LIBRARY University Michigan State This is to certify that the dissertation entitled Family and Internal Migration in Taiwan presented by Chun-Hao Li has been accepted towards fulfillment of the requirements for iology- Urban Studies ELL, 1% Datefléw. /y 02(r:}_ 0 MS U i: an Affirmative Action/Eq ual Opportunity Institution Major professor JzzKKw 0-12771 PLACE IN RETURN Box to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 6/01 cJCIRC/Dateouepes-sz FAMILY AND INTERNAL MIGRATION IN TAIWAN By Chun-Hao Li . AN ABSTRACT OF A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Sociology and Urban Affairs Programs 2001 Co-chairs: Dr. Brendan P. Mullan and Dr. Rita S. Gallin ABSTRACT FAMILY AND INTERNAL MIGRATION IN TAIWAN By Chun-Hao Li Three theoretical frameworks have dominated migration research -- the individual, the structural, and the family/household perspectives. In Taiwan the individual and the structural perspectives most frequently have been adopted. The family perspective has never been used to examine migration in Taiwan. This research uses this perspective to examine the rural-to-urban migration of one group of villagers at two points in time -- the mid-19608 and the late 1970s -- using data collected by Professors B. Gallin and R. S. Gallin. I argue in the dissertation that migration is a family sustenance and/or mobility strategy that deploys individuals on a selective basis to overcome the structural constraints of a changing economic structure. First, I examine migration at the family and structural levels. This analysis explores the associations between labor migration and influential factors such as family type, family landholding, and participation in the local labor market at the family level to illustrate the relationships among family structure, local and national economic structures, and labor migration. Second, I move beyond the traditional “cost-benefit ” argument of the individual perspective and consider how the migration process is affected by family power dynamics, as they are shaped by the intersection of gender and age. Here I discuss male-female power relations and power relations among female villagers. My analysis encourages an expansion of the individual perspective of migration, showing how personal characteristics are implicated in the decision-making process of migration. In the conclusion, I discuss the theoretical and empirical contributions of the research. ACKNOWLEDGMENTS I would like to express my deep appreciation to Dr. Rita Gallin, who as my committee co-chair, spent many hours helping me refine my dissertation. In addition, Dr. Gallin and her husband, Dr. Bernard Gallin, unselfishly provided to me over 40 years of their field data from Hsin-Hsing Village. This act of kindness and unselfish sharing of their research data is a true reflection of their scholarship and their deep devotion to their students. I wish to also acknowledge to my co-chair, Dr. Brendan Mullan, and his patient and enthusiastic support of my dissertation. I would also like to thank Dr. Stan Kaplowitz for paying extra attention on my chapter on statistical analyses, and Dr. John Schweitzer for serving on my dissertation committee. It was with their encouragement that I gained much confidence while finishing my dissertation. In addition, a special note of thanks goes to my parents, brother, sister, and many friends who gave me considerable emotional support. Finally, my deep debt is to my wife, Shu- Yao Hsu, for her support, encouragement, and patience during these years. iv TABLE OF CONTENTS LIST OF TABLES ........................................................................................................... ix LIST OF FIGURES ......................................................................................................... xi CHAPTER I INTRODUCTION ............................................................................... 1 1.1 Introduction 1 1.1.1 Causes of Migration 3 1.1.2 Consequences of Migration 5 1.2 Review of Migration Theories 7 1.2.1 The Individual Perspective 7 1.2.2 The Structural Perspective 9 1.2.3 F amily/Household Perspective 10 1.3 About Taiwan 12 1.3.1 Industrial and Economic Development 12 1.3.2 Family 14 1.3.3 Overview of Internal Migration Research in Taiwan 16 1.3.3.1 The Individual Perspective 16 1.3.3.2 The Structural Perspective 18 1.3.3.3 F amily/Household Perspective 18 1.4 Purposes and Significance of the Dissertation 20 1.4.1 Purpose of the Dissertation 20 1.4.2 Significance of the Dissertation 21 1.5 Organization of the Dissertation 21 CHAPTER II EARLY SETTLEMENT, ECONOMIC DEVELOPMENT, AND MIGRATION ........................................................................... 24 2.1 The Early Settlement Patterns in Taiwan -- Before 1895 24 2.2 Early Taiwanese Economy 26 2.2.1 The Taiwanese Economy during Japanese Occupation (1895-1945) 26 2.2.2 The Political Economy of Contemporary Taiwan (Afier 1945) 30 2.2.2.1 Land Reform and Land Ownership 31 2.2.2.2 Agricultural Development 33 2.2.3 Industrial Development and Policy in Taiwan 38 2.2.3.1 Rural Industrial Development 41 2.3 Population Distribution and Migration in Taiwan 43 2.4 Taiwanese Families and Migration 48 2.4.1 Family Dynamics in Taiwan 48 2.4.2 Migration Decision and Power Dynamics within Family 52 2.4.3 Consequences of Migration 54 2.5 Summary 56 CHAPTER III THEORIES, RESEARCH FRAMEWORKS, HYPOTHESES, DATA, AND METHODS ................................................................. 59 3.1 Introduction 59 3.2 A Review of Migration Theories 61 3.2.1 Individual Perspectives of Migration Theories 61 3.2.2 Structural Perspectives of Migration Theories 68 3.2.2.1 Intervening Opportunities 70 3.2.2.2 Uneven Economic Development 71 3.2.2.3 Regional Restructuring 72 3.2.3 Family Perspectives of Migration 75 3.2.3.1 Accessibility to Land and Participation in Local Wage Labor Markets 1 76 3.3 Research Framework and Hypotheses for Migration in Taiwan 78 3.3.1 Family Type 80 3.3.2 Accessibility to Land 82 3.3.3 Participation in Local Wage Labor Markets 84 3.3.4 The Integrated Research Model 85 3.3.5 Research Frameworks and Hypotheses on Family Power Dynamics 86 3.4 Data Sources 90 3.4.1 Surveys 91 3.4.2 Household Registration Records 93 3.4.3 Database and Variables 94 3.4.4 Operationalization of Variables 96 3.4.4.1 Landholding 96 3.4.4.2 Local Wage Labor Market Participation 100 3.4.4.3 Migration 102 3.4.4.4 Other Variables 102 3.5 Methods 103 3.5.1 Descriptive Methods 103 3.5.2 Inferential Statistics--Multivariate Analysis 105 3.6 Summary 105 CHAPTER IV THE PROFILE OF THE RESEARCH AREA - HSIN- , HSING VILLAGE- - ....... .......... - 107 4.1 Introduction 108 4.2 Climate and Spatial Layout of the Village 1 l 1 4.3 Family Structure 116 4.4 Economic Infrastructure 121 4.4.1 Agriculture Sector 121 4.4.2 Industry and Service Sectors 127 4.5 Demographic Infiastructure 135 4.5.1 Gender 136 4.5.2 Age 140 4.5.3 Demographic Process -- Fertility 148 4.6 Socio—economic Infrastructure 153 4.6.1 Education 153 vi 4.6.2 Occupation 4.6.2.1 Non-Migrant Villagers (1965) 4.6.2.2 Migrant Villagers (1965) 4.6.2.3 Non-Migrant Villagers (1979) 4.6.2.4 Migrant Villagers (1979) 4.7 Summary CHAPTER V THE DETERMINANTS OF RURAL-URBAN MIGRATION FROM HSIN-HSIN G VILLAGE ............................................. 5.1 Family Type and Migration 5.1.1 Theoretical Position and Hypothesis 5.1.2 Measurement 5.1.3 Analysis and Discussion 5.2 ‘ Accessibility to Land and Migration 5.2.1 Theoretical Position and Hypothesis 5.2.2 Measurement 5.2.3 Analysis and Discussion 5.3 Local Labor Market Participation and Migration 5.3.1 Theoretical Position and Hypothesis 5.3.2 Measurement 5.3.3 Analysis and Discussion 5.4 The Best Predictor of Labor Migration 5.5 Summary 210 CHAPTER VI MIGRATION AND FAMILY POWER DYNAMICS IN HSIN -HSIN G VILLAGE - .......................... - 6.1 Overview of the Individual Perspective of Migration Theory 6.1.1 Characteristics of Villagers in Urban Cities 6.2 Theoretical Position and Hypotheses 6.3 Measurement 6.4 Analysis and Discussion 6.4.1 Age and Migration among Female Villagers 6.4.1.1 Migration among Women 6.4.1.2 Employment Status among Married Female Migrants 6.4.1.3 Employment Status of Migrant and Non-Mi grant Women 6.4.2 Age Differentials between Male and Female Villagers in the Urban Wage Labor Market 6.5 Summary CHAPTER VII CONCLUSIONS AND DISCUSSION ..................................... 7.1 Summary of Dissertation 7.1.1 Research Intentions 7.1.2 Research Findings 7.2 Significance of Dissertation 7.3 Limitations of Dissertation 7.4 Some Additional Thoughts vii 158 159 163 165 169 172 175 177 177 178 179 190 190 191 193 198 . 198 199 200 206 214 214 216 ' 223 229 231 231 234 238 240 249 255 259 261 261 262 267 273 279 APPENDIX A 288 BIBLIOGRAPHY 290 viii 2.1 2.2 4.1 4.2 4.3 4.4 4.5 4.6 LIST OF TABLES Sources of Gross Domestic Product in Taiwan, 1952-1993 ..................................... 37 Population in Localities of 100,000 and more and between 20,000 and 100,000 Residents in Taiwan : 1962-1991 ............................................................................. 45 The Population and Number of Families by Family Types, 1957-1958, 1965, and 1979 .................................................................................................................. 117 Landholdings by Family Types, 1949, 1951, 1957, 1965, and 1979 ...................... 122 Descriptions of Land-holding: Hsin-Hsing, 1965 and 1979 ................................... 124 Descriptions of Local Labor Market Participation: Hsin-Hsing, 1965 and 1979129 The Population of Hsin-Hsing Village by Gender and Age Group, and the Mean Ages and Dependency Ratio, 1965 and 1979 ......................................................... 138 A Comparisons of Crude Birth Rate, General Fertility Rate, and Child-Woman Ratio of Hsin-Hsing Village, Taiwan Area, Chang-Hua County, and Pu-Yan Rural-Township : 1965 and 1979 i ......................................................................... 149 4.7 Average Years of Educational Attainment of Hsin- -Hsing Villagers, 1965 and 1979 ........................................................................................................................ 155 4.8 Hsin—Hsing Villagers’ Occupations by Migration Status and Labor Force: 1965 and 1979 .................................................................................................................. 161 4.9 Average Age and Educated Years of Hsin-Hsing Villagers by Selected Occupations, and Migration Status: 1965 and 1979 ............................................... 168 4.10 Entrepreneurship among resident villagers, Hsin-Hsing: 1965 and 1979 .............. 169 5.1 Analysis of Variance for Labor Migration by Family Types, Hsin-Hsing: 1965 and 1979 .................................................................................................................. 179 5.2 Comparison of the number of labor migrants between conjugal, stem, and joint families in Hsin-Hsing, 1979 .................................................................................. 181 5.3 Characteristics of Hsin-Hsing villagers working on family land within conjugal, stem, and joint families with labor migrants, 1979 ................................................. 186 5.4 Correlation between Accessibility to Land and Labor Migration in Hsin-Hsing: 1965 and 1979 ......................................................................................................... 194 5.5 Correlation between Local Wage Labor Markets and Labor Migration in Hsin- Hsing: 1965 and 1979 ..................................... 201 5.6 Labor Force and Local Labor Market Participation in Hsin-Hsing by Family Types: 1965 and 1979 ............................................................................................. 203 ix 5.7 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 Multiple regression for labor migration, family type, family’s local labor-force participation rate, proportion of family members working locally for paid wages, landholding per working family member, and landholding per family member for Hsin-Hsing Village: 1965 and 1979 .................................................... 208 Villagers’ Migration Status by Gender, Marital Status, Hsin-Hsing Village: 1965 and 1979 ......................................................................................................... 219 Mean Ages of Villagers by Gender, Marital and Migration Status, Hsin-Hsing Village: 1965 and 1979 ........................................................................................... 220 Logistic Regression of Labor Migration, Hsin-Hsing: 1965 and 1979 .................. 232 Hsin-Hsing’s Migrants in Conjugal Units and Migrant Conjugal Units, 1965 and 1979 .................................................................................................................. 241 Women Classified by Occupation, Migration Status, and Presence or Absence of Preschool Children, Hsin-Hsing Village, 1979 .................................................. 243 Support received by married women in the labor force, who had preschoolers, Hsin-Hsing: 1979 .................................................................................................... 244 Occupational Matrix of Mothers-in-law and Daughters-in-law, Hsin-Hsing: 1979 ........................................................................................................................ 245 Analysis of Variance for the Ages of Villagers in the Urban Wage Labor Market by Gender, Hsin-Hsing: 1965 and 1979 ..................................................... 250 Students and Non-students by Gender, Hsin-Hsing: 1965 and 1979 ..................... 252 LIST OF FIGURES 3.1 The Theoretical Integrative Model of Migration Analysis, Derived from the Family/Household and Structural Perspectives ........................................................ 80 3.2 The Theoretical Model of Migration Analysis, Derived from Individual Perspectives .............................................................................................................. 87 3.3 How Research Hypotheses Correspond to Research Questions ............................... 90 4.1 Population Pyramids of Hsin-Hsing Village, 1965 and 1979 ................................. 147 6. 1A Estimated Probabilities of Migration Hsin-Hsing Villagers Age 15-64, 1965 ....... 233 6.1B Estimated Probabilities of Migration Hsin-Hsing Villagers Age 15-64, 1979 ....... 233 xi CHAPTER I INTRODUCTION 1 . 