Ricefield intensity Outside-labor used Adults vocat. educ. Food crop sector Trade sector Land productivity Indust. and agric. pro- ducts Per capita income Seed used Home industry sector Draft animals used for cultivation Temporary house Large animals Ricefield int. guid. syst. Farmers using credit Fertiliser used Insecticide used New input Village staff Mutual act. participant Local contributions Government subsidies Public construction Adult Pupil/children ratio Fishery sector Dryland intensity Transportation sector Motorable road density New inputs for int. guid. syst. Public service sector Permanent house Population density Distance to larger city Village defense Excavation sector COOperatives Cooperative members People in the family Radio and television Animal husbandry sector Pupil/teacher ratio. Eroded and flooded land Credit used on each farm Credit used .87 .86 .66 .83 '5‘ .83 .80 .77 .65 .47 .82 .76 .75 .78 .57 .52 0‘7 .45 .75 .66 .65 .65 .63 .74 .71 .43 .52 .45 .59 .59 .52 .49 .58 .49 .47 .57 .51 .48 .46 .56 .50 .45 .47 .47 58 with each other, and together with industrial sector and industrial and agricultural product merchant ratio variables, indicated the stage of intermediate indus- trial development. Pattern Five: Extensive Farming Variables for large animals, draft animals used for cultivation and temporary-house family ratio were intercorrelated and emerged in the same pattern. This situation indicated the stage of extensive farming. Pattern Six: Trade Activiry Variables for the agricultural sector and the trade sector intercorrelated with each other. Agri- cultural sector preceded trade activity. Pattern Seven: Collective Action Size of family related to radio and television/ household ratio. This also intercorrelated with the cooperative and cooperative member/household ratio. Hence, this pattern was called collective action. Pattern Eight: Intensive Agriculture Adults having vocational education and outside labor used for cultivation intercorrelated with ricefield intensity. This preceded intensive agriculture. Pattern Nine: Dryland Intensity Only two variables emerged in this pattern with factor loadings of more than .40, showing that transpor- tation sector intercorrelated with dryland intensity. 59 Pattern Ten: Farm Credit Credit used on each farm and credit used/irri- gated land ratio correlated with each other and emerged in this pattern with low factor loadings. Pattern Eleven: Public Security Excavation sector, distance to larger city and village defense member/population ratio were intercor- related. Excavation activities, in fact, usually occur away from the cities. Pattern Twelve: Human Progress Variables for adult people, pupil/children ratio and fishery sector emerged in the same pattern. Pattern Thirteen: Degradation Pupil/teacher ratio, animal husbandry sector and eroded-flooded land existed in the same pattern. Animal husbandry sector is usually highly correlated with eroded land. From the rotated matrix it is known that Pattern One was the most common factor most influenced by the shared determinant. Pattern One had a total variance of 4.7 percent and Pattern Thirteen had the least total variance of 3.0 percent. Secondary Data, District Level Secondary data for the district level were col- lected in two regencies, the regency of Cianjur and the 60 regency of Bandung, both in West Java. District data for 1973 were collected in 1974 and comprised forty-four districts. See Figure 4. Environmental Variables Since the variables for the district level existed as "given" variables and had to be accepted by the village as well as by the household level, these district data are called environmental variables. Most of the data concerned soil types, current land use, population size, and transportation system. A complete list of these variables can be found in Table 5. There are thirty-three environmental variables. Pattern of Environmental Variables Thirty-three district level variables were analyzed by using factor analysis. From the unrotated matrix, the result found by free factoring with a minimum eigenvalue of 1.0 was eleven factors with cumulative total variance of 81.3 percent. By reducing the number of factors to five, there was still a cumulative total variance of 57.9 percent and a minimum eigenvalue of 2.2. The original results of the rotated matrix of the district factor analysis can be found in Appendix F. The main result of this factorization was an aggregation of thirty-three variables in five patterns. we?“ a- 61 TABLE 5 ENVIRONMENTAL VARIABLES OF CIANJUR AND BANDUNG REGENCIES (1973) T-l Alluvial, grumusol and gleyhumus soils, as a per- cent of total district land T-2 Latosol, red yellow podsolic, litosol and regosol, as a percent of total district land T-3 Andosol and regosol association soils, as a percent of total district land T-4 Settlements, as a percent of total district land T-5 Ricefield cropped once a year, as a percent of totaI district land T-6 Ricefield cropped twice a year, as a percent of totaI district land T-7 Miscellaneous crop arden hectarage, as a percent of total district land T-8 Dryland farm hectarage, as a percent of total dis- trict land T-9 Forest and shrubs hectarage, as a percent of total district land T-lO Plantations hectarage, as a percent of total dis- trict land T-ll Ran e and low productive land, as a percent of total district land T-12 Swamp and marsh land, as a percent of total dis- trict land T-l3 Eroded land, as a percent of total district land T-l4 Per capita income, in rupiah T-15 Number of children in family, in persons T-16 Children/population, in percent T-l7 Adult/pOpulation, in percent T-18 Farmers/population, in percent 62 TABLE 5--Continued T-l9 T-20 T-21 T-22 T-23 T-26 T-27 T-28 T-29 Businessmen/population, in percent Government staff/population, in percent Village staff/population, in percent Total workers/population, in percent r3 As halted (paved) rgads/total district land, in ?E_ kiIometers per hectare ,' Graveled roads/total district land, in kilometers per hectare Dirt (soft) roads/total district land, in g] kiIometers per Hectare Bicycle owners/total household, in percent Nonmotorized transportation owners/total household, in percent Motorcycle owners/total household, in percent Motorized transportation owners/total household, in percent Ricefield hectarage per ton of transportation capacity Ricefield as a percent of total district land Value of a ricultural products per hectare of total cuItivated Iand, in rupiah Production cost of principle crop per hectare of totaI cultivated land, in rupiah NOTE: T is dis"T"rict variable code number. 63 If all variables with factor loading lower than .40 are eliminated, twenty-eight variables remain. If the patterns and variables are constructed in rank order, the final results can be identified as those found in Table 6. By observing the data in Table 6, the district patterns can be named: Pattern One: Land Use Physical variables such as ricefields, settle- ments, forests and shrubs, plantations, and soil types were intercorrelated in one pattern which reflected the land use classification of the region. Pattern Two: Demographic Demographic variables such as number of adult people, children, farmers, government staff, businessmen, and bicycle owners emerged in the same pattern. There were also two land use variables that showed up in this pattern: swamp and marsh land, and miscellaneous crOp gardens. Pattern Three: Infra Structure Three variables emerged in this pattern: Graveled roads, paved roads, and nonmotorized transportation ownership. These variables are intercorrelated, and the pattern can be called infra structure. 64 TABLE 6 DISTRICT LEVEL FACTOR ANALYSIS Factor Numbers Variables 3 2 1 4 5 T-24 Graveled roads .96 T-23 Asphalted roads .96 T-27 Nonmot. trans. owners .92 T-l7 Adult .90 T-12 Swamp and marsh .85 T-16 Children .84 T-18 Farmers .68 T-20 Government staff .61 T-19 Businessmen .58 T-7 Miscellaneous gardens .49 T-26 Bicycle owners .49 .59 T-3l Ricefield .86 T-l Alluvial soils .83 T-4 Settlements .74 T-9 Forest and shrubs .71 T-5 Ricefield once a year .64 T-3 Andosol .63 .58 T-lO Plantations .58 T-33 Prod. cost principal crOp .79 T-32 Value agric. products .77 T-29 Motorized trans. owner .58 T-30 Transportation capacity .58 T-3 Latosol .57 T-l4 Per cap. income .54 T-8 Dryland farm .74 T-25 Dirt roads .56 T-6 Ricefield twice a year .52 T-l3 Eroded land .46 65 Pattern Four: Economics Variables for production cost of principal crops, value of agricultural products, per capita income and motorized transportation ownership are intercorrelated in this economics pattern. In this pattern there are also two other variables, transportation capacity and soil type. Pattern Five: Technology The variables for dryland farming, dirt roads, ricefield twice a year and eroded land reflected the technological condition of the region. These variables are intercorrelated and emerged in one pattern. All five patterns had a total variance ranging from 14.8 to 7.1 percent. aggragate Data, Multi-Level The data for variables used for the purpose of multi-level analysis were the same data for the house- hold level (1973), different data for the village level (1973), and the same data for the district level (1973). In short, the multi-level analysis used the 1973 data from the Cianjur and Bandung regencies of West Java. Variables were selected from those three levels of previous analysis which had factor loading higher than .55. There were seventy-five combined variables from the household and village as well as the district levels. 66 The village level was used as a data unit; hence, the necessity to average the household data for every village. The district data were duplicated because for every district there were two villages as data units. For the two regencies there were eighty-eight village data units. Combined Variables The variables in the multi-level analysis are called combined variables and were actually extracted from their original levels (lower levels). There are no new variables. A complete list of these variables can be found in Table 7. Pattern of Combined Variables Seventy-five combined variables were analyzed by using factor analysis. From the unrotated matrix, the result found by free factoring with a minimum eigenvalue of 1.0 was twenty-one factors with a cumu- lative total variance of 79.6 percent. By reducing the number of factors to ten there was still a cumulative total variance of 55.5 percent and a minimum eigenvalue of 2.1. The original results of the rotated matrix of the multi-level factor analysis can be found in Appendix G. The main result of this factorization was an aggregation of seventy-five variables in ten patterns. 