1 Introduction Urbanization and migration have been regarded as pressing population problems in most developing countries, even more pressing than high fertility and natural population growth rates (United Nations, 1985). It is generally believed that an excess growth of population in major cities in developing countries is more related to urbanization and ‘ migration than to other factors (Findley, 1987). Although rural-to-urban migration can stimulate economic grth and development in urban cities, it creates even more problems for the receiving areas as well as for the sending centers. The excess growth of population is accompanied by unbalanced regional development, deteriorating of urban habitats, and destruction of natural resources, and it imposes serious drains on governmental resources by heavy demands for infrastructure and housing (Findley, 1987). Rural areas can be adversely affected by this process because migration remains, by and large, selective, and it therefore draws away the more dynamic members of rural society (Oberai, Prasad, and Sardana, 1989). To solve the problems of excess growth of population in large urban cities and the continuing brain drain from rural areas in developing countries, understanding migration is crucial. Migration is one of the few mechanisms available for altering existing population distribution patterns. To this end, the concern with population re-distribution has been transformed into an interest in migration. The history of internal migration research can be traced back to 1885 and 1889 when Ravenstein published two papers on the “Laws of Migration” in the Journal of the Statistical Society. In these papers, Ravenstein listed a number of the “laws” which have remained key elements in theoretical explanations of migration that focus on the establishment of flows conditioned by a series of variables suchas distance, stages, transportation, and motives. Lee (1966), following Ravenstein, proposed the “Push-Pull Obstacles Model” to describe the causes of migration. Lee believed the decision to migrate is determined by the interaction of four dimensions: factors associated with the area of origin, those associated with the area of destination, intervening obstacles, and personal factors. Between Ravenstein and Lee, Stouffer (1944, 1960) introduced the “intervening opportunities” hypothesis proposing that “the number of persons going a given distance is directly proportional to the number of opportunities at that distance and inversely proportional to the number intervening opportunities” (Stouffer, 19402846). Stouffer (1940) believed the number of peOple out-migrating a given distance from an area is not a function of distance but rather a fimction of the spatial distribution of opportunities. From the first internal migration research done by Ravenstein at the end of the 19th century to the most current studies (e. g. Greenwood, Chalmers, and Graves, 1989; Greenwood and Hunt, 1984a, 1984b; Kitching, 1990; Sell, 1990), individual and structural perspectives have dominated migration research. Only in the past decades has the family/household perspective been applied to migration research. This dissertation is based on an integrated model which proposes that human migration is an important component of a family/household’s sustenance, survivability, and social mobility strategies to overcome societal structural constraints in rural areas. Community economic opportunities determine the extent to which the family can realize its subsistence or mobility needs locally, without migration (Guest, 1989). Therefore, community economic opportunities including accessibility to land and participation in local wage labor markets are considered to be the mechanisms through which a family/household determines and deploys its strategic responses to ensure the survivability of the kin unit and its members. This chapter begins with a brief review of migration in terms of causes, consequences, and theories. This review is followed by a brief introduction to Taiwan, in terms of industrial and economic development, family structure, and internal migration research. Then, the purpose and sociological significance of this dissertation is discussed. The final section of this chapter describes the organization of this dissertation. 1.1.1 Causes of Migration Economic determinism has dominated the study of internal migration. The overwhelming conclusion of most migration studies is that people in the Third World migrate primarily for economic reasons (Pamwell, 1993). Based upon the macro-economic perspective, spatial inequalities in expected earnings have been the dominant approach to the study of migration. The gap between rural and urban wages leads to migration from rural areas. High wage rates in the urban modern sector create high expected income returns from rural-urban migration (Harris and Todaro, 1970). The rural-urban wage differential, however, is institutionalized or politically determined, rather than market determined (Harris and Todaro, 1970; Montgomery, 1981). For example, Montgomery (1981) emphasized the rural-push side and argued that, in certain areas, agricultural markets were highly distorted by government policies. Rural incomes became artificially low, thereby stimulating rural-urban migration. A micro-economic perspective, in contrast, presumes that migration is a rational behavior; potential migrants decide to move or to stay according to the general rules of minimizing their costs and maximizing their returns (Lee, 1966; Sjaastad, 1962; Todaro, 1969, 1976, 1980). Two factors determine the potential net benefits of migration: rural- urban wage differentials, and the probability of obtaining an urban job. Potential migrants base their decision to migrate on personal human capital and accessibility to occupations in a potential destination. The net positive results of potential benefits from migration minus the potential costs would make people decide to migrate. Since rural-urban wage differentials exist, the probability of obtaining urban jobs plays a crucial role in the process of migration decision-making. In general, internal migration reflects the geographical allocation of occupation opportunities. The concentration of occupation opportunities that leads to a high demand for labor and high wages in certain areas attracts migrants.- Economic structures and systems in almost all societies, however, are changing on a daily basis. Human migration, therefore, might reflect a spatial shift in the organization of production (Frey, 1987, 1990). On the other hand, human migration could be related to a spatial shifi in the function that the affected areas perform within the organization of production (Frey, 1987, 1990). Furthermore, the restructuring of an economy is related to how newly developed urban sectors provide more occupational opportunities than places with declining economies. Following this particular point of view, internal migration is considered a demographic response to changes in the demand for labor in different economic sectors and geographic areas in a country. Migration can also be a result of non-economic factors. Gugler (1986) considers the aspect of social relations, and Massey and his associates (1987, and 1987) propose that migration is a social process. Social relations influence the decision to migrate because it is not reached in isolation (Gugler, 1986). They encourage direct moves, even over large distances (Gugler, 1986). Migration can also be influenced by environmental reasons. According to Gugler (1986), factors that bring immediate dangers, such as droughts, floods, earthquakes, cyclones, or volcanoes, as well as threatening hunger and disease, fi'equently make rural dwellers abandon their homes and seek relief in urban regions. In addition, psychological reasons have to be considered. Stark (1984) and his associates (1985 and 1988) and Rhoades (1978) emphasized social mobility in their migration studies. Aspiration for higher social status are thwarted by a lack of opportunities for advancement, particularly educational and occupational advancement, in rural areas. As a result, the decision to move, usually to an urban city, is made with the goal of enhancing opportunities for social mobility. 1.1.2 Consequences of Migration In the developing world, rural-to-urban migration dominates the migration flow. Nevertheless, migration influences both the sending and receiving centers. The receiving centers experience social and economic effects. The social effects of migration on the 5 destination areas include housing problems, urban restructuring that is derived from the increasing need for public facilities and infrastructure, and other problems such as crime (Deshpande and Arunachalam, 1981; Ayeni, 1981). Economically, the effects of migration in the receiving areas can be both positive and negative. On the positive side, migration directly causes an increase in population that provides a large enough labor force for economic development. The increase in population can increase the demand for goods and services, thereby stimulating economic growth and development in the receiving societies. Nevertheless, over-population, which has occurred in several large cities in under-developed Asian countries, has lead to high unemployment rates. In the sending centers, migration inevitably results in the periodic or permanent absence of people from their home areas. Unless the absentee is not economically active, migration also draws a potentially productive labor away from the sending area (Massey et al., 1987; Massey, 1988; Massey et al, 1993; Mines and Janvry, 1982; Pamwell, 1993). Migration may therefore lead to a reduction in a family’s ability to make the fullest use of productive resources such as land. Demographically, in most of the sending societies in developing countries, young people dominate out-migration flows, inevitably changing the age structure of the communities they leave. With their departure, old people make up a large proportion of the rural population. In addition, rural populations with high educational attainment are more likely than those with less education to migrate to urban cities in which more occupational opportunities are available for highly educated people. Therefore, migration leaves rural areas with inadequate human resources, in terms of education, for economic development. In short, rural-to-urban migration provides urban cities with an adequate labor force for economic development. Rural-to-urban migrants also increase consumption needs for industrial products in urban cities, further stimulating economic growth and development (Borts and Stein, 1964; Muth, 1968; Todaro, 1976). The increased population in urban areas also drains governmental resources to solve problems such as inadequate infrastructure and crimes, attributed to increasing migrants. In contrast, rural areas without an adequate labor force or human capital face barriers to economic development. As spatial inequalities increase, the unevenness of development between urban cities (receiving sectors) and rural areas (sending areas) becomes more disparate (Massey, et a1,1993) 1.2 Review of Migration Theories 1.2.1 The Individual Perspective The cost-benefit model of microeconomics has played a crucial role in migration research. Migration flows are the cumulative results of individual decisions based on the rational evaluation of the benefits to be gained and the costs entailed in moving (Wood, 1981). Migration is viewed as the outcome of a rational evaluation of the costs and benefits of movement (Massey, 1990a; Sjaastad, 1962; Todaro, 1976, 1980). The expected net return to migration has methodologically been used as an indicator to predict a potential migrant’s choice to move or to stay. In addition, the individual perspective of migration theory. suggests that the decision to migrate is an investment decision which involves an individual’s expectation to increase the productivity of human resources in terms of costs and returns over time (Massey, 1990a; Sjaastad, 1962; Todaro, 1976, 1980). People choose to move to where they can be more productive, given their skills. The individual cost-benefit model, therefore, is related to the computation of the expected costs and benefits of migration (see Speare, 1971). Todaro (1969, 1976) proposes that a migrant’s expected net return is a function of urban- rural expected income differentials and the likelihood of obtaining an urban job. The possibility of potentially obtaining a job in the modern urban sector is a crucial element in the decision-making process of a potential migrant; it is more important than the wage differential. Because of its emphasis on the importance of obtaining an urban job, the individual perspective turns the study of migration into a categorization of individual characteristics of migrants. This response to neoclassical microeconomic theory links human capital to the probability of obtaining a job and to the rate of remuneration. Early migration studies, therefore, focused on differences in human capital characteristics between migrants and non-migrants (Browning, 1969; Ladinsky, 1967; Long, 1973; Zachariah, 1966). In summary, this type of research is useful to demonstrate the characteristics of migrants. Nevertheless, it hardly provides a comprehensive picture of migration because people with similar human capital characteristics behave differently. To support the individual perspective, persons with similar characteristics should exhibit parallel migratory behavior. In contrast, persons who might be expected to migrate remain with their families in rural areas and people expected to remain non-mobile resort to internal migration. 