67 TABLE 7 COMBINED VARIABLES OF CIANJUR AND BANDUNG REGENCIES (1973) 8-4 8-5 8-15 8-22 8-26 8-28 8-29 8-30 8-32 8-33 S-34 8-37 8-38 8-45 8-46 8-47 S-48 8-49 8-61 Land ownership, in hectares Land operation, in hectares Total family income, in rupiah Total consumption expenditures, in rupiah Total production investment expenditures, in rupiah Membership in Boy Scout organization Membership in cosperative organizations Membership in sport organizations Membership in cultural organizations Membership in loan unions Number of organizations to which head of family belongs Number of organizations to which members of family belong Intensive guidance system understanding Intensive guidance system participation Radio listening frequency Kinds of radioprograms listened to Memory of last decade's land reformprogram Involvement in last land reform program Stage of involvement in last land reform program Consumption expenditures as a proportion of income, in percent House construction as a proportion of income, in percent 8-64 8-65 S-68 8-72 S-75 8-76 8-79 L-l4 L-l7 L-l9 L-2O L-30 68 TABLE 7--Continued Money savings as a proportion of income, in percent Per capita espenditures, in rupiah Fertilizer cost as a prOportion of production cost, in percent Land rent as a proportion of production cost, in percent Value of food crop production per hectare, in rupiah Income from nonagricultural sources, in rupiah Food cost as a proportion of consumption expendi- tures, in percent Clothing cost as a prOportion of consumption expenditures, in percent Education cost as a proportion of consumption expenditures, in percent Production investment in land as a prOportion of total investment, in percent Per capita income, in rupiah Ricefield intensity index, maximum 200 percent Dryland intensiry_index, maximum 200 percent Ricefield hectarage with intensive guidance system, as a percent of total village ricefield land Adult portion of total pOpulation, in percent Portion of adults with vocational education, in percent Pupil[children ratio Farmers using credit, as a percent of all farmers Outside labor used per hectare of cultivated land, in man days Large animal(s) per household 69 TABLE 7--Continued L-54 L-55 L-57 L-64 L-82 L-83 L-87 T-l Village funds for development per household, in rupiah Government subsidies for develOpment per household, in rupiah Temporary_houses/family ratio High yielding variety seed used, in quintals per hectare Distance from district city to larger city, in kilometers Mutual activity participants/household per year Number of cogperatives Village staff/household Value of industrial and agriculture products per merchant, in rupidh per person Food crop sector product, as a percent of gross viIIage product Animal husbandry sector product, as a percent of gross village product Public construction, as a percent of gross village product Draft animals used per hectare cultivated land, in’horsepower Motorable road density, in kilometers per hectare Land_productivity, in rupiah per hectare Alluvial, grumusol and gleyhumus soils, as a per- cent of total district land Latosol, red yellow podsolic, litosol and regosol, as a percent of total district land Andosol and regosol association soils, as a percent of total district land 70 TABLE 7--Continued T-4 Settlements, as a percent of total district land T-6 Ricefield cropped once a year, as a percent of total distriCt land T-8 Dryland farm hectarage, as a percent of total dis- trict landi T-9 Forest and shrubs hectarage as a percent of total district land T-lO Plantations hectarage, as a percent of total dis- trict landi T-12 Swam and marsh land, as a percent of total district an T-16 Children/population, in percent T-18 Farmers/population, in percent T-l9 Businessmen/population, in percent T-20 Government staff/pOpulation, in percent T-23 Asphalted (paved) roads/total district land, in kiIOmeters per hectare T-25 Dirt (soft) roads/total district land, in kilometers per hectare T-26 Bicycle owners/total household, in percent T-27 Nonmotorized transportation owners/total household, in percent T-29 Motorized transportation owners/total household, in percent T-30 Ricefield hectarage per ton of transportation capacity NOTE: 8 is the original hou"S"ehold variable code number L is the original vil"L"age variable code number T is the original dis"T"rict variable code number 71 If all variables with factor loading lower than .40 are eliminated, fifty-seven variables remain. If the patterns and variables are constructed in rank order, the final results can be identified as those found in Table 8. By observing the data in Table 8, the multi-level patterns based on a combination of previous pattern names from different levels can be identified in such a way that eight "hypothetical" statements and two extended conceptual variables can be constructed as follows: Pattern One: Pattern Two: Pattern Three: Pattern Four: Pattern Five: Pattern Six: Pattern Seven: PrOper land use and provision of develop- ment facilities lead to dryland intensity and industrial activity. Organizationalaparticipation Extensive farming is determined by the land tenure system, public security and economic conditions. Human progress is based on family income and primary expenditures. Development evolution Demographic composition influences agri- cultural intensity. Industrial activity is caused by agri- cultural development activity that is based on intensive agriculture and modernization. 72 TABLE 8 MULTI-LEVEL FACTOR ANALYSIS Factor Numbers Variables OP 9 6 4 8 2 1 3 5 7 10 S-76 Income from nonagric. (H6) .86 S-75 Value of food crop (H6) -.85 T-3 Andosol (Dl) -.65 T-2 Latosol (D4) .44 -.44 8-32 Loan union (H1) -.40 .55 T-12 Swamp and marsh (D2) .84 T-16 Children/pop. (D2) .79 L-3 Outside labor used (V8) .63 T-18 Farmers (D2) .62 .43 T-19 Businessmen (D2) .60 T-20 Government staff (D2) .51 .41 8-61 Consumption (H3) .83 S-79 Food cost (H3) .81 L-19 Pupil/children ratio (V12) .55 8-80 Clothing cost (H3) .47 8-6 Total family income (H2) -.45 T-23 Asphalted roads (D3) .83 T-27 Nonmot. trans. owners (D3) .82 L-35 Government subsidies (V1) .64 T-29 Motorized trans. owner (D4) .51 8-33 Org. head of family (H1) .82 8-34 Org. members of family (H1) .82 8-26 Boy Scout org. (H1) .81 8-28 Cooperatives org. (H1) .67 8-29 Sport org. (H1) .67 8-30 Cultural org. (H1) .62 T-l Alluvial soils (D1) .78 T-9 Forest and shrubs (D1) -.72 T-4 Settlements (D1) .65 T-6 Ricefield once a year (Dl) .61 T-lo Plantations (D1) -.60 T-26 Bicycle owners (D1) .58 L-S Dryland intensity (V9) -.55 L-83 Motorable road density (V3) .52 L-87 Land productivity (V4) .46 .44 8-4 Land ownership (H9) .77 8-5 Land operation (H9) .77 L-Sl .Distance to larger city (V11) .60 L-82 Draft animals used (V5) .57 L-3l Large animals (V5) .55 T-3O Transportation capacity (D4) .47 L-41 Temporary houses (VS) .75 L-46 Seed used (V4) .75 L-54 Mutual act. participants (V1) .55 8-62 House construction (H4) .46 8-87 Investment in land (H5) .40 8-38 Intensive guidance part. (H8) .62 L-6 Ricefield int. guid. sys. (V2) .61 8-34 Intensive guidance under. (H8) .60 L-4 Ricefield intensity (V8) .57 L-l Per capita income (V4) .45 8-68 Fertiliser cost (H8) .42 L-78 Public construction (V1) .41 T-25 Dirt roads (D5) -.62 L-20 Farmers using credit (V2) -.47 L-65 Food crop sector (V6) -.43 L-64 Indust. and agric. product (V4) .43 NOTE: OP is original pattern number H is household pattern V is village pattern 0 is district pattern 73 Pattern Eight: Infra structure influences economic development activity. Pattern Nine: Sources of income are influenced by economic organization of land use. Pattern Ten: Technology makes possible agricultural development and trade activity and affects the level of industrial activity. After the patterns of combined variables were analyzed, only one pattern remained that retained the original pattern name from its lower level analysis: that is, Pattern Two, Organizational Participation, is, as Pattern One, at the household level. The other nine patterns are an aggregation of variables that originally came from different levels. The dominant previous pattern names are underlined. Pattern One to Pattern Ten of this multi-level analysis each has a total variance range from 7.6 to 3.8 percent. Conclusion By using a limit value of .40 factor loading, the trivial variables were screened out and the out- standing variables retained. Out of the sixty-six familial variables, forty-nine variables (74%) were retained; from the sixty-six community variables, forty- five variables (68%) were retained; and out of the 74 thirty-three environmental variables, twenty-eight variables (84%) were retained. In addition, thirteen patterns of familial variables, thirteen patterns of community variables, and five patterns of environmental variables were found. Those patterns showed the grouping of characteristics of variables which were highly correlated. Factor analysis caused the emerging variation of character- istics between patterns and divided the regularity in the data into its distinct patterns. In the multi-level analysis, seventy-five com- bined variables were screened and were reduced to fifty- seven variables (76%). However, if a calculation is based on the original set of selected variables at the lower level (167 variables), it can be noted that this number was reduced to 35 percent (57 variables). A screening technique to condense the high number of variables was demonstrated without losing the objec- tivity in the analysis processing system. The multi- level analysis yielded a set of hypothetical statements based on the lower level analysis, which meant that the holistic milieu/ecology was conserved. CHAPTER IV THE APPLICATION OF MULTI-LEVEL ANALYSIS Description of BandungsRegency As has been mentioned before, the regency of Bandung is located in West Java, Indonesia, and encircles the municipality of Bandung in which is located the West Java Province administration office, the Bandung Regency office and the Bandung Mayor's office. The Regency of Bandung which will be used here for a pilot observation, with detailed explanations of the region for the appli- cation of multi-level analysis, consists of twenty-seven districts and 284 villages in which are 435,483 households. Complete observation was made at the regency and district level. There are fifty-four village samples (21.8%) and 431 household samples (0.1%) taken. The household samples consisted of 231 farmers, 39 traders, 24 entre- preneurs, 41 laborers, 73 public servants, 8 village staff members, and 15 other professions. See Figure 4. The government level above the regency level is comprised of the residency level (coordination function), the province level and the national level. The lower 75 76 government consists of the sub-regency level (coordi- nation function), the district level, and the village level. Based on the autonomous status, the regency level is the second autonomous government; the first, higher autonomous government is the provincial level, and the lower autonomous government, the third, is assigned to the district level.23 Regency government apparatus con- sists of the legislative body where the people's repre- sentatives work together; the administration office which executes the daily activity of public services; and the judicial body which has the power of law enforcement. The administration office is assisted by the regency planning unit. Physical Characteristics The regency of Bandung is a bowl-shaped region with a radial drainage pattern. Most of the settlements are in the center, scattered in the paddy ricefield region. This