1.2.2 The Structural Perspective The structural perspective of migration theory supports the individual perspective view that migration decisions are made by actors who weigh the costs and benefits of movement. Nevertheless, these theories diverge in their explanations of how decisions are made. The structural perspective suggests migration decisions are made within a specific social and economic environment that is determined by the larger structural relations in the political economy (Amin, 1974; Goldscheider, 1987, Massey, 1990a). The immediate socioeconomic context not only helps to determine parameters, such as the probability of employment and the costs of migration, but it also affects the way cost- benefit calculations are framed and conceptualized. In other words, while it may be that rational migration decisions are made to maximize expected returns, these decisions are ‘ always constrained by specific local structural conditions. Migration originates in structural change that affects the relations of production in the sending and receiving sectors. Population movement is a human behavior in response to the changing structure of the economy, and urbanization or population redistribution is a by-product of human migration. Economic development produces a pool of dislocated workers who respond to the rewards of increased productivity in developing urban economies. Cyclical economic growth in urban sectors, combined with inter-regional 9 differences in wage and cost reduction in transportation and communication, encourage emigration into the structure of economic development. Emigration assumes greater or lesser importance depending on the degree of economic connection between the sending and receiving areas. As economic integration grows, an inverse association between business cycles develops, networks of transportation and communication interlink, and labor recruitment becomes more frequent, producing’large-scale movements of labor between areas (Frey, 1987, 1990). Even if migration is stimulated by structural constraints and a changing economy, it is not likely that the structural perspective can comprehensively explain human migration. This is so because migrants do not all move toward the same destination. Most migrants do move to destinations with large numbers of employment opportunities and growing industrial and economic development. But, many migrants move elsewhere while knowing this fact. Therefore, the structural perspective does not completely explain why different receiving centers are chosen by migrants. Thus, while the structural perspective is useful in providing a broad framework for understanding the incidence of migration in relation to the development process, there is clearly also a need to show how these general macro-level processes translate into real-life situations (Pamwell, 1993). 1.2.3 Family/Household Perspective As seen, the individual and structural perspectives explain, an understanding of p0pulation movement must encompass both the broad structural societal parameters which affect behavior and the factors that motivate individual actors. In the context of rural areas, however, the unit of production and consumption is the household or the 10 family, not the individual. Consequently, an integration of individual and structural approaches can be accomplished through the analysis of household behavior as the unit interacts with its environment (Chant, 1992; Schmink, 1984; Wood, 1981). The dynamic character of household behavior can be conceptualized as a series of “sustenance strategies” by which a family actively strives to achieve a fit between its consumption necessities, the labor power at its disposal (both of which are determined by the number, age, gender, and skills of its members), and the alternatives for generating monetary and non-monetary income (Boyd, 1989; Grigg, 1980; Guest, 1989; Wood, 1981). Under conditions of structural change, the household must devise flexible and innovative strategies compatible with shifting productive opportunities. These strategies are a response to other factors that affect the sustenance of the unit. In other words, an agrarian family must provide its members opportunities for investing their labor power. The'outcome of labor invested must also meet the consumption necessities of the family. Once the balance between labor input and consumption necessities is achieved, the family does not need further sustenance strategies. Under conditions of structural change, an imbalance between these two key components are likely to occur and a family will have to seek an alternative sustenance strategy to achieve a new balance. The sustenance strategy for achieving a new balance between labor force available and consumption necessities in a family include seeking occupation opportunities in the local area as well as in other places away from the home. The family perspective on migration provides a theoretical framework to explain human migration as one strategy for family 11 sustenance. This strategy is linked to migration along four dimensions: accessibility to land and off-farm work (Grigg, 1980; Guest, 1989; Wood, 1981), risk diversification (Massey, et al., 1993; Stark, 1983, 1991; Stark and Bloom, 1985; Wood, 1981), relative deprivation and social stratification (Stark, 1984; Stark and Bloom, 1985; Stark and Taylor, 1989), and social networks (Dinerrnan, 1978; MacDonald and MacDonald, 1974; Massey et al., 1987; Massey and Espana, 1987; Massey, 1988, 1990a, 1990b; Massey et al., 1993; Mines and de Janvry, 1982; Mines, 1984; Mullan, 1989; Taylor, 1986; Tilly and Brown, 1967). The major limitation of this perSpective, however, is that it assumes migration decisions are made collectively. This perspective does not take into account the power hierarchy within which decision making occurs the family. For Chinese families, the authority for decision making usually is held by one or a few family members. The decision of migration, therefore, is hardly made by family members collectively. 1.3 About Taiwan 1.3.1 Industrial and Economic Development When Japan began its fifty-year rule of Taiwan in 1895, the island was a rural society with few settlements large enough to be considered a city. During the first three decades of colonial rule, Japan saw Taiwan mainly as a source of agricultural products. Most of its investments, therefore, went into agricultural development and construction of transportation facilities necessary to get farm products to ports so they could be shipped to Japan (Speare, Liu, and Tsay, 1988). With the end of World War H in 1945, the government of Republic of China replaced the Japanese government in Taiwan. There the government found a predominantly 12 agricultural society in which most people worked on farms and lived on the products from the land (Amsden, 1979; Ho, 1979; Lu, 1981). In one of its first moves to develop the island, the government, in 1953 enacted a land reform policy that began with the “Land-to-the-Tiller” Program. In addition, the government started to carry out the first of a series of four-year-economic development plans (Lu, 1981; Tsai, 1978). From 1953 to 1960, the government also adopted a policy of import substitution and concentrated on the domestic market (Ferdinand, 1996). The result was an average annual grth rate of 7.6 percent, that between 1961 and 1972, rose to 10.3 percent and, between 1973 and 1983, economic growth increased at an average of 12.8 percent per year (Ferdinand, 1996). To coordinate sequential economic programs, the government decided that industrial development and foreign trade expansion would be the two major directions of its efforts, and it provided many incentives to encourage industrial and trading investment, such as low interest rates and long term loans, tax reductions, transportation improvements, and the construction of infrastructure and power plants (Lu, 1981; Tsai, 1978, 1981). To support this industrial development, most of the time between the 1950s and 19905, agriculture was squeezed. The government invested far less in agricultural development in comparison to its total expenditures for industrial development (Tsai, 1978). As a consequence agricultural development was relatively slow in comparison to grth in the industrial and service sectors. Agriculture’s contribution to total gross domestic production by agriculture dramatically decreased from 32.3% in 1952 to 3.5% in 1993 13 (see Taiwan Statistical Data Book, 1994). The per capita income of agriculturists dropped much lower than that of the non-agriculturalists (Tsai, 1978). In response, the agricultural labor force increasingly migrated from rural areas to seek jobs in urban sectors between the 19505 and the early 19705 (Tsai, 1978). This rural-to- urban migration was caused by the geographical disparity of economic development; the booming industrial factories and commercial offices were geographically concentrated in cities (Tsai, 1978). The population flowing from rural villages and small towns to large urban cities, therefore, became the major stream of internal migration in Taiwan in the early stage of economic development (Speare, 1974; Tsai, 1978). Since the early 19705, industrial decentralization has contributed to population redistribution and decentralization (Liu and Tsai, 1990; Tsai, 1981; Tsai, 1990). The establishment of rural industrial zones has led to the increase in the number of employment opportunities in industries and factories for the agricultural labor force. As a result, the proportion of the labor force working in agricultural sectors dropped dramatically, decreasing from 56.1 percent of the labor force in 1952 to 11.5 percent in 1993 (see Taiwan Statistical Data Book, 1994). 1.3.2 Family The family has been described as the basic unit in Taiwanese society.1 The term “family” in Taiwanese or Chinese society usually refers to a unit consisting of members related to The family often coincides with the household, but the two terms are not identical. The family is an enduring kinship unit. By contrast, the household is a less permanent residential unit made up of any family members who happen to live together at a given time (Greenhalgh, 1990). 14 each other by blood, marriage, or adoption (Cohen, 1976; Lang, 1946). In general, families in Taiwan can be clustered into three types: conjugal, stem, and joint.2 The life span of a family is cyclical; a family can grow and become large and/or it can die. During the process of family growth, a large family can be divided into several small new families by family partition. Family, in Taiwanese society, is also a basic economic unit, in which members share a common estate and budget. Family division is thus an elaborate legal process in which many types of family obligations are terminated or re-defined, and family property is either equally or unequally divided (Cohen, 1976). Family land (including owned and tenanted), businesses (if any), livestock, tools, buildings and residence are all divided among the male heirs. Family partition thus inevitably leads to new divisions of labor. Economic obligations and responsibilities among family members have to be re-defined with changes in family composition. The economic behavior of a family can be conceptualized as a series of “sustenance strategies” by which its members collectively strive to achieve a fit between its consumption necessities and available labor power. Structural constraints and changes condition human behaviors. In postwar Taiwan, agricultural land was the major means of production. But as the Taiwanese government moved from an agricultural-based to an The conjugal family consists of a husband, a wife, and their unmarried children; the joint family adds two or more married sons and their wives and children to this core group. The stem family - a form that lies somewhere between the conjugal and joint family types - includes parents, their unmarried offspring, and one married son with his wife and children (see Lang, 1946). 15 industrial-based policy, agricultural land became less valuable than it had been in two ways. First, the value of outcomes/products of agricultural land became less than that of industrial products. Second, the average size of land per person became smaller while the rural population increased. These phenomena had two consequences: agricultural land could not support sufficient consumption needs, and family land became too small to accommodate all of the family members involved in the labor force. Because the value of agricultural products became less than that of industrial products and the average size of land per person decreased, family income from farming became insufficient to meet the needs of family consumption. Seeking additional financial resources became increasingly more important as the economic structure changed. Family members found occupational opportunities in their local areas or migrated to urban areas to find additional resources for family income. Allocating the human capital ’of family members is a family sustenance strategy to maximize familial collective interests, although family members could be made to migrate by a single member with disproportionate power or authority within a family. Deploying family members to seek job opportunities in major urban cities was the most p0pu1ar strategy adopted for increasing financial resources. As rural-to-urban migration led to the over-urbanization of certain cities, the attention of migration researchers was attracted. 1.3.3 Overview of Internal Migration Research in Taiwan 1.3.3.1 The Individual Perspective Internal migration research in Taiwan has focused on the demographic characteristics of migrants, such as gender, age, education, and the like (Chang, 1979; Chiang, 1978; Li, 16 1974; Liao, 1977; Liu, 1993; Speare, 1974; Tsai, 1978; Yin, 1978). Most Taiwanese migration studies indicate that migration rates are highest for young adults (Chang, 1979; Chiang, 1978; Tsai, 1978; Yin, 1978). Further, Chang (1979) suggested that men and women had different migration patterns in terms of age in the early 19705. Besides gender and age differentials, researchers also compared the educational attainment between migrants and non-migrants. These various migration studies yielded inconsistent results. Studies which focused on the relationship between migration and education did not clearly identify which level of education were held by persons more likely to migrate (see Chang, 1978; Chiang, 1978; Speare, 1974; Tsai, 1978; Yin, 1978). Occupation was another migrant characteristic examined in Taiwanese research (see Chang, 1978; Chiang, 1978; Yin, 1978). Nevertheless, research that focused on migrant’s occupation produced inconsistent results. In summary, migration research in Taiwan that was based only on the individual perspective demonstrated the characteristics of migrants, on the one hand. On the other hand, this type of migration research had three shortcomings. First, researchers generated diverse conclusions based upon their different samples. Second, their research only demonstrated the types of migrants who dominated the migration flows, not the reasons why people migrated. Third, they ignored the importance of the broader structural environment that is related to the individual behavior of migration. This individual perspective of migration research thus does not provide a comprehensive view of the process of migration. 17 1.3.3.2 The Structural Perspective Another major approach adopted to understand migration in Taiwan has been the structural perspective. This approach suggests that internal migration in Taiwan is a response to the geographical unevenness of industrial and commercial development. Between the late 19505 and the early 19705, the most popular destinations were two major cities: Taipei and Kaohsiung (Tsai, 197 8). Since the early 19705, industrial decentralization contributed to population redistribution and decentralization (Liu and Tsai, 1990; Tsai, 1981; Tsai, 1990). The establishment of rural industrial zones led to the increase in the number of employment opportunities in industries and factories for the agricultural labor force however (Tsai, 1981). Migration studies using the structural perspective have demonstrated that urban centers with more employment opportunities attract more labor migrants than those areas with a lack of job opportunities. Research using this framework has identified new migratory destinations (see Liu and Tsai, 1990; Tsai, 1981; Tsai, 1990). This approach, however, has over-emphasized the importance of economic factors. While it is useful in providing a broad framework for understanding the incidence of migration in relation to industrial and economic development processes, it ignores the social dimensions that contribute to migration. 1.3.3.3 F amily/I-Iousehold Perspective Both the individual and structural perspectives of migration studies fail to provide a comprehensive framework for understanding internal migration in Taiwan. This dissertation proposes an integrative model that is derived from the family perspective. 18 The family perspective never played a major role in migration studies in Taiwan. Nevertheless, the family perspective of migration is more appropriate than the individual and structural approaches for research on Taiwanese migration. First, the family is the fundamental and most important social unit in Taiwanese society. Second, the family in Taiwanese society is also a basic economic unit. Family members usually share a common estate and a common budget (Cohen, 1976) and act “collectively” to cope with the problems the unit faces. The family perspective thus maintains that human migration is a response adopted to ensure the survivability of a kin unit and its members. Migration of family members occurs because of an imbalance between family consumption needs and the labor force available. In general, the family/household perspective maintains that families deploy members into migratory streams as a strategic response to structural constraints. Essentially, families face life with a fixed short-term set of resources and a set of basic consumption and reproduction needs. The former include land, labor, and capital, and the latter include the family’s age-gender composition and its social and economic aspirations. Household resources are combined productively to meet the requirements of family maintenance and mobility, and migration can be a very effective way of capitalizing on the labor power a family has available. A family’s behavior in allocating workers to different productive pursuits may be viewed as a series of dynamic and flexible strategies that shift as needs and economic conditions change. 19 1.4 Purposes and Significance of the Dissertation 1.4.1 Purpose of the Dissertation Three theoretical frameworks have dominated migration research -- the individual, the structural, and the family/household perspectives. In Taiwan those most frequently adopted have been the individual and the structural perspectives to the study. The family perspective has never been used for migration research in Taiwan. The research results of studies adopting the individual and structural perspectives have produced either inconsistent conclusions or an incomplete picture of migration. Studies based on the individual perspective yielded inconsistent results. Studies based on the structural perspective did not consider individual behavior as a mitigating factor in migration decisions. In addition, they also completely ignored the importance of the social dimensions of migration. In this dissertation, migration will be approached as a family sustenance strategy that deploys individuals on a selective basis to overcome the structural constraints of a changing economic system. The first basic analytic unit will be family. The contributing factors to migration behavior will include family accessibility to local labor markets, including land and local wage labor markets. In addition, I will move beyond the traditional “cost-benefit” argument of the individual perspective and consider how the migration process is affected by family power dynamics, as they are shaped by the intersection of gender and age. I will discuss male- female power relations and power relations among female villagers. My analysis encourages an expansion of the individual perspective of migration, showing how personal characteristics are implicated in the decision-making process of migration. 20 1.4.2 Significance of the Dissertation This dissertation will expand the body of knowledge on internal migration in the theoretical way. Its theoretical approach is different from other Taiwanese migration studies. Internal migration was not a major issue in Taiwan until the over-urbanization of a few major cities was recognized in the 19705. Research on internal migration in Taiwan has been overly focused on the individual and the structural. Specifically, previous research can be classified into two groups: (1) migration selectivity and the characteristics of migrants (see Chang, 1979; Chiang, 1978; Li, 1974; Liao, 1977; Liu, 1993; Speare, 1974; Tsai, 1978; Yin, 1978), and (2) interrelationships between internal migration, economic development, and the process of urbanization (see Chang, 1984; Liu, 1982, 1983; Liu and Tsai, 1990; Sun and Tsai, 1981; Speare, 1974; Tsai, 1978, 1979, 1981; Tsai, 1990). Internal migration in Taiwan, in short, was attributed to specific geographical patterns of industrial and economic development. Urbanization and population concentration were strongly related to industrial and economic development. These studies do not explain how family power dynamics, which were shaped by certain personal characteristics such as gender and age, affected the decision-making of migration. This research moves beyond the neo-economical approach of “cost-benefit” to discuss how family power dynamics affected the migration process of rural villagers, and different employment status between men and women, and among female villagers. 1.5 Organization of the Dissertation This dissertation consists of seven chapters. Chapter 11 includes two major sections: a contextual introduction to Taiwan and a review of migration research in Taiwan. In the 21 first section a brief history of economic development in Taiwan is provided and family is discussed. This is followed, in the second section, by a review of internal migration research on Taiwan over the past few decades. Chapter IH describes the research framework. The theoretical framework includes the conceptual framework and the specific hypotheses that were developed from the research diagram. In addition, this chapter introduces the data and research methods. The data section describes the data sources, how data are organized into the database for analyses, and the studiedgvariables. Finally, statistical methods applied to this research are discussed. The main purpose of Chapter IV is to provide a profile for the research area -- Hsin- Hsing Village, Taiwan. This chapter describes the geographical location of the research area and discusses its demographic change and economic development. Specifically, in this chapter there are five topics introduced, including the village’s climate and spatial layout, family structure, and economic, demographic and socio-economic infrastructure. Chapter V focuses on how contributing factors relate to family decisions to migrate. Statistical techniques are applied in this chapter to examine the associations among labor migration, accessibility to land and accessibility to local labor market to answer the questions such as (1) whether labor migration was related to family type, (2) whether labor migration was related to a family’s accessibility to land, and (3) whether labor 22 migration was related to villagers’ local labor market participation. The analyses will be at the family level. In Chapter VI, I will examine the associations between labor migration and family power dynamics. The analyses will be at the individual level. These family power dynamics are analyzed in terms of (1) male-female relations and in terms of (2) relations among women In the final chapter, I summarize the analysis and discuss the theoretical and substantive contributions of the dissertation. In addition, I discuss the limitations of this dissertation, and the next logical research and policy steps. 23 CHAPTER II EARLY SETTLEMENT, ECONOMIC DEVELOPMENT, AND MIGRATION The main purpose of this chapter is to provide an introduction to Taiwan to lead to an understanding of its internal migration. Therefore, this chapter focuses on early settlement patterns, the economy, population distribution, and how family relates to internal migration. I briefly review Taiwanese immigration history prior to the end of 19th century. Second, the history of Taiwanese economy is discussed, separately during the period of the Japanese occupation and during the postwar Taiwan. This chapter especially focuses on the postwar Taiwanese economy with special emphasis on the twin processes of agricultural and industrial development. Third, this chapter presents a brief review of the interaction of economic development and migration in Taiwan. In addition, an assessment of the interaction of migration decisions and power dynamics within families will be demonstrated. Finally, the chapter reviews what consequences of migration are in Taiwan. 2.1 The Early Settlement Patterns in Taiwan -- Before 1895 The original indigenous peoples inhabiting Taiwan were the Proto-Malay3 (Common Wealth Magazine, 1991; The Government of Formosa, 1926; Shih, 1980). The Hakka and Hokkienese have been the major ethnic groups since the large waves of Chinese 3 Officially, there are nine indigenous groups: 1. Atayal; 2. Saisiyat; 3. Bunun; 4. Tsou; 5. Rukai; 6. Paiwan; 7. Puyuma; 8. Ami; and 9. Yami.‘ 24 immigration towards the end of the 17th century. Although the Hakkas migrated to Taiwan earlier than the Hokkienese, since the end of the 17th century the Hokkien - speaking immigrants have outnumbered previous Hakka settlers. Hakka immigration to Taiwan can be seen on the island as early as the 7th century (Kiang, 1991). The Hakka continued settling on the island in small communities up through the 13th century. The number of Hakka immigrants increased dramatically afier the 13th century and reached a climax in the 17th century (Common Wealth Magazine, 1991; Shih, 1980). The Hokkien speaking immigrants who had lived in the Southern Fukienese began to settle in Southwestern Taiwan as early as the 15905. The Hokkienese were the majority among Chinese immigrants toward Taiwan in the 17th century. Especially during the period 1662-1683, when Koxinga ruled Southern Taiwan, thousands of depressed peasants in Fukien flocked to Taiwan to find a new life. These new Hokkien immigrants soon outnumbered the Hakkas. These Hokkien newcomers displaced the earlier inhabitants - the indigenous, Proto-Malay and the earlier Hakka migrants. Soon after the surrender of the Koxinga Kingdom, the Manchu Ching dynasty promulgated a ban on out—migration to Taiwan from mainland China, specifically prohibiting the Hakka from crossing the sea to Taiwan (see Appendix 1). After approximately two hundred years of migration prohibition, all migration restrictions were eliminated in 1875. The Chinese were allowed to migrate to Taiwan freely. As a 25 consequence of banning Hakka immigration, the Hokkienese dominated the migration flows, and became the majority in today’s Taiwan. It is worth noting that early Hakka and Hokkienese immigrants landed on the southwestern Taiwan and settled on the western Taiwanese Plain. Early settlement penetrated by the geography has led to east—west uneven development; western Taiwan has dominated the economic and industrial development and received more attention than the. eastern part. In terms of population, most people reside in the western Taiwan. 2.2 Early Taiwanese Economy 2.2.1 The Taiwanese Economy during Japanese Occupation (1895-1945) Before the arrival of Chinese immigrants, the Taiwanese aborigines lived as hunter- gatherers. Hakka and Hokkien immigrants created rice and sugarcane fields, encroaching on the aborigines’ traditional hunting grounds (Ka, 1995: 1). By the eve of Japanese colonial rule in 1895, Chinese settlers had occupied the plains and established a booming agrarian economy centered on rice and sugarcane (Ka, 199521). In the course of half a century of colonial rule, Taiwan and Japan developed an unequal relationship in terms of division of labor (Ho, 1978229; Ka, 1995:1). Taiwan was developed to satisfy Japan’s economic needs. In Ho’s words, Taiwan became “an agricultural appendage of Japan, to help it feed its growing industrial population” (Ho, 1978:29). Taiwan, in'particular, was transformed into a food and raw material supplier in a new division of labor with Japan. On the one hand, the Taiwanese non-agricultural sector accepted/received Japanese manufactured goods (Ho, 1978:29). On the other hand, Taiwan exported sugar and rice to 26 Japan, which were the primary products exported from the Taiwanese agricultural sector (Ho, 1978229; Ka, 1995:l). The combination of rice and sugar accounted for 50-70 percent of Taiwan’s total exports between 1900 and 1939 (Ho, 1978230-1). According to Ho (1978:31), In every year, over 90 percent of Taiwan’s sugar output was exported, and in the 19305 about half of Taiwan’s rice output was also exported. Nearly all Taiwan’s rice and sugar exports went to Japan. Before World War II, rice and sugar comprised approximately 15 percent of Japan’s total imports, and Taiwan’s contribution to this was substantial. It was Japan’s major supplier of sugar; in the 19305 it provided nearly 75 percent of the sugar consumed in Japan. Before the sharp rise in agricultural productivity in the 19205, approximately 20 percent of Taiwan’s rice harvest was exported each year to Japan. Thereafier, as rice production per capita increased in Taiwan, more and more rice was exported. By the 19305, approximately, 45 percent of Taiwan’s annual rice harvest was exported to Japan, accounting for over 30 percent of Japan’s import requirements. During the Japanese occupation, the colonial government endeavored to apply scientific knowledge and modern techniques to the agricultural development of Taiwan (Ho, 1978258). For example, the Japanese government on Taiwan successfully introduced “seeds with higher yields, greater resistance to disease and high wind, and more receptivity to fertilizer and intensive care” (Ho, 1978:58). However, “it is impossible for technology to transform agriculture by itself without extensive restructuring of agricultural institutions” (Ho, 1978265). Among the agricultural institutions, the F arrners’ Association, which was first organized in Taiwan around 1900, played a significant role in the introduction of new agricultural techniques and scientific farming to Taiwan (Ho, 27 1978163). In addition, Japanese agricultural personnel provided the support to the agricultural transformation in Taiwan (Ho, 1978265). In the Japanese colonial period, a large proportion of Taiwanese population was living in agriculture. Land was the major means of production. The average size of land per farm household was relatively large and stable during the first forty years in the 20th century. Statistics show that between 1910 and 1940, the average size of land per farm household remained at around 1.97 hectares (JCRR, 1956:7-9). Nevertheless, the distribution of ownership of land was unequal (Ho, 1978:42). Survey data from 1920 and 1939 demonstrated out that about one-half of agricultural households had less than 1 chia of land.4 A large proportion of poor farmers held a very small amount of land. Specifically, in 1920, only 5.7 percent of land was held by the lower 42.7 percent of landowners (Ho, 1978:42 and 349-50). By contrast, 62.1 percent of land was held by the upper 11.5 percent of owners (H0, 197 8242). The 1939 survey revealed that the uneven distribution of land ownership was more serious; “64 percent of the landowners held less than 1 chia of land, and only 655 landowners possessed more than 50 chia (120 acres)” (Ho, 1978:43). ‘ In Ho’s (1978:42) words, approximately 25 percent of the farms had less than 0.5 chia (1.2 acres) of land; 20 percent between 0.5 and 1 chia of land; 39 percent between 1 and 3 chia of land, and 16 percent more than 3 chia of land. . Note that: l chia of land equals 0.9699 hectare. 