paddy ricefield is encircled by dryland farms, and the outer ring is forest and plantation. The whole region is about 311,045 hectares with a varying altitude range from 350 to more than 10,000 meters above sea level, a condition which enables various crops to grow in this region. The t0pography of the area varies, from a gently sloping (30.5%), 23This latter assignment is still being reviewed, an appropriate law will be forthcoming soon. 77 rolling area (33.8%) to mountainous region (35.7%). Annual rainfall averages about 2,250 millimeters. Humidity is about 70 to 78 percent, and the temperature ranges from 3° to 18° Celcius, according to the season and altitude. The mountainous region is covered by Andosol and Regosol soils (35.5%), while in the middle of the region from the western to the eastern part is found Alluvial soil (18.6%), and a mixture of Alluvial with Gley Humus (18.6%) and Grumusol (.l%) soils. In most of the western parts of the region there are Red Yellow Podsolic (16.8%), Latosol (27.1%) and Litosol (1.9%) soils. The average depth of the top-soil is about thirty to sixty centi- meters with various textures ranging from fine to coarse. A southern highway from the country capital (Jakarta) to central Java crosses about fifty-five kilometers of the region. Bandung Regency has 2,805 kilometers of motorable roads, with about 15 percent paved (asphalted) roads, 50 percent graveled (hardened) roads and the rest, soft (dirt) roads. The road density is about .9 kilometer per square kilometer. Demographic Characteristics Bandung Regency has a population of 2,031,209 people with the number of females 1.5 percent higher than the number of males. Out of the total number, 18 percent are infants, 25 percent are school age 78 children (five to fourteen years old), 45 percent are labor force age (fifteen to forty-four years old), and 12 percent are considered as elderly people (fifty-five years old and above). The population density ranges from 612 to 968 persons per square kilometer and several districts have more than 1,000 people per square kilometer. The average household consists of 4.7 persons. The distribution of professions is: 43 percent farmers, 30 percent laborers, 3 percent entrepreneurs, 9 percent traders, 9 percent public servants, 1 percent village staff members, and 5 percent other professions. About 8 percent of the labor force is unemployed. The population growth average is 2.5 percent; in some dis- tricts the average is 3.3 percent which might be caused by local in-migration. In every village there are one to three elementary schools though about 54 percent of the school age children do not benefit from education yet. Technology In an established developing region such as Bandung, the land use pattern can be considered as a reflection of the stage of technology, at least in the agricultural sector. About 27 percent of the region is production and reserve forest, 16 percent is plantations, mostly tea plantations. This amount of land use in these categories is thought to have a very positive 79 mcsocmm Co zocmmom CH mtaesrm .nmamm> Co coflpmooa are . m creams 5.! . L .11. IA.) .638 633....» .6 83.83 on I H . fix. a F P I e e's e I ,V. 2.-.... .-.- 5...... 3:... \W .‘s . .ll . II ..\ ’1, . fluowcg \e 0’. - /~ e c , .1. .2 c. o o -’ ‘4‘ . .\i!\ . :IESLFBE / 5 u\!\ .. Iv .\ ’n‘ O o. 5 o. ..(S‘ \ \ O Tee \ \ - a on ‘60 \ \ \ . .euoa \ l.. .. . _ on .7» O \ l. \\ 8 .\ a \ s. 1 s s I . .. singly \..J.. sit {8.3 \$ . a: .\ a f. N a I 3’. x \8 .\.. 3 _ I a \ m... “figurine ’ I \ \>o./ 3‘ o N“ m - ’- ‘ ' as I \ u ... I l/ \7 \I II on f \ \ on I .’0 :4 A.\ 1! .I \~ u I \I I .O I >% . I. ...-5‘.“ 3 \ .o ~ ...!I—u lu — t J J . .. .890 . 2 .. . . . a. a 2 .V.\I_!€lrx/ ) . .s LI: 88a a green, .I _ 0 8.0.5.3 \ ~ 93 u». 2 .(...l..bs| \ 4.. s/ ..N. ....» r: ... . a c.\\. adgx.\..\ 8 sluts 8 . 7.. .. 0 .v 3 «a /\ .\ \ \. I l./.a.lF\ a a 8883 1.8 s I ..afl’lladt:l:) » l\ . 5a \o)..|lo \o o ‘33“0 I . .. .. ... . - . c.-. . ... . \ . . 9:3: . . . . m E ~se’ . a New I}, J.. \ \..\..l.\ a O a.) s a .. 0. .I 3 :45. - .\. . . . I O . 1‘ l .. \ I. .\ 82.30! H0 . . . 50??le I Y. J x o. 88 ... 5...”. n 1,...) ~.«.\\ \ ..\ O 1», .\ «on. o «’0 .u I! 4‘1! Q. ~- 3- .o\ F) .\o on, o I‘3‘o .‘I w 40‘ )Msfl' us‘ ‘0 - ‘\ g1 ouufl‘u of / Egdounm an I..\...... Q Q a 0‘..n\.b / I. .. his \ ... \.1.. w l./ I..\-. \.. _ ~ .I. | .\ oo‘ ‘0. f o \u 0- ’. s’ \ ‘o \o as \ 08$ I l/ .s. 4/ w. ..5 Fines... «.1 o8.8« . a . 18m ’(K 1...... 5385...... u .u u - .ls. >uzusu¢ ezaezsH swim om. Ammo usmsuum>sa .ooum H8008 uwlm mm. Ammo swam ossaauo one we. Ammv macuuos3o 0:84 vum Na. Ammv sowuuuomo coda mum me. Am>v auaucoo ouch mansuouo: mmnq mm. Av>v >ua>uuosooum coca nmlq mm. Amos coon condonmuc nwna om. Anny uuoczo .ucduu .voasoz buns mm.l Am>v huwncoasw usmnhun mud mo. “Hoe use» 8 ooso uaoawoowm was nh.| Aaov ucowusucsam oaua on.u Aaov season was uuouom mus om. AHQV mucmfioHuuom via mm. “any udwou H8H>nad¢ HIE Hm. Acne muocso .ussuu vouduouoz ante mm. cm. Ame. modem uuwo mwne Hm.u pm. “An. Honovcd mus Nm. Away mono coon mo 09Hs> mhnm mm.: Ammo .oaumssos Bonn oEoosH whim mo. me. Ame. mucosa eachuam w~ua mm. Amov macaw acoscuo>oo owue mm. «o. Am>v can: wanna cowuuso onuq mo. “Nov cosmnoswusm mane Nb. Away uuafihsh mHIa mm. Away cause was ass3m Nana mm. Away .mom\:ouoafinu maua a cu m c a m m moanswun> muonesz Houosh Annmdv wuzmwfim UZDQde m0 MHMNA¢Z< AN>HAIHBADZ ho OZHNHMOBUdh mH mamdfi £33 mv. Ann. .umoss oossowsu o>wmcoucn «mum on. inane maacouusa chasm mqum um. Anamv mssumoum Owosm ovum mm. AH>V sowuosuuucoo Uganda ohnq om. ANHmV anon 036A whom Nm. 3.5 .0560 .uuog mvHDU< hand no. A>>v um>wusuonoou mmnq mv... 2.25 mascara ammun— non...— hv.l AHH>V huwo nomuma ou oossuuwo amid om. mv.| Away .uusm mossvasm o>aucousn onum vm.l Am>v nauaflcd ouudd Anna mo.| Away uuoo HouwaauHOh mwum mv. Av>v 0300:“ sawmso mom and wv. Away cowumEsucoo dance manm aw. loamy shaman coda usoam>ao>cH ovum He.| Aoamv ucs80>ao>ca omsum ovum an. Aoamv asuuoum su0uouocdq hvnm ow. Av>v can: comm ovna av. AH>V nucsmwowuusm .uouassusz vmnq vs. AH>V uuuum ommaah> pm-q an. Am>v amazon husuomsoa anq we. 1~H>e ohuou cuuuaaaoxaamam «and on. AHmv mCOflnd anon anm mv. mm. «any .muo o>wusuomooo ownm mm. A23 umoo cowuuusmvm lem h m m m o." N v H o uuonasz uouomh nuanusuu> Aomscaucoovunma mamas 94 Pattern Six: The sources of income are influenced by physical (land useT’teéhnoIogy as well as the economic con- ditions. Pattern Seven: Access to development program and having proper pro- duction tools will stimulate development activity. Pattern Eight: Intensive agriculture and organization participation are the base for coIlective action. Pattern Nine: Agriculture modernization and extensive farming are related to public secufity. Pattern Ten: Development activity through extensive farming will lead to industrial activity. Higher Order Factor Analysis As was mentioned in the earlier chapter, the oblique rotation technique was utilized in factorization of variables to enable production of a factor correlation matrix. In oblique rotation the factors that emerged were correlated to each other, and the factor correlation matrix can be found in Table 16. An examination of the factor correlation matrix shows that there was some correlation between factors; however, in general, the correlation is small. The relationships between factor 95 00. n0.| n0.l 50. «0. A0. 00.H 00.! 00.A v0.| v0. no. «0. 00.A n0. «0.! m0. v0. N0.I 00.d 00.! 00.0 v0.l #0.! «0.: 00.H 00.I A0.I 00.! 00. «0. n0. m0.l 00.~ H0.I n0.l 50. 00. AH. n0. m0.l ~0. 00.H NA.I m0. Ad. H0.I v0. 00. no.l no. 00. 00.H aua>auon «swan-nose 0» once «Au: modsuuu o>wocovxo sosoasu huu>auon uslsnoao>oo auwusoon uuansm ou oousaou can usdaucu o>uusouxc vac sodasnasuOGOB ousuasowuo< sowuuc c>wuuoH~oo HOu cash on» on. co«us&«o«uusa acqusnacduuo can ousuaso«uos o>quc0ucn huu>uuos vac! 1.338 313.3- 35 .83 souuusooum nomoum uca>sn and Hammond usasmoao>oo ou I-ooofl occauwvsoo calosooo on Ads) on 0.: 65:23.3." an 8.06335 «HI 3006.... NO nah—won 3H. >u4>wuos Huwuuusvaa and noncomxo\06005 .3 «58.». noon. Iguana asuuoun cocoon anon usoaoosd>vn «coaquos 1:00» oucooun nucofluuo>sa canon and sou-an cannon coda ousuanuquuc o>qusdusa oocosausa dad) couuwnomsou oagmsuuosco can .n-ouooua sass: uou haw>avoc usua00ao>ou ms xoun nousuwocoaxo aunflaum as. caduamuoauusm soausnwcdouo auw>wuuc acduuusvsa opossum Add: huqawnduwsa owcmsumoaou has: oosuasou .uoauwaaosu usuamoHo>ov no coauu> noun and cusuosuu- sauna 65633 .23 on... also: .0H .N 0H m nonsdz uooosm nuouodh uo cowumwuo-oo anbadv NDIHURK 025924” ho mHuaR424 Afl>MAIHBADI ~NH¢H¢I BOHB‘AEKKOU IOFU‘h 0H fldfldfi 96 one and factor five and factor two and factor six both have positive correlations. Factor three and factor eight and also factor four and factor nine have negative correlations. Lastly, factor seven and factor ten have a positive correlation. By using the above correlation matrix, a second order of factorization can be further analyzed. Running the factorization with a minimum eigenvalue of 1.0 resulted in five factors with a cumulative total variance of 56.3 percent. The original results of the second order factorization can be found in Appendix I. The main result of this second order factorization was an aggregation of ten patterns to five patterns of the second order. Patterns with factor loadings lower than .50 were eliminated. The patterns and the second order factors were constructed in rank order, and the result which was achieved can be found on Table 17. An aggre- gation of hypothetical statements can be structured, based on the first order pattern hypothetical statements. The aggregate hypothetical statements are as follows: First Aggregate Hypothetical Statement: Last decade program (in land reform) which influences family income also influences the "conditio sine qua non" of industrial activity. 97 TABLE 17 SECOND ORDER FACTORIZATION OF MULTI-LEVEL ANALYSIS OF BANDUNG REGENCY (1973) Factor Numbers Description of Patterns 5 3 4 1 2 10. The demographic compo- sition will influence intensive agriculture .81 Intensive agriculture and organizational partici- pation are the base for collective action. -.56 Agriculture modernization and extensive farming are related to public security -.79 Land tenure system and basic investments precede techno- logical advancement .57 Organization participation and primary expenditures back up the development activity for human pro- gress —,73 The sources of income are influenced by physical and use as well as the economic conditions -.72 Last decade program influ- ences family income/expense and industrial activity .72 PrOper land use, building infra structure and provi- sion of development facil- ities, combined with demo— graphic suitability, will precede industrial activity .60 Development activity through extensive farming will lead to industrial activity .68 Access to development program and having proper production tools will stimulate develop- ment activity .61 98 Second Aggregate Hypothetical Statement: Access to development program and having proper production tools will stimulate development activity. If the develOpment leads to industrial activity, then some extensive farming may result because labor will go to industrial jobs and the hectarage of farm operation will also increase. Third Aggregate Hypothetical Statement: Proper land tenure system and wise basic investments will stimulate technological advancement. However, if, conversely, there is improper land tenure system and unwise basic investment policies which do not stimulate technological advancement, then this situation will cause agricultural development of the extensive farming type which will require con— siderable public security. Fourth Aggregate Hypothetical Statement: The physical and technological as well as the economic conditions determine the sources of income, and together with organizational participation and wise primary expenditures, these elements precede development activity for human progress. Fifth Aggregate Hypothetical Statement: Imbalance in local demographic composition will negatively influence intensive agriculture. If this occurs, it will also influence the intensity of agricultural production and organization par- ticipation, both of which are the base for col- lective action. Since the second order factorization of this multi-level analysis used oblique rotation techniques, a second order of factor correlation matrix was produced as in Table 18. Based on that 5 x 5 factor correlation matrix, a third order factorization was further analyzed. It resulted in two factors having a minimum eigenvalue Description of Factors 99 TABLE 18 SECOND ORDER FACTOR CORRELATION MATRIX, MULTI-LEVEL ANALYSIS OF BANDUNG REGENCY (4:973) Factor Number 1 2 3 4 1. Last decade program (in land reform) which influences family in- come also influences the ”conditio sine qua non" of industrial activity 1.000 Access to develop- ment program and having proper pro- duction tools will stimulate develop- ment activity: if development leads to industrial activity, then intensive farming will be needed. .006 1.000 Improper land tenure system and unwise basic invest- ment policy will not stimulate techno- logical advancement, but will cause agri- cultural develOpment of the extensive farm- ing type which will require considerable public security. -.072 .029 1.000 The physical and technological as well as the economic con- ditions determine the sources of income, and together with organi- zational participation and wise primary expen- ditures, these elements precede development activity for human pqurOlI. -0101 0032 0095 10000 Imbalance in local demographic compo- sition will negatively influence intensive agriculture. This situation will also influence the intensity of agricultural produc- tion and organization participation, both of which are the base for collective action. .029 .045 -.047 -.004 1.000 100 of 1.05 and a cumulative total variance of 44.9 percent. The results of the third order factorization can be found in Table 19. The chief result of this third order factorization was an aggregation of five patterns to two patterns of third order. Patterns with factor loading lower than .50 were eliminated. An aggregation of hypothetical statements can be structured based on the second order pattern hypothetical statements as follows: Final Aggregate Hypothetical Statement One: It was understood that physical and technological as well as economic conditions determine the sources of income, and together with organizational participation and wise primary expenditures, these elements will precede development activity for human progress. It seems that the unjustified last decade program in land reform which influenced the family income also influences the "conditio sine qua non" of the industrial activity. Further, it should be noted that the imprOper land tenure system and unwise basic investment policy will not stimu- late technological advancement, but will cause agri- cultural development of the extensive farming type which will require considerable public security. Final Aggregate Hypothetical Statement Two: Imbalance in local demographic composition will negatively influence intensive agriculture. This situation will also influence the intensity of agricultural production and organization par- ticipation, both of which are the base for col- lective action. Further, access to development programs and having proper production tools will stimulate development activity. Later, if development leads to industrial activity, extensive farming will be needed. 101 TABLE 19 THIRD ORDER FACTORIZATION, MULTI-LEVEL ANALYSIS OF BANDUNG REGENCY __4_1973) Description of Patterns Factor Number Last decade program in land reform which influences the family income also influ- ences the "conditio sine qua non" of industrial activity. Access to development program and having proper production tools will stimulate development activity; if development leads to industrial activity, then extensive farming will be needed. Improper land tenure system and unwise investment policy will not stimulate technological advancement, but will cause agricultural development of the extensive farming type which will require considerable public security. The physical and technological as well as the economic conditions determine the sources of income, and together with organizational participation and wise primary expenditures, these ele- ments precede development activity for human progress. Imbalance in local demographic compo- sition will negatively influence intensive agriculture. This situation will also influence the intensity of agricultural production and organi- zation participation both of which are the base for collective action. Sum of squares -.59 .12 .17 .73 .61 -.04 .65 .15 -.18 .69 1.20 1.05 102 Strategic Pattern for Development The result from the first order factorization process of multi-level analysis is the aggregation of variables into ten factors/patterns. A question is then raised about the relation between the variable of family income from the original variables and the ten patterns. It should be remembered that the use of the oblique rotation method in factor analysis resulted in patterns that are related to each other, since oblique is non- orthogonal. To make possible an analysis of the patterns together with the original variables, a factor score is needed which can determine the score for a case on each pattern/factor. The factor scores are derived in the following way. Each variable is weighed proportionally to its involvement in a factor; the more involved a variable, the higher the weight. Variables not at all related to a factor would be weighted near zero. To determine the score for a case on a factor, then, the case's data on each variable is multiplied by the factor weight for that variable. The sum of these weight-times-data- products for all the variables yields the factor score. This weighted summation will give cases high (or low) scores if their values are high (or low) on the variables involved with a factor. The factor score could be used for further analysis such as classification of cases (to arrange the cases according to their involvement in every 27R. J. Rummel, Applied Factor Analysis (Evanston: Northwestern University Press, 1970), P. 150. 103 pattern) and multiple regression. In this analysis the patterns are examined by using linear multiple regression to find the relationship to family income. In this way the patterns that have the most influence on income gen- eration can be found. To facilitate the analysis, data for ten patterns which explain cases (that is the factor score) were aggregated to the data of seventy-five variables of cases (which is the original data). These new data set were run through multiple regression analysis wherein total family income was the dependent variable and the ten patterns were independent variables. The result shows that the computed F value indicated that only one pattern, Pattern Three, was significant at the a = .1 level. The equation is: Y = 2.27 - 5037.81 Xl The equation gave only a multiple correlation (R) of .34, and the coefficient of determination (R2) was .11. The significance of the equation was .01, and the other nine patterns were deleted since their significance levels were .14 and higher. It seems that the patterns could not strongly explain total family income. Strategic Variables Which Influence Famin_Income The main result of factorization in each level in the previous chapter is the identification of 104 outstanding variables in each level. This technique enables the construction of a data set of multi-level variables that is limited only to those variables that remain after screening by factor analysis. The sources of data set that will be used to find out those strategic variables which influence to family income are: (l) Factors/patterns that have high factor loading of family income variable, that is Pattern One (.42), Pattern Five (.33), and Pattern Three (.28); (2) Simple correlation matrix of the original seventy- five variables (75 x 75 matrix) that has been formulated before the factor analysis of Bandung Regency was run. The comparison of the four data sources to get the best equation model is as found in Table 20. Since the equation from the simple correlation matrix yielded a regression model with the highest F level (13.82), highest coefficient of determination (.59) and lowest coefficient of variability, in this disser- tation the equation for the regression model of Bandung Regency will use those variables that originally resulted from the simple correlation matrix data. The correlation between the variable of total family income 105 m me.m~ am. he. ooo.o ~m.ma xfluumz coaumamuuou mansam m ee.mm am. am. Hoo.o oe.m m>nm cuouumm v mm.am mm. om. ooo.o mm.m mmune cumuumm m mm.om me. we. ooo.o Hm.e mco cumuumm A my soap coaumswm om>ao>sH muflaflo8flum> Imcaaumumo Amy on» How coaumsvm mmHQMHHm> «0 .MO coaumamuuou Hm>mq may now moonsom no umngsz unmaonummoo nameoemmmoo meanness mocmoemaamem m ewusmsoo Amemav wozmomm ozoozmm mo ammo: onmmmmomm mom mmomoom 0N mqmdfi 106 and other variables having a correlation coefficient (r) of more than .20 can be seen in Table 21. Based on the above correlations found in Table 21, nineteen variables were run in linear multiple regression analysis. The dependent variable was total family income and eighteen variables were independent variables. The variable of total consumption expenditures was excluded because it has a very high correlation to total family income; if the variable of total consumption expenditures had been brought to the equation, it was understood that it would bring an unbeneficial effect to the total result. Hence, this variable's exclusion. The cor- relation matrix of the nineteen variables that had been run in the regression analysis can be found in Table 22. By using the stepwise method, and with the sig- nificance level at the a = .10 (F 1,54 Table = 2.80) only five variables remained in the equation. The equation is: Y = -72921.7 - 1475.5 Xl + 255.1 X2 + 29014.1 X + 13736.1 X 3 4 + 19787475.0 X 5 where: Y = family income X1 = Food cost as proportion of consumption expenditures 107 TABLE 21 SIMPLE CORRELATION MATRIX BETWEEN VARIABLE OF TOTAL FAMILY INCOME AND OTHER VARIABLES (BANDUNG REGENCY, 1973) Correlation Name of variables Coefficient (r) Total famil ncome (1) to: Total consumption expenditures (20). .88 Per capita expenditures (7) . . . .36 Land productivity (13) . . . . . .35 Adult portion of total population s s s s s s s s s s o -033 Membership in Boy Scout organi- zation (2) . . . . . . . . . .33 Plantations hectarage (17). . . . -.32 Membership in cooperative organi- zations (3). . . . . . . . . .31 Settlements (15) . . . . . . . .31 Membership in loan unions (5). . . .31 Asphalted (paved) roads per hectare (18) O O O O O O O O O O .29 Distance from district city to Iarger city (12) . . . . . . . .27 Food Cost as a proportion of con- sumption expenditures (8) . . . . -.27 Education Cost as a proportion of consumption expenditures (9) . . . .25 Nonmotorized transportation owners/ population (197* . . . . . . . .25 Ricefield cropped once a year (16) . .23 Alluvial, grumusol and gleyhumus 80118 (14) s s s s s e e s e 022 Per capita income (10) . . . . . .21 Memory of last decade's land reform program (6). . . . . . . . . .21 Membership in sport organization (4) O O O O O O O O O I O .20 108 00.0 00. mm.