28 Those who owned a great amount of land did not necessarily farm all of their land. Those who held small pieces of land or no land might have rented some land in. In the case that most farmers held small pieces of land, during the period of 1920 and 1939, most agricultural population were tenants than other types of farmers (Ho, 1978:43). Ho (1978) pointed out that “in 1910, 33.7 percent of the agricultural population were owner- cultivators, 42.8 percent were tenants, and 23.5 percent were part-owners and part- tenants” (Ho, 1978:43). By 1941, the proportion of tenants among the agricultural population was reduced to 37.4 percent (Ho, 1978243). The proportions of landowners and those who were part-owners and part-tenants changed to 30.4 percent and 32.2 percent of the total agricultural population, respectively (Ho, 1978:43). Based on several surveys between 1920 and 1940, Ho showed that “land cultivated by tenants as a percentage of total cultivated area remained fairly stable at around 57 percent” (Ho, 1978; 43). Unequal distribution of land ownership had caused an unfair land tax system. To establish new land tax policy that attempted to properly tax legal landowners, the Japanese government changed the traditional tenure arrangement from the three-level tenancy (composed of ta-tsu-hu, hsiao-tsu-hu, and subtenants) system to a two-level tenancy (composed of hsiao-tsu-hu and subtenants) system.5 The Japanese government ’ According to Ho (1978: 12-3), continue to the next page... 29 started the reform in 1904. In 1905, it “brought out the ta-tsu-hu and made the hsiao-tsu- hit the legal owners of the land and directly responsible for the land tax” (Ho, 1978:44). The elimination of the ta-tsu-hu transferred income streams from those who were neither involved nor interested in agriculture to those who had a direct stake in agriculture and were therefore more likely to use the resOurces productively (Ho, 1978244). But the change of the land tenancy system in the early 20th century by the Japanese government did not prevent. the development of inequality between tenants and landlords. 2.2.2 The Political Economy of Contemporary Taiwan (After 1945) ' The Nationalist government arrived in Taiwan in 1949 as an “outsider” with no ties or commitments to the established local elites. Therefore, to establish political stability and prevent the reocCurrence of defeat by communists, the Nationalist government felt a strong need to establish a solid political base among the Taiwanese peasants. Under these circumstances, a redistribution of land ownership was adopted, which made most tenants become landowners. The first land reform in the late 19405 and early 19505 was aimed at Land was settled during the Ching dynasty under several arrangements, which in later years determined the tenure system. Some plots of land were cleared by individual farmers who held the land in private ownership and some were rented from pacified aborigines and cleared by individual settlers who then became the tenants. However, most were settled under the sponsorship of wealthy Chinese or that of the government. Land settlement required labor, capital, and protection from the unfriendly aborigines. The immigrants were able to provide labor, but for capital and protection they had to turn to the wealthy Chinese and the government for help and sponsorship. In return for the sponsors’ capital (such as tools, draft animals, weapons) and protection (which sometimes meant costly punitive expeditions against the aborigines), the reclaimed land became the property of the sponsor, but the settlers had the perpetual right to work the land at a fixed rent. Regardless of the details of the initial arrangement, the sponsor of land settlement was called ta-tsu-hu and the settlers who cleared the land were called hsiao-tsu-hu. As more immigrants arrived from the mainland, the hsiao-tsu-hu leased a part or all of their holdings to the newcomers and also became landloards. In this fastion a three-level tenancy system evolved: the cultivators (the subtenants), the hsiao-tsu-hu (the tenant-landlords), and the ta-tsu-hu (the great landlords). This complex system remained until it was revised by the Japanese in 1904. 30 guaranteeing political stability through the achievement of the social and economic security of tenant families (Greenhalgh, 1990). 2.2.2.1 ‘ Land Reform and Land Ownership Land reform policies began in 1949. Based on the first land reform regulation, rents were limited to 37.5 percent of the annual production of main crops (Chen, 1961; Chen, 1994; Greenhalgh, 1990; Ho, 19781160; Ho, 1987:234; Lu, 1981). Two years later in 1951 the Taiwanese government took the first step to transfer ownership of farmland taken over from the Japanese to families which had actually tenanted it. More than 50 percent of the farmland owned by the “government was eventually affected by this policy” (Ho, 19782161). This policy legislated that “the price of the public land was 2.5 times the annual yield of principal crops and was to be paid in 20 installments over a period of 1.0 years” (Ho, 19782161; Ho, 1987:234). Finally, the most important land reform policy in Taiwan was carried out in 1953, which was the land-to-the-tiller program (Ho, 1978; Ho, 19872234; Lu, 1981; Tsai, 1978). This policy destroyed the power of the landlords, which the government compulsorily purchased all land in excess of 3 chia of medium quality at a low cost (Chen, 1961; Greenhalgh, 1990; Ho, 19782162). The land later was redistributed to tenants who paid at the price of 2.5 times of the annual yield (Ho, 1978:162-163). As a consequence of the land reform policies, the ownership of land in Taiwan was significantly redistributed. A great number of tenants became landowners and most of farmland became cultivated by owners. As Ho (19782164) indicates, In 1948 tenants farmed 44 percent of the total cultivated area; by 1953 the percentage of tenant cultivated land had decreased to 17 percent. Almost 50% of Taiwan’s farm households, or about 75 percent of tenant and part- 31 tenant farm households, were able to purchase some land. The percentage of tenant farm households among total farm households was 41 percent in 1947, 21 percent in 1953, and 10 percent in 1970. Owner-cultivators households as a percentage of total farm households significantly increased from 32 percent in 1947, 55 percent in 1953, and 78 percent in 1970. The Nationalist land reform policies reduced the size of the landless peasant class and created a more equalized income distribution in Taiwanese agricultural sector. Nevertheless, the operational sizes of farms could not increased by the adoption of land reform policies. First, it was almost impossible to increase any amount of arable land on the island. As Ho (1978) states, “[b]y the 19405, nearly all the available land economically suitable for farming was being cultivated” (p. 147). From 1952 to 1973, cultivated land area increased by less than 3 percent. Second, the decline in the average farm size per household was reinforced by the steady growth of farm population in the postwar period (Ho, 1978:147). More specifically, from 1946 to 1950, more than 1 million mainland refugees (including military and civilian) arrived in Taiwan. In the 19505, the rate of natural population increase was also high; about 3.4 percent per year (Ho, 19782156). During the 19605, although the rate of natural increase in rural areas fell, the continuous rise in population put a significant pressure on the un-growing and limited land resources (Ho, 1978:156). The average size of farm decreased. In fact, [s]ince the colonial period the size of the average farm has more than halved, decreasing from about 2 hectares to less than 1 hectare of land. In 1939 about 25 percent of the Taiwanese farms had less than 0.5 hectare of land but by 1960 the farms in this category had climbed to 37 percent (Ho, 1978: 1 56). 32 In sum, the series of land reform policies significantly contributed to the redistribution of the land ownership. Nevertheless, the stable size of farming land and increased population and caused to the reduction of the average size of farm. 2.2.2.2 Agricultural Development With the sudden increase of the population in the late 19405, the need for food was obvious and urgent. In the early postwar period, the importance of agriculture was apparent. Its functions were twofold. It could not only produce food for an increasing population, but also provide opportunities for labor input. The redistribution of land ownership afier the land reform did provide labor input opportunities for more people. Nevertheless, how to increase agricultural production on the constant cultivated areas, in terms of size, to face the increased population pressure became an important issue. Farmers responded to the increased p0pu1ation pressure and the decreasing farm size by adopting labor-intensive production techniques, which allowed more crops to be grown in a single year. Farmers adopted new cropping patterns and inter-cropping (planting a second crop between rows of the first crop before the latter is harvested). The new cropping patterns made “the total crop area increase about 40% from 1.2 million hectares in 1940 to 1.7 million hectares in the late 19605” (Ho, 19782150) In addition, the government applied new technology to agricultural production to respond to the crisis of rapid increased agricultural consumption. For example, the Taiwanese 33 government encouraged farmers replacing animal with mechanical power; applying power tillers.6 Nevertheless, the utilization of power tillers diminished the importance of animal manure as a source of fertilizer. It led to the dependence of agricultural production on chemical fertilizers. In the 19705, the government kept prompting farm mechanization with a view to solving the labor shortage problem in rural areas (Lu, 1981 :8).7 Although new agricultural technology and cropping methods instantly increased agricultural production, over time, agricultural development was extremely slow. In the early stage of the postWar period, the Taiwanese government totally controlled the distribution of chemical fertilizer; it was “the sole source of chemical fertilizer in Taiwan” (Ho, 1978:180). However, Taiwan did not produce chemical fertilizers, which were exclusively imported by the government, and then distributed by two governmental organizations (Ho, 1978:181). These two organizations -- Taiwan Sugar Corporation and Farmers’ Association -- distributed chemical fertilizers according to crops (Ho, 19782181). Most chemical fertilizer was distributed to rice farmers. Ho (1978) reported Since the implementation of land reform, most farmers owned land, but land holdings was fragmented. Lands of a same owner were scattered at several places. Each piece of land was small in size. Furthermore, irrigation was not easy accessible. To solve these problems, the Taiwanese government launched the rural land consolidation project in the early 19605. In the 19705, the rural land consolidation project became a part of the plan for rural reconstruction. As a consequence, “[flragmental plots have been transformed into rectangular shapes, offering an easy access to farming facilities, such as irrigation...” (unknown, 1977250). Also, “this new farming structure provided a chance for the use of agricultural machines and the application of new techniques” (unknown, 1977:52), which enhance the likelihood of migration. As Ho (19782159) mentions, by the late 19605, migration and part-time off-farm jobs had pretty much depleted the pool of surplus labor in rural Taiwan. The need for labor-saving equipment became obvious, and in 1970 the government drafted a program to promote the adoption of farm machinery (garden tractors, rice transplanters, and harvesters). 34 that “about 70-80 percent of the fertilizer is allocated to rice and is distributed to rice farmers through a rice-fertilizer barter system” (p. 181). Nevertheless, the rice-fertilizer barter program had serious disincentive effects on agricultural development. In addition, agriculture was squeezed to support industrial development. To promote the industry of chemical fertilizer, the government artificially ' kept the price of fertilizer high. In the early years of the postwar period, fertilizers utilized in Taiwan were imported. Taiwanese government began in the mid-19505 to produce “sizable quantities of fertilizers at costs substantially higher than imported fertilizers” (Ho, 19782181). The higher price made the domestically produced fertilizer undesirable.8 The high costs of fertilizers reduced the stimulation to farmers to expand production through the application of more fertilizer. In fact, farmers also hesitated to apply high yield seeds which required intensive fertilization, because of the high fertilizer Costs (Ho, 19782183). Obviously, agricultural development slowed. Besides the uneven development policies between industry and agriculture, there were other policies squeezing and slowing down the agricultural development and leading to increasing income differentials between the nonagricultural and agricultural sectors. For example, the government operated certain programs to control a sizable share of the rice crop produced each year. As Ho (1978:180) argues: According to the government policies, farmers were often required to purchase fertilizers they did not need (Ho, 1978). 35 Because peasants were not fully compensated for the rice collected by the government, these programs were in effect extractive instruments. The main methods used by the government to collect rice [were] taxation, compulsory purchase of rice at prices substantially below the market price, and the bartering of fertilizers for rice at ratios stipulated by the government... Although the official purchase price increased steadily in the 19505 and 19605, it nevertheless remained consistently 25-30% below the wholesale market price of paddy. The difference between the two prices [was] of course essentially a tax on the cultivators. .. [In terms of land tax,] in 1946 landowners paid 11.5 kg of paddy rice for every yen (dollar) of land tax owed to the government; the rate significantly increased to 27 kg per yen by the late 19605. In general, agriculture received less attention from the Taiwanese government, compared to industry and services. The government investment in agricultural development was small in comparison to its total expenditures for industrial development (Tsai, 1978). Consequently, agricultural growth was relatively slower than industry and services. Based on the governmental records, Table 2.1 shows the growth of agricultural, industrial, and service sectors from the early 19505. Gross domestic production of agriculture increased from NT$ 55,558 million in 1952 to NT$ 114,556 million in 1980, and to NT$ 197,794 million in 1993. Nevertheless, the contribution of agriculture to the total gross domestic production dramatically decreased from 32.3 percent in 1952 to 7.7 percent in 1980, and to only 3.5 percent in 1993 (see Table 2.1). On the other hand, the foreign trade of agricultural production dramatically “changed from a yearly average surplus of $54 million [in 1960-64] to a deficit of $1.6 billion in [1985-89]” (Huang, 1993:44).9 According to Huang (1993), agricultural trade in 1975-79 was a deficit of $609 million and decreased to $1478 million in 1980-84. 36 “-0— c-...-«— ”#P Table 2.1 Sources of Gross Domestic Product in Taiwan, 1952-1993 Agriculture Industry Services Total Year GDP % GDP % GDP % GDP 1952 5,558 32.2 3,396 19.7 8,297 48.1 17,251 1955 8,720 29.1 6,966 23.2 14,295 47.7 29,981 1960 17,838 28.5 16,796 26.9 27,873 44.6 62,507 1965 26,611 23.6 34,025 30.2 51,991 46.2 1 12,627 1966 28,379 22.5 38,494 30.5 59,149 46.9 126,022 1967 30,057 20.6 48,053 33.0 67,707 46.4 145,817 1968 32,308 19.0 58,524 34.4 79,072 46.5 169,904 1969 31,276 15.9 72,565 36.9 93,004 47.2 196,845 1970 35,076 15.5 83,530 36.8 108,199 47.7 226,805 1971 34,455 13.1 102,680 38.9 126,541 48.0 263,676 1972 38,619 12.2 131,670 41.6 145,883 46.1 316,172 1973 49,678 12.1 179,893 43.8 180,834 44.1 410,405 1974 68,279 12.4 223,609 40.7 257,689 46.9 549,577 1975 74,875 12.7 235,419 39.9 279,357 47.4 589,651 1976 80,504 11.4 305,443 43.2 321,763 45.5 707,710 1977 87,875 10.6 364,393 44.0 376,727 45.4 828,995 1978 93,033 9.4 448,007 45.2 450,562 45 .4 991,602 1979 102,248 8.6 542,210 45.3 551,380 46.1 1,195,838 1980 114,556 7.7 682,1 14 45.7 694,389 46.6 1,491,059 1981 129,487 7.3 807,242 45.5 837,202 47.2 1,773,931 1982 147,016 7.7 843,022 44.4 909,933 47.9 , 1,899,971 1983 153,289 7.3 944,691 45.0 1,002,025 47.7 2,100,005 1984 148,351 6.3 1,081,913 46.2 1,112,814 47.5 2,343,078 1985 142,999 5.8 1,144,824 46.3 1,185,963 47.9 2,473,786 1986 158,224 5.5 1,360,196 47.6 1,336,760 46.8 2,855,180 1987 171,234 5.3 1,528,714 47.4 1,523,045 47.3 3,222,993 1988 175,624 5.0 1,597,457 45.7 1,723,870 49.3 3,496,951 1989 189,567 4.9 1,690,913 43.6 1,998,067 51.5 3,878,547 1990 174,242 4.1 1,795,742 42.5 2,252,020 53 .3 4,222,004 1991 173,927 3.7 1,997,973 42.5 ' 2,532,237 53.8 4,704,137 1992 183,162 3.5 2,153,799 41.4 2,861,544 55.0 5,198,505 1993 197,794 3.5 2,320,874 40.6 3,193,851 55.9 5,712,519 Source: Council for Economic Planning and Development, Republic of China, Taiwan Statistical Data Book (1994). Unit: NTS million In sum, the Japanese government could not improve the uneven distribution of land ownership through the change of tenancy in 1904. To establish its regime and political stability, the Nationalist government adopted a series of land reform policies in the late 37 19408 and early 19505 after immigrating to Taiwan. Although these land reform policies redistributed land ownership, they could not increase the area of arable land. To solve the food problems caused by the rapid increase of population, the Nationalist government introduced new labor-intensive production techniques, new agricultural equipment, and chemical fertilization. Nevertheless, economic policy favored industrial development and squeezed agricultural development. Farmers paid a high price for fertilizers. Also, land taxes paid on paddy rice significantly increased from 1946 to the late 19603, which led to the great gap of income differentials between agricultural and nonagricultural sectors. 2.2.3 Industrial Development and Policy in Taiwan The rapid Taiwanese economic development in the postwar period has been described as an economic miracle. In fact, while the importance of agriculture in the economic development was declining, industrial development was growing in importance and received more and more attention from the government. The industrial development started in the early 19508. From the early 1950s to the late 1970s, the industry was developed along with the economic development that can be divided into three major stages: import substitution (1952-1960), export substitution (1961-1973), and technology- intensive (after 1974) (Lu, 1981). Starting in the early 19503, the Taiwanese government “followed an inward-looking development strategy based on import substitution in manufacturing” (Ho, 1978:186). In 1952, the government launched its first four-year economic plan and started the second one in 1956 (Lu, 1981). During this period, the strategies for economic development could be summarized by a government slogan: “Developing agriculture by virtue of 38 industry and fostering industry by virtue of agriculture” (see Ho, 1978:105). Agriculture was the base of the Taiwanese economy and industrial development depended on agricultural production. The primary economic policy was, therefore, to process agricultural commodities and export agricultural products (Lin, 1995: 184). Meanwhile, the government allowed imported manufactured commodities (e. g., cotton yarn, cotton fabric) for industrial production and industrial products (e. g., bicycles, electric bulbs). By importing commodities from foreign countries, Taiwan learned and developed new industrial technology. In the 19605, economic development strategies changed to “Developing agriculture by virtue of industry, and fostering industry by virtue of foreign trade” (see Ho, 1978: 106). The government’s economic development policies were shifted to export substitution in which the production of industry was not only for domestic consumption, but also for exporting. Nevertheless, agricultural products still played a major role in the foreign trade in the early stage of this period. As Ho (1978) notes, “in 1964, at the beginning of the outward-looking phase of Taiwan’s industrialization, agricultural and processed agricultural products comprised nearly 60 percent of Taiwan’s exports” (p. 210). Later ”Taiwan shifted from producing and exporting commodities with high natural resource content to producing and exporting commodities with high labor and skill content” (Ho, 1978:211). Two export-processing zones were established in rural areas, which were outside of two major cities in the central and south Taiwan in the same period. With the development of labor-intensive industries, the manufacturing sector was able to absorb a great number of workers including both males and females. Meanwhile, a large 39 proportion (about 83 percent) of female workers concentrated in four industries: food, chemicals and chemical products, textiles, and electrical equipment (Ho, 1978:211). Taiwanese economy was not growing constantly. In 1973 -- the year of the first oil crisis in the world -- oil price increased dramatically in Taiwan. In the following two years, Taiwan experienced the economic recession and the foreign trade including both exports and imports significantly shrank (Lu, 1981 :6-7). Up to this time, most of exported items were labor-intensive products such as textiles and plastic commodities. In 1976 the government launched a new strategy for economic development. It decided to “replace the nation’s labor-intensive products with high-technology goods so as to better fight protectionism in the overseas markets” (Lu, 1981 :7). Meanwhile, the pressure of inadequate infrastructure was recognized. In the 19703, the government, therefore, launched ten major national constructional projects for improving transportation and communication, and necessary infrastructure (e. g., nuclear power plant, steel mill, petrochemical complex, and shipbuilding yard) for industrialization and economic development. Besides the construction of infrastructure and power plants for industrial development, the government provided many other incentives to encourage industrial and trading investment, such as low interest rates and long term loans, tax reductions (Lu, 1981; Tsai, 1978; 1981). Compared to agricultural development, industrial development received much more attention. The growth of agriculture was far behind and much slower than the industrial development. The uneven development between industry and agriculture led to income differentials and inequality between agricultural and nonagricultural activities. As noted by Ho (1978), 40 [i]n the early 19508 the per capita real income originating in Taiwan’s nonagricultural sector was twice that in agriculture. Despite a steady rise in agricultural productivity, rapid industrialization after the mid-19508 widened this differential considerably so that in 1966-70 the average per capita real income in the nonagricultural sector, at NT$ 11,791 in 1966 prices was more than three time that in agriculture (p. 140). Along with the agricultural-nonagricultural income differentials, the geographical location of new industry, the pressure of population on land, and surplus labor in the agricultural sector, combined to become a powerful economic incentive for people to move off farms and into the cities. In the 19508 and 19608, as rural population increased and the size of farms diminished, an increasing number of farmers were seeking off-farm occupation oppOrtunities. During this period, “surplus labor was a seasonal phenomenon, [but] by the mid-19608, an increasing number of farm workers had moved to other occupations or were finding it profitable to take short-term jobs in cities or in factories” (Ho, 1978:158-9). 2.2.3.1 Rural Industrial Development In the early postwar Taiwan, major cities were usually primary destinations of migrants. Industrial decentralization which contributed to population redistribution and de- concentration (Liu and Tsai, 1990; Tsai, 1981; Tsai, 1990) did not occur until the mid- 19608. Rural industrial development changed human migration patterns and directions (Tsai, 1990). In the early postwar period, significantly smaller shares of the labor force in manufacturing, commerce and transportation and communication were employed in the 41 rural areas. The cities were centers for modern industries (Ho, 1979:81; Tsai, 1981 :20). Industries began to move to rural areas in the mid-19608 in search of low cost labor and land. As of the 1971 industrial and commercial census, manufacturing establishments located outside Taiwan’s 5 provincial and 11 other cities accounted for 50 percent of the manufacturing employment (Ho, 1979:83).lo By 1979, about 43 percent of the total number of factories in Taiwan were located in 12 rural counties (Tsai, 1981 :21). The development of rural industry was a crucial influence on the rural population migration. The growth of industries in a spatially decentralized manner has enabled an increasing number of farm households to combine farming with part-time or full-time employment in nonfarm activities. Thus, it has helped to ease both the pressure of population on land and the corresponding pressure on farm household members to migrate to cities for jobs (Ho, 1979:88). Meanwhile, Taiwan’s decentralized pattern of industrialization in combination with its relatively well-developed transportation system has made it possible for many of the country’s rural household members to shift to nonagricultural employment without changing their residence -- they commute to work from rural areas. '0 According to Tsai (1981), “the spatial structure of industries in rural areas of Taiwan has been greatly affected by government policies relating to regional planning, land use, agricultural development, etc.” (p. 20). For the rural industrialization, the government established industrial zones near rural areas. From 1953 to 1979, the government planned to establish 42 locations covering 7,145 hectares for the establishment of industrial zones in rural areas. By 1979, only were 18 rural industrial zones set up in 12 rural counties. The total amount of land used for rural industrial zones were 2,032 hectares of land, which comprised 28.4% of the total land within the planned rural industrial zones (Tsai, 1981). In addition, another 11 rural industrial zones were developing in 1979 (Tsai, 1981). 42 In sum, the introduction of new techniques (e.g., power tillers and herbicides)ll to agriculture has made it possible to substitute new inputs for labor, 80 members of farm households have been free to participate in nonfarm activities (Ho, 1978:94). Income differentials and geographical locations of urban industries caused rural-to-urban migration. Because of the industrial decentralization, rural industries emerged. Increasing employment opportunity in nonagricultural activities for rural households enabled some farmers to change occupation or work in industry in the off-season without moving fiom the rural areas (Ho, 1978:158-9). It helped to raise average rural income, diversified family income sources, and also had the laudable effect of moving the distribution of income among farm households in the direction of greater equality (Ho, 1979:92-3; Lin, 1985; Park and Johnston, 1995: 184). 