- 00. 00. 00. «0. 00.- 00.- 00. 00. 00. 00. 00. 00.- no. 00. 00. mu. uuonzo 00004000000400 .00 00.0 00.- 00. 00. 00. 00. n~.- m0.- «0. 00.- 00. «0. 00. 00.- 00.- «0. 00.- 00. annex .00 00.0 00.- e0.- 00.- 00.- 00. 00.- 00. s0.- 00. 00. 00. 00.- 00.- 00.- m0.- 00.- 00000000400 .00 00.0 00. 00. 00. 00.- 00.- 00.- 00. n0.- 00. m0.- 00. no. 00. «a. 00. 00000000: .00 00.0 mm. 00. m~.- 00.- 00. 00. 00.- 00. 00.- 00. 00.- 00. 00. 00. nunoao0uuom .00 00.0 00. 00.- 00.- 00. 00. 00. 00. 00.- 00. 00. 00. 00. an. 000om .00 00.0 00.- m0.- an. 00. 00.- 00. 00. 00.- «0.- 00. 00. mm. mu0>0uoseoua .00 00.0 00.- 00.- 00. no.- «0. 00. 00. 00.- n~.- «0. an. nouns-00 .00 00.0 00. 00.- n0. .0. 00.- 00.- 00.- 00.- 0n.- nn.- 00:00 .00 00.0 00. 00.- 00. nu. no. 00. 00. m0.- 00. osooc0 000000 nos .00 00.0 00. 00.- 00. 00. en. 00. 00. 00. uuoo 0o0uuoaem .0 00.0 00.- 00. 00. mm. 00. #0. 00.- noon soon .0 00.0 00. 0a. 00. 00. 00. on. nausu0ocoaxo «00000 ans .0 00.0 00. an. 00. 00.- 00. 5400000 0.000006 0.40 .0 00.0 00. 00. 0m. 0m. co0an-cuoa .0 00.0 00. an. 00. 00000 .0 00.0 0m. 0m. o>0uauomooo .n - 00.0 mm. unoom mom .a 00.0 usauc0 000540 00000 .0 00 00 00 00 m0 .0 n0 «0 00 00 0 0 0 0 m 0 n a 0 .o0nc0un> no can: Shady Nuzmomfl 020..ng ho mHqudzd AERIHBADZ smademdg zmmauZHz ho x2842 ZOHEEEOU 82% BUDDONm NN Ban. 109 X2 = Distance from district city to larger city X3 = Membership in cooperative organizations X4 = Education cost as proportion of consumption expenditures X = Asphalted (paved) roads density The multiple correlation (R) is .768 and the normalized/standardized coefficient of determination (R2) for this equation is .590. The standard deviation for the equation is 59832.461 with the significance level for the total equation being 0.000 (F computed = 13.821). The coefficient of variability is 25.790 percent. The intercept (constant) of the equation is -72921.7. The beta coefficient and other facts of data in the equation can be found in Table 23. By analyzing the data in Table 23, it can be seen that the standard error is much smaller in magni- tude than the regression coefficient; thus, the sign of the regression coefficient can be interpreted with some confidence. In Table 24, the contribution of the variables to the coefficient of determination (R2) and to the total significance level can be found. The importance of the independent variables to total family income is indicated by their respective regression coefficients. There are four variables that positively related and one that negatively related to 110 00. 00. 00.50 om.mhmao>v oo.mneemhm0 hyamcmo momou omuamnmmd .m m0. mm. mo.m em.0~vv 00.0mem0 monsuwocmmxm c000 IQESmGOU mo cowuuomonm m mm umoo coaumosom .0 mv.0 00. mm.50 vo.vmmo 00.000mm mGOHuMNflsmmuo m>0u Imnmmooo s0 magnumnfioz .m 00. 00. 00.00 00.00 00.000 0000 ummum0 on 0000 uofiuumflo Scum museumao .m mm.- mm.- 00.m~ 00.000 00.0000- amusu0ecomxm coaumESmcoo mo s00uuom noun 8 mm umoo ooom .0 huwoflummHm omuflamauoz 000800 .mmoo .mwm . Imacmfim mo 0000M umwwmwwwmmm moaomdum> mo mfimz ucmfiowmmmou ovum m oumocmum . Amemav uozmumm uzsazsm ho mHmNA¢Z¢ AN>QAIHBADS ZH momma QMddZde mBH 924 BZWHUHmhmOU ZOHmmmmUmm mmB mm mqm¢fi 111 000. 000. 0mm.m0 mom. 0m0. omm. mob. momou omu0mnmm¢ .m 000. 000. omm.m mvm. who. mmq. ~00. mousu I0©cmmx0 s00umadm 1:00 mo c00uuomoum m mm umoo 000umosom .v 000. ooo. onm.m m0m. mmm. mwm. vow. mc00u Isu0smmho m>00mummo loo :0 m0nmumnsmz .m 000. 000. 000.0 000. 000. 000. 000. 0000 “80000 on 0006 00000 Im0o Eon“ mosmum0o .N 000. 000. 000.0 000.- 000. 000. 000. mwusu0ecmmxo 0000 Imfismcou mo s00uuom I000 H mm umoo ooom .0 08909 0s0uumm m 008 c00um0 mmsmwo mumsvm m 0o>mq lum>o wwwflmm on“: m m 00000002 mo0h~008> Mo 0882 mocm00m0cm0m . 0m>m0IHBQDZ ZH Ammv ZOHBdZHZMMan m0 BZMHUHhmmOU Mme OB mmqdeMd> ho ZOHBDmHMBZOU mmfi “00000 uuzmomm ozsozsm mo 00000024 0N m0m¢9 112 the dependent variable. For a better understanding of these important variables, each will be described in the following section. Food Cost as a Proportion of Consumption Expenditures This variable is considered a familial variable and was originally extracted from the household level. The average food cost as a proportion of consumption expenditures is 83.2 percent; however, it actually varies according to the profession of the head of the family, ranging from 79.5 to 88.0 percent. This variable has a negative regression coefficient which is to be expected since the higher the income, the lower the food as a prOportion of consumption expen- ditures. The lower the income, the higher the per- centage food as a proportion of consumption expenditures. Distance to Larger City This variable is considered a community variable and was originally collected from the village level, although it can be also considered an environmental variable. The average distance of the village samples to the larger city is 95.7 kilometers. This variable has a positive regression coefficient to the dependent variable, that is, the further the location of the village, the higher the income of the family. This 113 also means that the rural regions in Bandung Regency still have the potential of absorbing employment oppor— tunity, compared to the urban regions. Membership in Cooperative Organization This variable is one of the familial variables that originally was collected from the household level. Out of the 430 household samples, there were 216 villages in which the development stage was above average and the rest were villages in which the develop- ment stage was below average. The response distribution of the household samples in both levels reveals that there is no significant difference in cooperative membership between the villages of above average and the villages of below average development stage. There were 7.2 percent of the household samples which said that both the head and the members of the family join the cooperative organizations. About 13.8 percent of the samples mentioned that only the head of the family joins the cooperative organization, and 1.4 percent mentioned that only the members of the family join the cooperative organization. In other words, there were 77.6 percent who admitted that neither the head nor the members of the family join the cooperative organi— zations, or an average of 22.4 percent of the household population who join the cooperative organizations. 114 Education Cost as a Proportion of Consumption Expenditures This fourth variable, which has a positive cor- relation with the dependent variable, is also a familial variable that originally was collected from the household level. The education cost as a proportion of consumption expenditures ranges from .3 to 4.7 percent or represents, on the average, 2.5 percent of the total consumption expenditures. As with the food expenditures, the edu- cation expenditures also depend on the profession of the head family. The education expenditures variable has a positive regression coefficient which means that the higher the income, the higher the education expenditures, and vice versa. Asphalted Roads This variable concerns the length of asphalted/ paved roads within one hectare. The average is 0.0019 kilometer per hectare, or 1.9 meters per 10,000 square meters of land. The standard deviation is 1.8 meters. Paved roads make transportation easier and raise the mobility of the rural peOple so that the village pro- ducts can reach the consumer easier and faster. An increase in the length of paved roads in the region will also enable greater accessability of the product to the market which will, in turn, cause an increase in the income of the farmers in rural regions. 115 By using the intercept and the regression coefficient and by applying the regression equation, total family income in other villages as well as in the same villages, but in different time series, can be estimated. The average difference in estimated value and actual value of total family income is 22.3 percent. Those values of the fifty-four village samples in Bandung Regency can be found in Appendix J. The variables that are significant in the linear multiple regression were tested in polynomial multiple regression, and the equation is as follows: y = a + b x + b x2 + b x3 + b x + b x2 + b x3 + b x 1 1 2 1 3 1 4 2 5 2 6 2 7 3 + b x2 + b x3 + b x + b x2 + b x3 + b x 8 3 9 3 10 4 11 4 12 4 13 5 2 3 + b14x5 + b15x5 Y, X1, X2, X3, X4, and X is the same as in 5 the linear equation. By running the equation through the polynomial multiple regression analysis, using the significance level of a = .1, only five out of the fifteen variables still exist in the equation. The other ten variables were deleted due to the fact that the significance level of those variables is too large, ranging from .225 to .890. The contribution of those five variables to the coefficient of determination (R2) can be seen in Table 25. ooo. ooo. ooo. 116 ooo. oHo. 00009 00>0q moo. Hoo. ooo. Noo. oHo. 0000000 00000 -000am0m mmo.m0 ¢o.¢0 mmn.m0 mvm.m hmo.h m HHMH$>O mvN. hmm.l 0mm. mam. hvm. 00000 1000000 000E0m one. 000. «m0. «ma. oNH. 000000 00msvm m 0mm. «em. vmv. mum. 0N0. 00msqm m mob. mmn. mmm. NNm. hvm. m 00000022 000500oc0mx0 00000800000 00 0000000000 0 00 00000 0000mosom 0000000000 1x0 00000800000 00 0000000000 0 mm 0000 ooom >00ms0o 0000 000003000 00 00000m 0000000000000 0>00000mooo 00 m0£m00nfi0z >000 000000 00 >000 00000000 00 00000000 00 0000 .0 m0000000> mo 0802 quom NSF ZH Ava ZOHBdZHZMMBMQ m0 BZmHUHhmmOU HEB OB mmadem4> ho ZOHBDmHMBZOU mma AMNmHV Nuzmomm UZDQZ4Q m0 mHmwfldzd AM>MAIHBADZ .ZOHBdDOH AdHZOz mm wqmda 117 The polynomial equation has a multiple cor- relation (R) of .788, and the number of the standardized coefficient of determination (R2) for the regression is .621. This means that the polynomial regression equation raises the coefficient of determination by .031. The standard deviation for the equation is 57575000, with a significance level for the whole equation of 0.000 (F computed = 15.694). The coefficient of variability is 24.766 percent and the intercept of the equation is -108166820. The regression coefficient, the beta coefficient and other data in the equation can be found in Table 26. Conclusion The technique of multi-level analysis, using real life data from Bandung Regency, was demonstrated in this chapter. The variables identified in the previous chapter were used to construct hypothetical statements. In the first order of factorization, ten hypothetical statements were found. By further analysis with second order factorization, five aggregated hypothetical statements were constructed. And later, a third order factorization produced final aggregations of those five aggregated hypothetical statements. These statements can be used to evaluate the general condition and situation of the Bandung regency. 118 0000000000 00. mm. hv.| hv.l mo. N0. 00.0 mv. vo. 0v. 00000000000 0000 0000008002 0m.m 0m.0m mm.0m 0N.m0 mw.m0 000000m0cm0m m mm.vmmv mm.0m~ mn.ovvmhm mm.0mmm ommmmwhm0 .0000 .000 mo 0o00m o0mocm0m mn.mmmm0 0000000000x0 00000850000 00 0000000000 0 00 0000 00000000m mn.v0m0| 0005000000x0 00000850000 00 0000000000 0 00 0000 0000 om.mmmmmmm 0000000 0000 000003000 00 000:0m m0.mmmmm 0000000000000 0>000000000 :0 0000000802 ooonmmmmm 0000 000000 00 0000 00000000 00 00000000 00 0000 00000000000 0000000000 m0000000> mo 0802 Amnm0v Wozmwmm UZDQde ho mHmNA¢Z¢ 0m>m0lHBADZ ~ZOHB 500 meters asl. 144 Since the majority of land use had been designated according to the real condition of the village samples, the type breakdown for the 300 villages is shown in the following table. TABLE C-Z THE TYPE AND NUMBER OF VILLAGE SAMPLES: EXPECTED AND ACTUAL Type Number of Stratum Description Total 1 2 3 4 5 6 7 8 9 Distribution of dis- trict types 111 32 37 25 48 32 15 28 33 361 Scattering of dis- trict types (where village samples are located) 30 15 15 15 16 15 12 17 15 150 Expected village samples 60 30 30 30 32 30 24 34 30 300 Village sample types according to the field work 67 46 44 29 27 28 18 22 19 300 Percentage differ- ence from expected samples +128 +538 +478 -38 -l68 -78 -258 -358 -378 M 268 District samples of district types 308 728 598 568 278 448 608 398 278 M 468 By examining the above table, it can be seen that the range of village samples as a result of the field work, is 12 to 53 percent higher and 3 to 37 percent lower than the expected village types, or an average deviation of 26 percent. If the two village samples are considered to represent each district (two village samples, above and below average), the district samples range from 27 to 72 percent, or an average of 46 percent of district types. APPENDIX D ORIGINAL RESULTS OF HOUSEHOLD LEVEL FACTOR ANALYSIS so. so. ss. mm. oo. oo.: so.- oo.- oo. os.: so. so. so. sausa>so>cs omauw ovum ss.- so.- so.- o. oo.- so. so. oo. oo.- so. so. oo.- so. euosmu cans ucoao>so>cs ooum pm. ss. ss. oo.- oo. vs. oo. ss.- so. so.. so. oo. oo. gaseous suosou onus scum mm. oo.- so. so. ss. oo. so. so. so.- so. oo.- os. oo. nauseous osoom ovum mm. oo.- oo.- os. ss.- so. oo.- so.- so.- so. so.- so.. os. ocscouoss asses mono ms. so.- so.- so.- ss.- oo.- os. ss.- so.- oo. so. ss.- ss. summonses mascomm eons om. ss.- ss.: so. so.- mm. os.- oo. oo.- so.- so.. ss.- so. cosumsuoucs «0 mousom sv-m HM. os. ss. ss.- ss.- bk. so.- so. so.u so.. so. so.. ss. .uuua .moooususaz soum oh. so.- so.- oo.- so.- om. so. so.. so.. so.- oo. so.u ss. .uwucs .aooo-sussz scum Hs. so.- so. so. ss.- so.- os. ss.- so.. so. oo. oo.- ss. .uuaa masseuse .sus osum om. so. ss.- oo.- oo.- bh. so. so. oo.- oo. oo. so.- so. .uoocs masseuse .sus ssnm so.- os.- ss. oo. so.- .mm. oo.- oo.- oo.- so. so. ss.- so. .uuua .osso o>sucoucs osum ss. so. ss.- oo.- so. so. so. oo.- so. so. so. so. so. .uoncs .osso o>sunmucs ssum oo. oo. so. ss. ss.- so. ss.- so.- oo.- so. so. ss.- so. cossmasosuuua cosuuusm osum so.. so.- ss. oo. so. so. so. so. so.- so. so.u so. Mb.- nosuooso oocouous msnm so.- so. oo.- so.- so. ss. oo. so. so. so.- so. os.- we. assess so eschews .muo .sum oo. so.- oo.- oo. so.- oo. so. so. so.- oo. oo.- ss. Mb. assess so econ .muo ss-m so. so. oo.- so. so.- oo.- os. ss.- oo.- so. so. ss.- mm. nose: sacs ss-m ss.- oo.- oo.- so.- so. so. os. so. ss.- so. oo.- oo. ss. .oso sacosnuosous ssum so. oo.- so.- so.- so. ss.- oo.- oo.- so. so. so. so. mm. .muo saususso os-m so. so.. oo. oo.- so.: oo.- oo.- oo.: so. so.: oo. so.: mm. .ouo uuoam osnm ss. so. so. oo.- os. oo. so. so.- so.- oo.- so.- ss.- mm. .mso o>suouoaooo osum so. os.n os. so.- so. so. oo.- so. so.- so.. so.. as. ms. .ouo oocououussom ssum oo. ss. oo. so. so.u .mb. oo.- oo. so. so.- so. ss.- so. .ouo uaoum mom os-m ss.- oo.- os.- so. so. oo. oo. ss. ss. Mb. oo.- bm. es. .mso uaosossum ssum oo. so. so. so. so.- so. mm. so. os. ss. so.u ss.- so. nouausocoaxo sauce «sum oo.- oo.- oo.- so.- so. mo.- so. so. o.: oo. so.u oo.- so.- mocs>au sagas sauce ssum so.- so.- so. oo. oo. oo. so.. so.- so. we. so. oo.- mo. ucmaumu>cs .voua sauce ssnm so. oo. oo.- so.- oo.- so. so.- so. oo.- so. os.- mb.n so.n cosoosuumcoo canon sauce usum so. so.- ss. so.- os. so. so. so. so. so. so. .. ss. cosuaaancoo sauce msum so. so.- os. oo.- MH. so. oo. so.- so.- so. ss.- ss.- oo. 0585 assess sauce onm so. so. oo.- oo.- on. so. so. so. oo. oo.- so.u ss.- so.u cosuuuomo cans sum so. ss. oo.- so.- oo. so. so. oo. so. oo.- so.u ss.- oo.- asnuuoc3o onus o-m so. so. ss.- so. oo. oo. so. so.. so. ss.- ss. oo.- oo. oucoocoamo ssseus s-m ss ss ss os s o s o s o s s s ukwgz HOanfih moanesum> mHmwddzd mOBU¢m am>wq anmmmaom mo mBADmmm AdzHUHmO HID mdmdfi Q NHQZMhNfi Ju45 lu46 os. oo. oo. oo.- oo.- oo.- oo. so. so. so.- so.. so. so. .suumsa as unmasoo>cs ssuo so. os. os. oo.- oo.- ss.- so. os.- oo.- oo. so.: so. so.- mocsosssn as scoeusu>cs osuo oo. so.- oo.- so. oo.- oo.- oo.- so. so. so. oo.: so.. so. mosesna> as possumo>cs souo os.n so. ss. ss.- oo. ss. so.n os. bb.n so.n oo.- oo.: oo. osoou guns as unusumo>cs souo oo.- so.. oo. oo.- so. so. so. so.- ss. so. so. oo. so. ones as unusuom>cs soum ss.- os.- so. ss.- oo.- so.- so. oo.- oo.- so.u ss. ss.- oo. ooocoaxu sense oouo so. os. ss. so.- so.- so. ss. oo.- os. so.u os. os.u oo.- masocuam sscosoousomm mono so.- oo.- Mb. so.u os. oo.- oo.- os.- oo.- os. oo. oo. ss. noose soum so.- oo.- so. ss.- so. ss.- so.. so.. ss. so.- .bb.u so. oo. ooooo assuage souo so. so. so. oo. oo. so.- oo.- so. so. Mb.u so. ss.: so. soon cosuuoaom souo oo. oo.- so.. so. oo. oo.- so. so.: oo.- so. bl. ss. so. sooo seasons souo oo.- ss. ss. so.- so. oo.- oo. oo.- so. so. Nb. so. so. soon massuoso oouo oo. ss.- oo.- so. oo.- so. so. bb. so.- so. so. ss. ss.- sooo woos osno so. oo. so. so.- oo.: so.- so. bb. oo.- so. oo.- ss.- so. .osuoscoc sous osoocs osno oo.- ss.- oo.- oo.- ss.- so. so.. ss.- so. so.. oo.- so.u so.- mono oooo so ossc> osnm ss.- bb. ss. os.u ss.- os. so.u ss.- oo. so. oo. oo.- oo. umoo usomnossooosz vsno ss.- bk. ss.: oo.- oo. oo. oo.- so. oo.- oo.- so.n ss.- os. nexus onus ssuo so.n bl.u oo.- so.n oo.: ss.- so. ss.- so. so.. so. oo.- so.u scan onus ss-o oo.: oo. os.- ss.- so.. .u so.: so.. so.- oo.- oo. oo.- so.- soon soon anus ssuo oo.- so. oo. oo.: ss. oo. oo.- so. oo.- oo. oo. so. os.- soon noses osno so. so. oo.- os. so. bb. os. oo.- so.- so.. so.: so. so. soon cososuoumcs sono oo.- oo. os. ss. so. so. so. ss.- so. so.- oo.- so.u oo. soon uossssuuoo oo-o ss. Mb. ss. os. ss. so. so. os.u so. oo.- so.- os. oo. sooo omoo soum oo.- oo.- so.u os.- ss.- os. oo.- os.u so. bb.n so. oo.- so. soon .ooua moso woos oono so.- so. so. ss. so.. so. bb. so. ss. ss. ss.- ss.- so.- soususocomxo unseen nos souu so.. so.- so. so.. so.- oo.- oo. oo. bb. so. so. so. so.- socs>uo smcoz so-o os. so.- so.- oo.- oo. so. oo.- so.- so. Hb. oo. oo. oo. unwaumo>as cosuosooua soum so. oo. os.u so.- oo.- so. oo.: oo. oo.- oo. bu. ss. so.- :osuusuumcoo mono: souo oo. ss.- so. so. so.- oo.- so. oo. so.- so. Mb. ob. so.- cosumaaocou sons so. so.- ss. so. bb. so.- so. so.- oo.- ss. ss.- ss.- so.- meoucs «usamu nos oono so. ss.- so. so.. so. ss. oo.: bb.- so. so. oo.- oo.. so. sooo .ooua sauce osuo so. so.- so.. so. os. so. so.- oo.- so. so. os.u os.- so. .ooua aouu woos sauce oouo ss ss ss os s o s o s v s s s MHOQESZ ROHOME omsnussu> “it soonesucouv suo msmda “I APPENDIX E ORIGINAL RESULTS OF VILLAGE LEVEL FACTOR ANALYSIS I .... E. . .I . o o . .... .l . . ..Il.l . NH.I .Osmfl os ss os oo os oo os so ss so so ss .ossm .ucs nos cosmos smz oo-s 00.! mo.! 00. 00.! mH. mo. ov.! 00. mo.! bb.! Nm.! v. 00. .GOHQ OUHH MOM muDQCH 3wz 5vlq 5H. 50.! N0. 50. 00. N0.! NH.I o0. H0. Mb. mN.! bb. Ho. coma 600m owla 00. oo.! MN.I oo.! 0H.! 0H. m0. mo. vH.I 00. N0. bk. #0. wood OUHOHDOOQGH mVIH mo.! v0.! mo.! 00.! mo.! mo.! 00.! 50.! no.! on. 00.! Nm. mo. coma HONHHHDHOh VVIH H0.! H0. 00. MH. No. v0. VN. v0.! 05. 50. MM. No.! mo.! omnon hheuomEOB Hvln 5H.! NN. m0.! Hb.! no.! 00.! mo. N0. 0H.! no. oo.! v0. HN. oozon DCOCdfihmm owlq No.! mH.! mo. 5v.! 00. 00. No. 00.! 00.! m0.! mo. 90.! 0.! Shaw SUMO Go coma uwvouo oMIH oo.! mo.| MH.! mH.! mo.! NH. m0. no. 00. N0. N0. v0. MM. mOHvaQDm usmfiduo>00 mMIA oo.! 50. 00.! ON. vH. mo.! 00.! v0. no.1 00.! H0.! No.! 00. QCOHDDQHHUSOO HuUOH vnla H0.I No.! mo. mo. 0H.! H0. H0. No.! Ml. N0.! N0.! vH. m0. mvcdu ameHHH> MMIH m0. mo. 00. v0.! 00.! .! oo.! HH.! m5. No.! 00. NH.! mN. mHoEHdd mound HMIA bb. mo. N0.! oo.! vH.! mo. no. no. no. oo. 00. mo.! v0. bums MoneH muHuUDO oNIA MN. H0.I N0.! 00. mo. 50.! no. no. 00.! v0.! 5H. 5N. 00.! coma uouoeua 0NIH Nv.l N0. mo. bb. 5H. ON. 50. 00.! 0H.! mo.o 50.! mN. H0. umoo cowuosvoum oNIH mH.! v0. 50. 5v.! m0.! mo.! mo. 00. 00. oo.! bl. vn. vo.! flown UvaHU mNIH oo.! 0H. mm.! mo.! 00.! 00.! HH. VH. vn. HN. mv.! HN. HH.! huwmcoc noHueHdmom vNIH ss. so.- so. ss. so. ss.- so. ss.! so.- ss. ss.! bH.n ss. assaus one as osmoms ssus NH.! NH. MH. mo.! v0.! N0.! mN. 00.! mH. v0. mo. 5m. 00. uHUouo mcwms mumauch 0NIA Mb. H5. HH. 00. mo. 50. 0H. NH. Ho. mo.l N0.! mo. oo. OHUUH couvHHno\HHm5m QHIH om. mo. 00. NH. MH.! 0.! HH. mo. 5H.! mo.! 00.! mH.! 0H.! OHUGH Hanocmu\HHmam QHIH mo. NH.I H . No.! H0. 00. NH. VH. 00. No. No.! 00. 50. .00 .ud00> NUHnfld 5HIA 00. H0. Mb. MH. 5N. N0. 50. NH. v0. no. NH.I NM. #0.! .00 .60Ho quad OHOB mUH50< oHIA HH.! .! Nv. N0.! 5H. HH. «0. No.! oo.! v0. m0. NM. 00.! .U0 .EOHO uuHfidd mHIA 0H.! v5. mH.! No. 50. m0. oH.! Ho. 5H. vo.! N0. 50. NH.I uH504 QHIH Ho. mo. 5m. Ho.! Nb. 0H.! 0N. 0N.! 0H.! 00.! 00.! 0N.I mm. UCUH .UHDO ommua>¢ NHIA 0H.! mH. em. m0.! mv. 0H.! VH. mo.! H0.! mo.l 0N.! HH.! 0H. OQHEHUH oneHauo HHIH 50.! mm.! mo. Hm. 0H.! 50. v0.! 0N. Ho.! 50.! 0N. 00. oo.! GOHuduade can umOHOh 0HIA N0. HH. 0N. 5N.! Ho.! 90.! 00.! mm. 00. No.! on. .! mo. QHOHNOOHH wounHflm m!A No.! MH. N0. 00.! .! mo.! v0. N0.! mH.! mo.! m0. 05. N0.! .mhm .vwam .UCH @Howuoowm oIH HH.! mo.! 00. m0. mm. m .! HH. 00.! H0. v0. oo. mo.! No. a9HmcoucH UGloHo mIH Mb.! no. 50.! H0.! 50.! 50. N0. 0H.! 00.! oo.! 5H.! 00. m0. huHmcquH vHoHHOUHm VIA me. HH. 0H.! mm.! «H.! Ho.! 0H.! 0H.! mo. 00.! bH.! m0.! N0.! QGMH GOQOOHM was UOUORH NIH Ho.! 00.! mm.! mN. no. mo. Ho.! 00.! oo.! bH.I vv.! 0H. H0. UHGHHOOHH “mummHuuH NIH mH.! NH. 50. mo. 00. mH. Ho. 0H.! v0.! 55. NH. H0.! H0.! .OcH nuHQmo Hem HIA NH NH HH 0H m m 5 o m v m N H muonEdz Houoem mossssss> mHmMH ho mBHDmmm HdZHUHmO Hlm mqmda m Xanmmsflw 1J47 lu48 ss.- so. ss.