2.3 Population Distribution and Migration in Taiwan The uneven economic development policies not only increased income disparities between agricultural and nonagricultural sectors, but also acted as the impetus for rural- to-urban migration. The agricultural labor force migrating from rural areas to seek non- agricultural jobs in urban sectors between the 19508 and the early 19708 responded to the unbalanced development of economy (Tsai, 1978). Because the booming industrial factories and commercial offices were geographically polarized (Tsai, 1978), the population flowing from rural villages and small towns to large urban cities became the major streams of internal migration in Taiwan following World War II (Speare, 1974; Tsai, 1978). ” See Footnote 6. - 43 The population redistribution accompanied both the economic development and rapid urbanization (Tsai, 1978). After World War 11, Taiwan was transformed from a largely rural country to one in which the majority of the population worked in industrial production and commercial services. This change resulted in most Taiwanese living in metropolitan areas or cities with populations over 100,000 people. The proportion of population living in cities of over 100,000 inhabitants increased from 29.2% in 1961 to 55.5% in 1991 (see Table 2.2). In contrast, the proportion of population living in small rural and urban townships of less than 20,000 residents dropped from 34.0% in 1962 to 5.9% in 1991. The number of urban cities containing 100,000 and more inhabitants increased from 10 in 1962 to 30 in 1991.12 ‘2 See Table 13. Population of Cities of 100,000 and More Inhabitants in Statistical Yearbook ofthe Republic of China, 1992, DireCtorate-General of Budget, Accounting and Statistics, Executive Yuan, Republic of China. 44 Table 2.2 Population in Localities of100,000 and more and between 20,000 and 100,000 Residents in Taiwan : 1962-1991 Year 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 Localities of 100,000+ Localities of 20,000 - Localities of 20,000- Total pop. residents ' 100,000 residents residents' in Taiwan Pi 96 Pl 96 14 96 14 3404003 29.6 4192849 36.4 3914876 34.0 1 1511728 3532766 29.7 4394217 37.0 3956540 33.3 11883523 3695619 30.2 4588900 37.4 3972163 32.4 12256682 3835165 30.4 4755378 37.7 4037805 32.0 12628348 4027974 31.0 4978013 38.3 3986776 30.7 12992763 4468528 33.6 5442875 40.9 3385168 25.5 13296571 5002220 36.6 6015567 44.1 2632583 19.3 13650370 5348876 37.3 6379303 44.5 2606683 18.2 14334862 5565674 37.9 6634032 45.2 2476258 16.9 14675964 5788164 38.6 6936802 46.3 2269857 15.1 14994823 6005021 39.3 7141451 46.7 2142576 14.0 15289048 6441356 41.4 8177887 52.5 945587 6.1 15564830 6778464 42.8 8204503 51.8 869257 5.5 15852224 7094900 43 .9 821 1307 50.8 843495 5.2 16149702 7322262 44.4 8096064 49.0 1089864 6.6 16508190 7536872 44.8 8212631 48.8 1063624 6.3 16813127 7774450 45.4 8308216 48.5 1053048 6.1 17135714 8187655 46.8 8248843 47.2 ‘ 1042816 6.0 17479314 8395674 47.2 8356380 46.9 1053013 5.9 17805067 8709894 48.0 8354624 46.1 1070990 5 .9 18135508 9076507 49.2 8298942 45.0 1082474 5 .9 18457923 9281463 49.5 8376843 44.7 1074632 5.7 18732938 9483960 49.9 8403338 44.2 1125214 5.9 19012512 9771617 50.7 8356626 43.4 1129810 5.9 19258053 10176497 52.3 8184215 42.1 1093898 5.6 19454610 10510749 53.4 7986523 40.6 1 175340 6.0 19672612 10833999 54.4 7844509 39.4 1225304 6.2 19903812 1 1004061 54.7 7957024 39.6 1 146355 5.7 20107440 1 1280802 55.4 7872408 38.7 1 199756 5.9 20352966 11403827 55.5 7934197 38.6 1218818 5.9 20556842 Source: Directorate-General of Budget, Accounting and Statistics, Republic of China, Statistical Yearbook of the Republic of China (1992). These two columns are calculated by the author. There were specific trends for migrants in terms of sex, age, and education between late 19508 and early 19708. Migration rates were highest for young adults (Chang, 1979; Chiang, 1978; Tsai, 1978; Yin, 1978), although males and females had different migration patterns in terms of age in the early 19708 (Chang, 1979). Males in the age 45 cohorts of 25-34, and 4044 had the highest migration rate, while females were most likely to migrate at the age 20-29. Most employed female migrants were in the age group 15-29, with the peak at ages 20-24 (Chiang, 1978). In contrast, employed male migrants were generally older than the female, with the largest concentration at ages 20-34. Besides sex and age differentials, there were certain unique patterns of educational attainment among migrants. In the early 19708 a large proportion of migrants were primary school graduates or illiterate, and migrants with a college education were few (Tsai, 1978). The better educated, however, were more likely to migrate (Chang, 1978; Chiang, 1978; Speare, 1974). Migrants who went to the cities had the most education; and those who went to nrral towns had the least education (Speare, 1974). In addition, migrants tended to be better educated than residents at the place of origin, however, they were not necessarily better educated than the residents of the place of destination (Speare, 1974) Structurally, internal migration in Taiwan responded to geographical patterns of industrial and commercial development (Liu, 1982a and 1983b; Liu and Tsai, 1990; Speare et al., 1988; Sun and Tsai, 1981; Tsai, 1978; Tsai, 1981; Tsai, 1990). Between the late 19508 and the early 19708 only major cities and a few counties around Taipei and Kaohsiung metropolitan areas had positive migration rates (Tsai, 1978). The population concentration was based on push and pull factors. Pulling migrants was the large increase in number of industrial factories and commercial offices. This was due to industrial development and trade expansion being geographically concentrated to those major cities 46 and areas around two municipalities, Taipei and Kaohsiung (Tsai, 1978; Tsai, 1990). In addition, rural migrants were pushed out by the relatively small government investment in agricultural development (Tsai, 1978). This led to the decrease in agricultural income and farmers’ lives becoming increasingly difficult (Tsai, 1978). Rural-to-urban migration was a response to the geographical unevenness of economic development. A great deal of agricultural population migrated to major cities and areas around two municipalities where better-pay jobs were expected (Tsai, 1978). A8 a consequence, these areas experienced more rapid urbanization in the early 19708. Economic development continuously influenced population redistribution in Taiwan during the 19708 and later. The industrial decentralization contributed to the population redistribution (Liu and Tsai, 1990; Tsai, 1981; Tsai, 1990). Rural industrial development changed human migration patterns and directions (Tsai, 1990). Incorporating with the rural industrialization, the Taiwanese government in 1973 launched an agricultural reconstruction program which was designed to accelerate rural development, including the improvement of rural transportation, agricultural development, and rural community services (Ho, 1978; Yu, 1977).13 The agricultural reconstruction program firrther improved the living standards of rural population. In sum, the establishment of rural industrial zones led to an increase in the number of employment opportunities in industries and factories for agricultural labor forces in rural areas (Tsai, 1981). '3 See “The Accelerated Rural Development Program in Taiwan” by Terry Y. H. Yu. In this publication, Yu describe the governmental rural development policies in the mid-19708 in Taiwan. 47 The accessibility to the labor market of industrial production reduced the out-migration likelihood of potential movers from the rural farming sectors. In the early stage of the postwar years, a great number of rural migrants moved toward major cities, the two largest municipalities, and their satellite urban townships. Since the 19708, more and more potential migrants might have chosen to stay in rural villages and commuted to workplaces, while the other would have migrated to nearby loci where the rural industrial zones were established. Migration was not the only means to compensate for poverty. In addition, small-size urban areas where industrial zones were usually located had experienced a consistently higher rate of population growth since the early 19708 (Liu and Tsai, 1990; Tsai, 1990). In sum, internal migration patterns in Taiwan reveal that human migration is a behavior responding to economic development. 2.4 Taiwanese Families and Migration Traditionally, migration has been considered as an individual matter. If the individual perceives the personal benefits of migration to exceed the costs incurred in the act of migration, the individual is expected to migrate (Sjaastad, 1962; Bowles, 1970; DaVanzo, 1981). However, in the Taiwan Context, it is important to ask what is the role of family in human migration? 2.4.1 Family Dynamics in Taiwan As a subsistence unit, a family carries two major functions. One is to assign jobs in the family labor force (Harbison, 1981), and the other is to meet consumption needs for its members. As the family resources are defined as the means for providing labor force input and meeting consumption needs of family members, it seems that the more limited 48 the resources of the family, the stronger the motive and incentive to migration. However, within the family framework, whether a family can provide adequate opportunities for labor input and meet needs for consumption for its members is determined by the interaction of three factors: the individual, family, and structural levels (Harbison, 1981). The individual level factor is “the relative status of the individual within the family, governing differential access to the total production of the family” (Harbison, 198 1 :23 8- 9). The family level factor is “the size and quality of the resources held by the family (that is, the direct link between the family and the environment)” (Harbison, 1981 :23 8). The structural level factor is “the available technology and other aspects of the socio- cultural system that determine productivity” (Harbison, 198 1 :23 8). Access to land is of prime importance in agricultural communities. Focusing on the family level factor, the size and quality of farming land play a significant role in the decision making of sending family members away for maintaining the family or maximizing the familial welfare. B. Gallin and R. S. Gallin (1974:338) based on their studies on Hsin-Hsing Village to indicate that the families of migrants moving to large cities are holding less land than those families without migrants. Speare (1974) demonstrated that “a higher proportion of migrants than nonmigrants came from nonfarm families” in his study on the central Taiwan (p. 322). However, it is doubtful that the size of family landholdings has an absolute power in determining migration decisions. It is extremely likely that migration decisions are a function of family landholdings and other familial and structural factors, and probably some factors related to individual characteristics. 49 Holding a small amount of farming land could not lead to a migration decision until at least one of two possible factors emerge. The family, first, perhaps has to be unable to provide its members adequate opportunities for inputting their labor force. Second, the production from the small piece of land could not meet the consumption needs of its members. These two factors are associated with the family composition. A family with more family members in the labor force needs more land than a family with less members participating in labor force. Furthermore, larger families need more land than smaller families to produce agricultural production for consumption. Inadequate agricultural production or inadequate land for labor input could cause in the migration of family members; Nevertheless, long-distance migration is not necessary until resources which provide job opportunities and generate production for family consumption are extremely limited not only within families, but also in migrants’ hometowns. B. Gallin and R. S. Gallin (19742337) pointed out that having no local industries and insufficient job opportunities caused Hsin—Hsing villagers to migrate to large cities. Speare (1974) argued, in a similar vein, that “men were not forced to migrate to one of the large cities if the family landholdings were insufficient to support them, but could choose to move to a nearby urban center or even remain in their traditional home and commute to work” (p. 305). In Taiwan before the 19708, a great number of rural people migrated to large cities, which were far away from their hometowns. B. Gallin and R. S. Gallin (1974) argued that 50 it was because of that “the largest cities are most attractive to migrants with little capital and few skills, for only there can they find relatively high-paying employment” (p. 337). By contrast, they pointed out that “those who migrate only a short distance tend to come from families of comparatively high socio-economic status,” and “have ample landholdings, surplus capital, or both, and family members are not tied to land” (p. 335). Speare (1974) suggested that “for most [short-distance migrant] men the decision to move to the city was based more on the comparison of job opportunities in the city with those in the rural area than on an absolute need to find employment” (p. 305). Based upon these observations, it is clear that structural constraints could push rural poor people to migrate large cities to seek occupation opportunities. Nevertheless, it is not necessary that everyone responds to same structural difficulties in the same way. In addition, migration decision is not necessarily based upon same considerations/rationales. Poor people adopt migration as a family sustenance strategy. People with a higher socio-economic status choose migration to diversify family income sources and to supplement family incomes. Another motivation for rural people to migrate could be the relative deprivation and risk diversification (which will be discussed in Chapter HI). However, what roles did Taiwanese families play in response to the urban-rural uneven development? Although migration is considered an individual behavior, migration decision is a family strategy. Further, we should ask how migration decision is made within the Taiwanese family. 