- oo. so. so. oo. oo.: so. bb. oo.: oo. oo.- sus>suoaooum egos sous oo.- ss. so.n os. os. oo. oo.- ss.- oo.- so. ss. so.u ss.- seasons xssm oous oo. ss. oo.- so. ss.- os. bl. os. oo.: oo.- ss.! ss.- so.u osssuou was seasons oous so.u ss. ss. oo. ss. so.n oo. ss.- os.u oo.- .. so.u os. cosos>as¢u can osoam sous os. so.. so.. ss.- ss.- oo. so. ss. bl. os. oo.: oo.: oo.! susocmo ouou osnnsouoz sons oo. so. so. oo. so. oo.- so.- so.. so. oo.- ss.: so.! so. woo: osussaa posse sous oo.- oo. oo.: oo. oo. oo. oo.- oo.- oo.- oo. oo. oo.- oo. souoou unsosom oons so.- so. os. ss.- os. oo. ss.: ss. oo. oo.- ss.- oo. HI. cosuoauuoaoo mus>sss ssns oo.- ss.- ss.- ss.- os. so. ss.! ss. oo. ss.: HH. so. so. cosuuauuocoo assess osas oo.- oo. so. bb. ss.- oo.- ss. oo.! ss.- so.. so.: ss.! oo.- uouooo ous>uom assess ssus ss.- ss.: oo. oo. ss. os.- oo.- so. so.- oo.: ss. so.u ss.! scans osus os. oo. os.n ss. oo.- so. os. so. so. oo.: so. oo.! oo.- uouomo ensue osus ss.- oo. ss. ss.- s.. so.- so. ss. so.u ss. oo.! ss. so.u uouoom usoouo ssus ss. so.- os.n ss.! oo. so. so. ss. oo.- Mb. oo. oo. so. uouomu cosuouuomocnua ssus oo.: ss.- ob. ss.! ss.- ss.- so. ss. oo. so. oo.- ss.- so.- uouoom suumsocs macs osus s .u so. ss.- ss.! ss. so. oo.- ss. ss.- ss. so. os.u so. uouooo cosum>aoxm sous oo.! s .u so. ss. ss.: so.- ss.! os. oo.- oo.! ss. oo. ss. souomo .uausm one pauses sous bb. so. ss. so. oo. oo.- ss.- os. ss.- so. ss. os. oo. uouoos snoross sous oo. ss.- so. so.. so. so.- oo.- bu. ss. os.n ss.: so.u oo. uouooo successor smasce oous ss. so.: so. os.- oo.- ss. os. ss.- so. .u os.- so. oo.: uouooo mono woos oous so. ss.- oo. oo.- so. so.. so.. os. oo. oo. oo. oo.! oo. .uosm .osuon can .uosoas sons ss.- so.» .bb. so.- oo.- so. oo.- ss.- so. so.. so. so.. ss.- noose» sum .oosa .ossm< sous so. ss. so. ss.- so. ss. so.n oo.- oo.- so. so. os. so. museums omssss> oo-s ss.- os. ss. oo. oo.- so.- vs. oo.- ss.- oo. oo.- ss. bl. .uooou ooozuonnosoz so-s oo. oo. os. oo.- ss.- so.- Mb. oo.: so. os. so. so. os. woman ooosss> sous oo.- oo.- oo. oo.- os. ss. Hm. os. ss. oo.- oo. os. so. mumnaos m>suuumaoou oous so. oo.- oo.- ss.- ss. ss. so. ss. oo. oo.- oo. ss. oo.- oo>suosomooo oons so.- so.- so. ss. so.n ss. oo.- oo. oo. so. so. oo.- oo. oucmasosuusa .uus snaps: sous so.- oo. os. ss. so.- ss. oo.- os. oo.- so.n ss.- ss.- os. mmsus>suoo snaps: so-s so. ss. Mb. ss. so. so. oo.- so. ss. so.u os. ss.! so.u .auu socoomu as coasuoso so-s ss.: ss. oo. ss.- so.- os. oo. oo.- os.u ss. so. ss.- oo. suso gasses 0» oocuuoso so-s ss ss ss os s o s o o o s s s muonsdz uouomm moHnessm> sumacsuaooo s-m NH mama. APPENDIX F ORIGINAL RESULTS OF DISTRICT LEVEL FACTOR ANALYSIS so.- mm. oo. so.n os. .aouo smmsocssa so soon .oosa ssua NH.I MM. mo. v0. MM. nuosoosm .oHumm msHs> lea ss.- Mb. so. os.- oo. osmsomosm ssua oo.! oo.- Mb.u so.- so. .moo cosponsommcosa osua ss.- mm. so. ss. ss.! osoazo .mcssu omssuosos ssua ss. os. bb.u so. ss. msoazo msososouoz ssae 00.! v0. Nm. HH. .HH. mHOG3O .mGMHu .uOEdOZ 5NIH Mb. ss. oo.- so. so. ssmnzo msososm osna oo. ss. bM.u so.! ss. memos usso osua so.- so.: MI. oo.- oo.- memos omsm>ouo osns oo.- bb. oo. so.u oo. meson omssooams ssna os. so. oo.- ss. ss.: mumxsoz ssua os.- os. ss.- Mb.- ss. soaps mossss> ssus ss.- os. ss. bM. ss. soaps ucmscsm>oo osue ss.- os. ss.- .bM. so.u cmesmmnsmsm ssua oo. os. ss. MM. os. museums ssue ss. so.u oo.- bb. ss. .moo\usao< ssus os. ss.- so. so. so. .aooxcmsossno osua ss. MM. os. so.- ss. sssaso as :msossno osua Mb. so. os. ss.- ss. msoocs .aso sma osua oo. ss.- ss.- Mb. os.u bass omoosm ssna os. so. os.n oo. so.n rouse was mamas ssua oo.- ss.- oo.- so.- bb.u amass ssua so. os. oo.- ss.- bM.- scosumucssa osne HH.- oo.: os.u ss.- ss.- masses was ssmsos sue ss. os.- ss.- .1 oo.: sums seasons one Mb.- ss.- oo. oo. ss.- scmouoo snomcsssooosz sue bM. ss. os. os.- Mb. sows m moss» osmsomoss one oo.- oo.- os.- so.- NM. sums m mono osmssmosm one oo.- bb. ss. so.u Abs. ssamEmsuumo one sm.- MM. ss.- so.- so.: sooooqa sue so. so.. os. oo.- Mb.u somouss sue ss.- oo.- so.- so. so. mssom sos>ssss sue o o s s s mHOQESZ HOUOflh mmsnsssm> mHmwadzd moaodm Hm>mq BUHmBmHQ Hlm mnmda m XHQmem4 ho mBHDmmm AdszHmo 149 ORIGINAL APPENDIX G RESULTS OF MULTI-LEVEL FACTOR ANALYSIS ‘---—-—- oo.- so. ss.- Hm. ss. ss. so.- ss.- oo.- Mb. .ssm .osso .sas osmsomosm o-s ss. os.- ss.- Mb.- so.- ss.- ss.- os.u _so.u oo.- susocmuas newsman ous so. so.- os.- MM. so.- oo.- os.- ss. oo.: oo. sossomucs ososomuss ous os. os.- oo.- oo. os.- Mb. so.- ss.- so.- so. masons sesame uma sus ss.- so.- ss.- oo. so. oo. oo.- os. ss.- os. onus as scmsusm>cs solo ss. ss.- oo.- oo.- ss. so. MM. ss. ss. ss. usoo cosusosom ss-m so.- ss.- os.- ss.- ss. so.u MM. os. ss. so. soon scseuoso oo-m so. Mb. so. oo.- ss. so.- so. oo.: ss. ss.- soon ooos ssum os. MI. oo. oo.- oo.- os. so. oo.w oo. oo. .ossoscoc sous msooas osuo ss.- oo.- oo. os. so.- oo.- so.- so.- so. so. aouo oooo so osss> osuo os.u oo.- oo. MM. ss.- os. ss. ss.- so. os. saws onus ss-o oo.- ss.- oo. so. os. ss.- so.- os.- so. ss.- smoo susssssums sous os. oo.- so. os. os.- ss. ss.- ss.: ss. ss. mousssocmaxo «assoc nos oonm so.- ss. os.- ss. ss.- Mb. oo. oo.- so.! so. socs>oo soaoz oouo so.- oo. oo. oo.- oo.- oo. os. ss.: so.- so. cosuosuuucou mason sous oo. oo.- so. so.- ss. so. so. so.- os. oo.- :osuaeascoo sous os.- ss. so. so. ss.- ss.- so.- ss.- os. so.u scosm>so>cs oosuo sous os. so. so. ss.- ss.- ss.- so. os.- so.» os. auosmu ocss squam>so>cs sous ss.- ss.- ss. ss. ss. ss. oo. ss. ss. os.a seasons auosos onus solo ss.- os. os. ss.- oo. ss. ss.- so. ss. os.n nauseous osomm ooum so. ss.- os. Mb.- ss. ss. ss. os. ss. ss. masomuoss osomm oouo os.- oo.- os. MM.- os. so.u ss. so.- ss. so. .uusm .osso m>sscoucs osuo so.- so.- ss. oo. so. so.. so.. os. MI. ss.- .smoas .osso m>sscoucs ssuo so. so.. oo.- os. oo.- so.- ss. so. MM. oo.- ssssss so smegma .ouo osuo so. Mb. so. so.. os.- ss.- so. os.- MM. oo.- assess so some .ouo ssum oo. oo.- so.- so.- os.- oo.- oo.- oo.- .MM. so. noses onus ssuo so. ss.- oo. oo.: ss.- oo.- os. ss.- MM. oo.- .mso sonausso osuo so.- ss.- oo.- ss.- oo.- os. os. oo. MM. so. .suo «scam ssuo oo. oo.- oo.- os. so. so.: oo.- so.- MM. so. .ouo o>suusoaooo osno oo.- ss. os.- so. so.- ss.- so. ss. so. so. .oso uoooo som osuo os.- oo.- ss.- os.- oo.- oo. ss.- os. so. ss. scmsusm>cs .ooua sauce ssno ss. ss. ss.- ss. so. so. MM.- vs. os. ss. cosuoesonou ssuoa osuo ss. os.- ss. ss. so.- so. oo.- bM. os. ss. mEooas sssEso sauce ouo so.- so. so.- ss.- so. oo.- os.- MM. ss. os.u cosposoao onus ouo so. ss. so.- os.- oo. ss.- ss.- ss. ss. ss.- asnoumnzo onus ouo os s o s o o o s s s nonesz souoem mmHnsHHe> mHmNH¢z< m0804h HM>WHIHBHDZ ho mBHDmmm HdZHUHmO U xHQmemd HIU mHfldB 15() 1151 oo.- ss. MM.- os. ss.: so. ss. so. ss.- so. .mmo cosumuqumcmue osns ss. ss.- MM. ss. oo. so. so. ss.- ss. so.u mumcso .mcmuu omssuouoz osua oo. oo. ss. ss. Mb. so.. so.. so.- so.. MM. .mumc3o .mcuuu .uoecoz ss-s Mb. os. ss. so. so. oo. oo. ss.- oo. om. mumczo osusosm osua ss.- ss.- Mb.- so. so. so.. ss.- oo.- so. ss. mono“ usso ss-s oo. so. .MM. ss. bb.- os.- so.- so.. os.- so. memos omusunmmc ssna ss. oo.- MM. so. .MM. ss.- os. os.- oo.- os. “sous unassum>oo os-s ss. oo.- ss.- oo. o . oo.- ss. ss.- ss.- o .- swammacsuam ssns so. so. vs. so. .MM. oo.- ss. so.: ss.- so. muoeums os-a ss.- so. so. so. MM. ss.- ss. oo. oo.- oo. mom\couossnu os-a ss.- oo.- so.- so. oo. so.- so. so. ss.- bM. amuse was Mensa ssus oo.- os. so.- so. so.: ss. so. ss.- ss. bM.u muosuuucusm osus so. ss.- ss.- so. oo.- so. so. vs. so. ss.- mnsunu can unmuos sue ss.- oo. so.- ss. ss.- so. ss.- os. os.u bb. guns unassso o-a ss. ss.- os.- oo. oo.- ss.- oo.- oo.- so.- MM. you» a coco osmsswusm ona oo.- bb.- os. so. so.- ss.- ss.- ss.- so.- so. mugmfimsuumm «.9 ss. MM.- so.- ss. so.- ss. oo. oo.- so.- MM.- somouca s-s ss.- so. ss. so. oo.- ss.- os.- ss. ss.- MM.- somoums sus Mb. ss. ss.- ss. oo. oo.- oo. ss.- ss. MM. assou sus>ass4 sus oo. so.- ss. vs. so. so.. ss.- ss.- so. MM. sus>suosooum cams sous oo.- so.- so.- oo. oo. ss. ss.- MM.- so. MM. susmcoo coon osnuuouoz so's ss. so.- oo.- MM. ss.- ss.- os. ss. ss.- oo.- cums assascm usmuo ssns oo. so. os. so. ss.- so. ss. ss. ss.: so. cosuossuncoo assess os-s Mb. ss.- ss.- ss. ss.- os.- ss. ss. ss.- so.u subconasg suasa< oous s .- oo.- ss.- ss. ss. ss.- so.- ss.- oo.: os.u uouomm mouo boom sous so. ss. os.- so. so. oo.- ss.- so.- vs. ss. msosnosm .usumu .uaocs sous so. oo.- oo. ss. ss.- os. os. os.- so.. so. sums» mousss> ssus ss. so.- oo.- oo. oo. Mb. ss.- so. ss. so. mo>suaummooo ssus ss. so. oo.- ss. so. ss. ss.- Mb.- oo.- ss.- mucuMsosuumm .uou sagas: song so. so.. oo.- so. ss.- Mb.- os. oo. ss.- oo.- suso somuus o» magnumso ssus so. so.- so.- so. so.: .MM. oo.- so.- so.- so.. can: cams owns so.- so.- M|.- so. so.- ss. so. so.- ss.- ss.- mason sumsomaos sous ss.- so. so. so. ss.- ss. os.- so. ss.- oo. mmsosmnsm unwecum>oo ssns os. so.- MM. ss. so.- ss. oo.- Mb. oo.- ss. mucus mousss> ssus so. so. oo. os. MM.- oo.- ss. ss. ss.- oo.- «smascu omums ssus MM. oo. so.- ss. so. ss. ss.- oo.- so. oo.- cams moans mosmuao os-s ss.- ss. so.- os. so.- so. bb. ss.- vs. so. usomuo mass: mumsuam os-s so. so. os.- oo. oo. ss. ss. ss. so. oo. osuuu couvssso\ssmsm ss-s ss. so. os.- oo. oo. ss. ss.- oo.- ss. ss.- .osoo .uuoo> nusaoc ssus ss.- os. ss.- ss. so.- so.- ss. ss.- os.- oo.- asses os-s os o o s o s v s s s Honesz uouomm wasnusuu> somscsucooo sno msm4a APPENDIX H ORIGINAL RESULTS OF FACTORIZATION, MULTI-LEVEL ANALYSIS 0N BANDUNG REGENCY wuzmumm UZDQZ¢m ZO mHmMA¢24 AM>NAIHHADS sZOHaduHmoaodh m0 mBADmmM Him mnmda m XHQZNAMd AfiZHUHMO ss. so. ss. os. oo.- ss. so.u ss. ss.- MI. .msm .usss .ucs usossooss sus so. os.- os. ss.- so. os.- oo. so.- so.. ss.: sussnouas unassua sus so. ss.- ss. so.u oo. ss. so. so. ss.- ss. susscoucs asmssmosm ous .bb. so. ss. ss.- ss. so. ss.- os.- ss.- so.u oeoocs assume sum sus ss. ss. ss.- so.. so.- os. so. ss. Mb.- ss. cams as unosusm>cs so-s ss. so.- os.- ss.- oo.- os. so. so. .MM. os. usoo cosuuosum sons so.u oo.- ss.- ss.: ss. ss. ss. ss. MM. oo. usoo sasnuosu osns ss.- so. ss.- ss. MM.- so. ss.- ss. ss. os.- uuoo woos ssns oo.- oo. so. so. MM.- oo.- oo.- so. so. os. .osssuaoa sous osoucs os-s oo.- oo.- so.- bb.- so. os. so.u os.u so. oo. souu woos so mssu> ssus ss.. .MM. ss.- ss. ss. so. oo.: ss.: so. so. soon cams ssus ss.- ss.- so.- oo.- ss. ss.- oo. ss. so. ss.- usou nosssssums sous so. so. os. so.- ss.- ss. so.- ss.- so. ss. mmususocmmxo assume nos ssus oo. ss. so. ss.: os.u oo. so. ss.- so. so.- sscs>us smcoz oo-s so.- oo. oo.- ss. oo.- ss. so.u ss.- Mb. os.- cosuoauumcoo «use: ss-s ss.- so. ss.- ss. oo.- Mb. os.n os. so. ss.- cosumasscoo sous so.. so.- ss. so. ss. s.u os.u ss.- so. so. scuam>so>cs coups sous ss.- oo.- oo. so.u oo.: s .. ss.- oo.- oo.- ss. enemas cans uaoam>so>cs sous so. oo.- os.- Mb. ss. ss. so. ss. ss. ss.- summons euosou vans sons ss. ss. ss. MM. os. os.- ss. so. oo. ss.- sensuous osuus sons so.- so.. oo. bM. ss. ss. so. so. ss. ss. mascouuss osoum sons oo.- oo.- os. os. so. ss. os.s os. ss. ss. .uuum .osss o>smaouas ssus so.- ss.- ss. so. ss. ss. ss.: so.u MM. os. .saoca .osas o>sscoucs ssus so. os.- ss. so.- so. so. ss. so.- bM. so.- assess so usages .suo osus ss.- so. os. ss. so. ss.- so.u ss.- MM. oo.- assess «0 soon .suo ssus so. ss.- os. so.u os. os.- os. ss.- bM. so. scosas zoos ssus oo.- ss. ss. ss. ss. ss.- so.- ss.- MM. ss.: .suo suuausno osum ss. os. so. so.- ss. so. ss. so.- bM. so.- .suo amass ssns so. so.u so. so.. so. ss.: so.. so. MM. so.n .suo m>suusmmooo ssnm oo.- os.- oo. oo.- ss.- ss.- .MM. oo.- so. ss. .suo usoos som ssus ss. ss. so.- ss.- oo.- so. os. so.- so. os. unoeumo>cs .oous sauce ss-s so.- ss.- ss. os.- so. so. so. os.- ss. so. cossmssscoo sauce ss-s os.- ss.-. ss. ss.- so. ss. MM. ss.- so. so. «goons sssaus suuoa s-s oo.- ss.- ss. ss. oo.- ss. MM. so. os. so.- cosuuuomo sans sus oo.- ss.- ss. so. os.- ss. ss. so. ss. so.- asnmsocao cans ons os s s s s s o s s s mosnmsuu> Hmnfidz HOuOMh . ssswmfit 1J52 .1553 ss. ss.- ss.- os. ss.- os. ss. ss. oo. so. .ooo nosuouuosocoua osus so.- ss. ss. oo.- ss. so.u ss.- ss. ss. oo.- ouocuo .mMcnuu oousuouo: ss-a ss.: so. ss. oo. so.| mo. on.| oo. oo. hm. uuoczo .mucnuu .uoadoz ssua ss.- so. so. os. so.- so. bb.- so. oo.- so. ouooso osososm osua oo. ss. os.- ss. os. so.- os. so. os.- MM. Moos unso ssua ss.- os. so. oo.- ss. ss. ss.- bb.- os.- os. one» oousoooos ssna os.- oo. ss.- so. os. so.- ss.- MM. os. ss. Humus unoscuo>ou os-a os. so.- oo. ss.- so.- ss. os.- MM. so. os.- caaooocsoso ss-a oo.- ss. ss.- ss.- so. ss. ss. MM. oo. so. ascends ss-a oo. oo.- so. so. oo. oo.- os. MM. oo.- so. .moo\couossno osua so. so.- so. so.- so.- ss. oo. oo. oo.- .Hb. cause can Mensa ssna so.- os. so. os.- ss.- ss. oo.- oo.- so. MM.- oaosuoucoss osua ss. ss.- so. oo.- ss. so. .. ss.- so. ss.- saunas can saunas sua ss.: ss. so. os. so. os.- ss. so.- ss.- Mb.- suns unassun o-a os. ss.- os.- os.- so.- ss.- so. so.- so. MM. sous a coco osossoosm sun so. ss. oo.- oo.- os. ss.- so.- oo.- ss.- bM. oucoaosuuos oua ss. so. so. so.- ss. so.- oo.- so.- os. s.u soooono sua oo.- so.» os.u ss. oo.- ss. os. oo.- ss.- bM.- sooouus sun so. oo.- oo. ss.- ss.- so.n os.- so. ss. MM. asses sos>sss< sua ss.- os. os. ss.- so. ss. ss.- so.: so.- ss. sus>suoaooso onus so-s oo. ss. so. ss. oo.- so.n ss. oo. so. so. susscoo coon osaououox so-s so.- so.- so. Mb. oo.- so. so. so.. so.. ss.- coma ssoascu among sous oo.- ss.- so.- ss. os. oo. so. so.- so.- so. cosuossuocoo ossnss os-s so. ss.- ss.- os. os. ss.- os. ss.- ss.- so.u uouoos suoconoso soasas so-s so. ss.- ss.- os.u oo. ss.: so. os. ss.- ss.- uouooo mono ooos so-s so.- os. oo. ss.- ss.- so. so.- so. os. os. success .osuso can .ssoos oo-s so. ss.- .. ss. so.. so.- ss.- ss.- so. so. sun». ooosss> ssus so.- os. so. oo. so.- oo.- ss. oo. os. so.u oo>suosooooo ssus so. ss.- ss. os. os.u os. ss.- so. ss.- ss.: sucumsosusos .uuo scans: os-s so.- so.- os.- oo. ss.- ss. os. ss.- oo. ss.- suso Humans 0» ooaouuso ssus ss. oo. ss. os. oo. os. so.- so.. so.. so. can: ooos sous os. ss. so.- ss. oo. os. os.- oo. ss.- os.n cocoon suosooeoa sous ss.- oo. so. so. so. so.- ss. ss.- ss.: so. ousosunsu uaoacuo>ou ss-s ss.- Mb. so.- ss. oo. os. ss.- ss.- so.n ss. songs ooosss> ssus oo. os.- s .- so. ss.- so.- os. MM.- ss.- oo.- usuasou moans ssus so. so.- ss. ss. so.- so. so.- so. ss.- oo. owns scans oosouso ss-s os. ss.- so. os. ss.- ss.- ss.- so.- oo. so. ssomuo ocsss museums os-s ss. so. oo.- oo. ss.- ss. so.- so. so. so.- osuou cosossno\ssoos ss-s os. os.- ss. so. oo.- so. so. oo.- so.- ss.- .oooo .uooo> masons ssus oo. oo. ss.- so.- ss.- os.- ss.- so. so. os.- usooo osus os s o s o s o s s s Hmnfidz uouowh :uwsscwucouv Him Hams“. oosnosuo> 7' J" I": ‘P ..d ‘HQE? APPENDIX I ORIGINAL RESULTS OF SECOND ORDER FACTORIZATION, MULTI-LEVEL ANALYSIS OF BANDUNG REGENCY APPENDIX I TABLE I-l ORIGINAL RESULTS OF SECOND ORDER FACTORIZATION, MULTI-LEVEL ANALYSIS OF BANDUNG REGENCY (1273) Description of Patterns Factor Number 3 10. Proper land use, building infra structure, and provision of development facilities combined with demographic suitability, will pre— cede industrial activity. Organization participation and pri- mary expenditure back-up the development activity for human progress. The demographic composition will influence intensive agriculture. Land tenure system and basic investments precede technological advancement. Last decade program influences family income/expenses and industrial activity. The sources of income are influenced by physical land use as well as the economic conditions. Access to development program and having proper production tools will stimulate development activity. Intensive agriculture and organi- zation participation are the base for collective action. Agricultural modernization and extensive farming are related to public security. Development activity through extensive farming will lead to industrial activity. Sum of squares .10 .13 .43 -.11 -.13 -.29 .22 -.07 1.14 -.16 .19 -.OO .16 -.10 .14 -008 -.21 .13 .05 1.12 -019 1.08 154 APPENDIX J ACTUAL AND ESTIMATED VALUE OF TOTAL FAMILY INCOME (Y), AND THE RESIDUALS FROM THE LINEAR MULTIPLE REGRESSION ANALYSIS (in rupiah, $1.00 = Rp. 415.00) OF BANDUNG REGENCY (1973) APPENDIX J TABLE J-l ACTUAL AND ESTIMATED VALUE OF TOTAL FAMILY INCOME (Y), AND THE RESIDUALS FROM THE LINEAR MULTIPLE REGRESSION ANALYSIS (in rupiah, $1.00 - Rp. 415.00) OF BANDUNG REGENCY (1973) Difference Difference gillage Actual Estimated Between Actual as Percentage amples Value Value Name of Y of Y and Estimated of Actual Value of Y Value of Y 1. Nyalindung 80.663.130.00 142.547.800.00 -61.884,670.00 76.7 2. Cipatat 82.426.250.00 36.274.430.00 46.168.820.00 56.0 3. Cimareme 486.916.000.00 354.211.000.00 132.705.000.00 27.3 4. Ngamprah 289.939.100.00 253.192.200.00 36.746.970.00 12.8 5. Cibabat 221.096.100.00 263.227.200.00 -42.131.040.00 19.1 6. Melong 295.977.300.00 271.158.800.00 24.818.500.00 8.4 7. Antapani 286.326.700.00 201.928.400.00 84.398.300.00 29.5 8. Ciburial 212.991.600.00 209.815.600.00 3,175,954.00 1.5 9. Cipadung 182.169.400.00 204.004.400.00 -21.835.000.00 12.0 10. Cinunuk 263.435.300.00 201.987.300.00 61.447.950.00 23.4 11. Bojongloa 181.653.100.00 247.965.700.00 -66.312.600.00 36.5 12. Haurpugur 192.510.000.00 276.334.300.00 -83.824.260.00 43.5 13. Cikasungka 128.076.300.00 208.295.900.00 -80.219,650.00 62.6 14. Tenjolaya 215.677.400.00 337.126.500.00 -121,449.100.00 56.3 15. Nanggeleng 219.412.800.00 202.036.800.00 17.375.900.00 7.9 16. Ciroyom 119.122.500.00 166.221.700.00 -47,099,210.00 39.5 17. Rajamandala 193.066.300.00 176.974.100.00 16.092.190.00 8.3 18. Cikalong 206.049.500.00 172.619.800.00 33.429.710.00 16.2 19. Cihideung 218.455.600.00 181.766.100.00 36.689.500.00 16.8 20. Cihanjuang 213.550.600.00 246.563.500.00 -33.012.830.00 15.5 21. Cikahuripan 324.447.000.00 333.173.500.00 -8.726.543.00 2.7 22. Cibodas 244.621.900.00 284.869.000.00 -40,247,110.00 16.5 23. Buahbatu 505.021.900.00 364.138.800.00 140.883.100.00 27.9 24. Cipagalo 230.606.900.00 310.472.200.00 -79.865.280.00 34.6 25. Sukamenak 352.888.800.00 325.647.300.00 27.241.450.00 7.7 26. Sukapura 346.582.500.00 272.740.900.00 73.841.640.00 21.3 27. Sukagalih 189.988.100.00 148.502.800.00 41.485.360.00 21.8 28. Sukasari 270.519.400.00 266.769.500.00 3,749,837.00 1.4 29. Kamasan 237.038.500.00 181.367.800.00 55.670.740.00 23.5 30. Nagrak 365.275.000.00 293.862.500.00 71.412.460.00 19.6 31. Sukamaju 155.864.400.00 196.510.600.00 -40.646.230.00 26.1 32. Warnasari 112.097.700.00 225.143.000.00 -113.045,200.00 100.8 33. Cipeuyeum 178.256.400.00 174.683.300.00 3,573,096.00 18.8 34. Cikawao 137.492.500.00 180.640.600.00 -43.148,130.00 31.4 35. Ibun 213.332.600.00 183.415.700.00 29.916.950.00 14.0 36. Cipedes 174.183.600.00 177.683.600.00 -3.500,019.00 2.0 37. Padasuka 297.836.300.00 367.711.800.00 -69.875,510.00 23.5 38. Wangisagara 267.834.100.00 262.215.700.00 5,618,403.00 2.1 39. Manggahang 229.530.000.00 196.098.000.00 33.432.020.00 14.6 40. Babakan 177.503.100.00 172.317.800.00 5,185,344.00 2.9 41. Sadu 246.846.400.00 241.998.400.00 4,847,978.00 2.0 42. Cilegon 189.350.000.00 235.971.400.00 -46.621.410.00 24.6 43. Panyocokan 230.835.600.00 257.425.600.00 -26.589,980.00 11.6 44. Cipelah 201.512.500.00 172.083.000.00 29.429.530.00 14.6 45. Sukamulya 195.052.900.00 160.466.100.00 34.586.800.00 17.7 46. Cikoneng 191.357.500.00 193.130.600.00 -1,773,140.00 0.9 47. Selacau 195.740.600.00 221.066.400.00 -25.325.750.00 12.9 48. Utama 201.915.900.00 239.504.800.00 -37.588.920.00 18.6 49. Tanjungjaya 150.918.400.00 148.568.800.00 2,349,562.00 1.6 50. Tanjungwangi 119.559.400.00 189.089.700.00 -69.530.360.00 58.2 51. Sindangkerta 355.154.400.00 249.645.500.00 105.508.900.00 29.7 52. Cikadu 216.898.800.00 252.708.800.00 -35.810,090.00 16.5 53. Gununghalu 351.742.800.00 329.709.100.00 22.033.610.00 6.3 54. Sodong 406.304.600.00 390.058.100.00 16.246.500.00 4.0 155 SELECTED BIBLIOGRAPHY SELECTED BIBLIOGRAPHY Arndt. H. W. "Survey of Recent Development." Bulletin of Indonesian Economic Studies 7 (July 1971): 3. Cattel, Raymond B. "Factor Analysis: An Introduction to Essentials II. The Role of Factor Analysis in Research." Biometrics 21 (June 1965): 405-35. Draper. N. R., and Smith, H. Applied Regression Analysis. New York: John Wiley & Sons, Inc., 1966. Fruchter, Benjamin. Introduction to Factor Analysis. Princetown. N.J.: D. van Nostrand'Co.. Inc., 1954. Gregg. Philip M.. and Banks. Arthur S. "Dimension of Political System: Factor Analysis of a Cross Polity Survey." American Political Science Review 59 (September 1956): 602-14. Hadi, Hariri. "Pembangunan Daerah Dalam Repelita II." Prisma 3 (April 1974): 69. Higgins, Benjamin. Indonesia's Economic Stabilization and Development. New York: Institute of Oacific Réiation, 1957. King. Dwight Y. Social Development in Indonesia. Jakarta: Biro Pusat Statistifi, 1973: Koentjaraningrat. ed. Villages in Indonesia. Ithaca. N.Y.: Cornell University Press. 1967. Kusumadewa, Arie L.. et a1. Laporan Penelitian Padat Karya di Jawa dan Madura. Jakarta: Team Penelitian Padat Karya. 1972. Mangkusuwondo. Suhadi. "Dilemmas in Indonesian Economic Development." Bulletin of Indonesian Economic Studies 9 (July 1973): 30-31. 156 157 Mubiyarto. Pengantar Ilmu Pertanian. Jakarta: Lembaga PeneIitian. Pendidikan dan Penerangan Ekonomi & Sosial. 1973. Nie, Norman. ed. SPSS: Statistical Package for the Social Science. New York: McGraw Hill Inc., 1970. Rummel, R. J. "Understanding Factor Analysis." Conflict Resolution 11 (December 1967): 444-80. . Applied Factor Analysis. Evanston: North- western University Press, 1970. . H“ I ill. '- Sajogyo. "Modernization Without Development in Rural Java." Bogor, Indonesia: Bogor Agricultural University, 1973. A mimeographed paper con- tributed to the study on changes in agrarian 5 structure. organized by F.A.O. of the U.N., 1972-1973. Sibero, Atar. "Program Bantuan Pembangunan Kabupaten/ Kotamadya." Ekonomi dan Keuangan Indonesia 21 (June 1973): 954112. Subroto. "Kebijaksanaan di Bidang Kesempatan Kerja & Transmigrasi Dalam Repelita II." Prisma 3 (April 1973): 18-29. Unnonimous. Indonesia: Perspective and Proposals for United States Economic Aid. New Haven: Yale University Southeast Asia Studies. 1963. Wileden, Arthur F. Community Development. Totawa. N.J.: The Bedmister Press. 1970. llf'l1111WIWIWIIIWH‘W|H1H||VIIWI1|1|11.1 3129310786 4195