51 2.4.2 Migration Decision and Power Dynamics within Family Viewing the family as the decision-making unit whose corporate welfare is being maximized, family strategies are developed for coping with the constraints imposed by the larger labor markets (F emandez-Kelly, 1982). Family strategies are also developed with the collective good in mind, family members must accept those decisions and carry them out (Wolf, 1991). By contrast, individuals must sublimate their own wishes for larger goals (Wolf, 1991). Although it is often asserted that the family makes decisions or makes very precise calculations about allocating labor (Pahl and Wallace, 1985; Guest, 1989), the family can neither decide nor think (Davidson, 1991; Wolf, 1991). For Taiwanese families, the authority for decision making usually is held by one or a few family members. Certain people within the household make decisions. One or more persons with enough power to implement them makes decisions and other, less empowered family members follow them. Wolf (1991) argues that “household[/family] strategies necessarily embody relationships of power, domination, and subordination if a strategy is formulated by the decision maker(s) and successfully executed by those for whom decisions are made” (p. 32). Under the assumption that families adopt migration as a part of a survival or mobility strategy, money transfers are likely to be involved (Findley, 1987). It is suggested that the persons who make decisions about money matter in the family are likely to be the ones who make migration decisions (Findley, 1987). In Taiwanese society, this person would usually be the family head. The migration decision may be made solely by the family 52 head, but as a part of family survival and mobility strategy the decision of migration is a family matter, responsive to the needs of the entire family (Findley, 1987). No matter who is the decision maker, the migration decision is inevitably made based on the individual, family, and structural characteristics. Focusing on the individual and family characteristics, family power dynamics of migration decision should be discussed. In Taiwan, the authority for migration decision making could be traditionally held in the hands of family heads. Their decisions on migration, however, concern not only the motivation of potential migrants, but also their capabilities. In general, migration in terms of labor allocation inevitably influences the division of labor within a family, which is related to the status of the life cycle of a family. The duty of a migrant must be taken care by other family members staying at home. Those family members with un-transferable duties, therefore, are not allowed to move. Those people without proper capabilities for the labor market in the destination are not allowed to move either. In sum, although family heads in Taiwan usually hold the authority to decide who can migrate and where to go, their decisions are determined by the interactions of the characteristics of individual members and families, and structural factors. A family at the later state of life cycle might have more family members. The increase of family members could lead to the needs for more land for labor input and for agricultural production for family consumption. Further, migration could be a more-acceptable family strategy, because more family members would be available for dealing with the change of 53 division of labor, which is caused by the migration of certain family members. On the other hand, migration decision makers would give migration permissions to those who are likely to fit in the labor markets in the destination. 2.4.3 Consequences of Migration A large proportion of this chapter has dealt with the history of Taiwanese economic development and causes of migration. The consequences of migration and how they reinforce migration also deserve attention. The first change in migrant families is the division labor within family, which is caused by the change of family composition and the reduction of available labor force. Bernard Gallin and Rita Gallin pointed out that a large proportion of migrants still owned and rented some land in their hometowns (B. Gallin, 1966; B. Gallin and R. S. Gallin, 1974). Usually, migrants retained their land more for security than as an economic investment (B. Gallin, 1966:38; B. Gallin and R. S. Gallin, 1974:345). However, it is inevitable that married migrant males left their work to their wives at homes. It increased women’s responsibility of taking care of family land. B. Gallin (1966) pointed out that “[v]illage women whose husbands spend a good deal of time working in Taipei are especially active on their own land” (p. 65). Since migration moves a part of family labor force away from farming family land, labor shortage for agricultural production frequently happens. Especially, during harvest seasons, migrant families have to hire necessary labor to overcome the problem of labor shortage. 54 Family partition could be another consequence of migration. Speare (1974) argued that migration leads to the breakup of the extended families. He pointed out that “[m]ost married migrants moved with their wives and children” (p. 319). Only is a very small proportion of married male migrants lived apart from their wives (Speare, 1974). Therefore, Speare (1974) suggested that the movement of the entire conjugal units to cities resulted in the division of existing extended families. '4 However, Chuang (1972) argued that migration did not necessarily result in family partition, but the emergence of a new type of family in rural areas, which was the federated family. Traditionally, a family in the Taiwanese or Chinese society refers to a unit consisting of members related to. each other by blood, marriage, or adoption (Cohen, 1976; Lang, 1946). Family members live under the same roof. As an economic unit, family members usually share a common estate and common budget (Cohen, 1976; Lang, 1946). By contrast, a federated family is a group of conjugal units surrounding parents (Chuang, 1972). People in different conjugal units live under different roofs and cook separately. Although people in different conjugal units live separately, they have a very close relationship. Also, each conjugal unit is economically independent from other units; they manage budgets separately. They don’t share a common budget, but the group of conjugal units does not go through the process of family division. Migrant conjugal units leave family land to those conjugal units, which remain in hometowns, if they do have '4 Nevertheless, Speare did not address the definition of “family division.” Of course, rrrigration causes in that certain family members could not live with their parents and married and unmarried siblings. However, it is not necessary to mean the occurrence of family division, which is defined as the division of family property and the termination of obligations of family members. 55 land. The father who usually is the federated family head represents the entire federated family to participate ritual and other activities in villages. Finally, parents and those conjugal units remaining in rural areas usually are security resources for migrant units in financial and psychological matters. 2.5 Summary Taiwanese immigration started as early as the 7th century. In 1895, Japanese occupied Taiwan. In the course of Japanese colonial rule, Taiwan was transformed into a supplier of food and raw materials, and a receiver of manufactured and commercial goods. Rice and sugar accounted for a great proportion of Taiwan’s total exports. In order to use the cultivated land more efficiently and change the land tax system, the Japanese government changed the three-level tenancy system to the two-level system. However, the Japanese land reform policy did not prevent the occurrence of unequal distribution of cultivated land. After World War II, the Nationalist government migrated to Taiwan in 1949. The new government adopted a series of land reform policies to redistribute the ownership of land _ to establish its political base with the peasants. After the land reform, a great number of tenants became landowners. Furthermore, since holding their own land, the new landowners worked hard and farm households increased their agricultural production. However, for farrrr households, availability of cultivated land straightly relates to labor opportunity. Inadequate cultivated land could cause either inadequate opportunity for labor force input or inadequate production for consumption, or both. Under these 56 circumstances, family members might develop strategies to cope with the problems (Grigg, 1980; Guest, 1989; Wood, 1981). Migration usually is one of the strategies. The first question this research will deal with is how the amount of cultivated land available relates to migration. The amount of cultivated land is assumed to be responsive to family consumption needs and the opportunity for labor force input. Once either one is unbalanced, family migration could occur to cope with the substance problem. This research, therefore, will examine whether households with smaller sizes of cultivated land are more likely to have migrant family members. The second question this research will discuss is how the development of rural industries influences the migration of the rural population. It is assumed that inadequate land could not produce enough products to meet family consumption needs or provide enough opportunities for labor force input of family members. The emergence of rural industries is assumed to provide nonagricultural occupation opportunities for rural population and reduce the income inequality. Therefore, this research is interested in understanding how rural industry influences the rural migration patterns, for example, if industrial decentralization stopped the rural-to-urban migration. Since a family is the basic social unit for Taiwanese, this research will further study migration under the family perspective. Under the assumption that families adopt migration as part of a survival or mobility strategy, no matter who is the decision maker, the migration decision should be made based on not only the motivation of potential migrants, but also the individual and familial capabilities. Considering that migration is a 57 process of family labor allocation, which would influence the division of labor within a family, the third research question, therefore, assesses how the status of the life span of a family influence the migration decisions and patterns. For example, the research will examine if members of extended families are more likely to migrate and if more migrants are from extended families than other types of families, which typically have less members than extended families. Furthermore, another research question related to family should be raised, which is how family power dynamics influence migration decisions. Although data on family dynamics are unavailable, this research will explore the role such a process takeslin‘ decisions about migration, based on the following assumptions. First, it is assumed that individual characteristics such as age and gender reflect the position of family member within the authoritarian hierarchy of the family. Second, it is assumed that decisions within the household are shaped by considerations such as family continuity over time (the idea here is that daughters leave the family to become members of their husbands’ households while sons remain to provide old age security - theoretically at least). Give these assumptions, who migrates and to pursue what type of employment may suggest the dynamics embedded in seeming neutral family strategies. 58 CHAPTER III THEORIES, RESEARCH FRAMEWORKS, HYPOTHESES, DATA, AND METHODS 3. 1 Introduction This chapter will describe and explain the theoretical research frameworks, hypotheses, data and methods to examine the research questions raised in Chapter II. In the second chapter, in addition to providing an introduction to early immigration in Taiwan, economic development including both agricultural and industrial development, changes of the economic structure, and its families following the historical perspectives, four research questions regarding human migration, which form the core of this dissertation, were raised. These four research questions cover three dimensions: family, structural, and individual perspectives. At the family level domestic units are seen as being responsible for providing adequate cultivated land for family consumption needs and the opportunity for labor for input. Family migration strategies could be developed to cope with inadequate land for family consumption and labor input. The first research question deals with how the amount of cultivation land available relates to the decision of migration. The demands on land for family consumption and labor input are determined by family composition. Families, primarily living on farm, with more people consuming or participating in labor force need more family land than those with fewer family members. Thus, this dissertation will examine how family type influences the migration decisions of family members. 59 At the structural level, the emergence of rural industry provides more occupation opportunities for surplus labor in rural villages than agriculture. This research, therefore, will examine how the development of rural industries influences the migration of the rural population. In addition, this research will rely on the individual theoretical perspectives of migration to examine family power dynamics within the family, how they relate to migration decisions, who in the family participates in human migration, and why. While these four questions generally focus on the causes of human migration, the first two questions deal directly with economic issues. They assess if a family’s land ownership influences the decision to migrate, and how structural economic factors determine migration behaviors. Questions three and four are related to family issues. Specifically, question three deals with the relationship between family structure and human migration, while question four deals with power dynamics within Taiwanese families. The research question on power dynamics examines how factors, which contribute to an authoritarian family hierarchy, influence migration. All of these four questions are intended to explore what caused migration from a rural village in Taiwan. This chapter opens with a literature review of migration. With such an understanding of migration theories, reasonable and rational research frameworks to address my four research questions can be constructed. Following the literature review, two research 60 frameworks are constructed and series of research hypotheses are derived based on the research frameworks. Finally, this chapter discusses the data and the appropriate statistical methods from univariate descriptive statistics to complicated statistical modeling. The data section discusses what data are used, how they were collected, and the Operationalization of variables. The methods section accesses the statistical methods to be applied throughout the dissertation. 3.2 A Review of Migration Theories In the past, migration research has dealt mainly with the following questions: Who dominates migration flows? Where do people migrate? Where did they come from? Why do people migrate? What are the causes and Consequences of migration for the areas of origin and destination? These questions have been addressed largely within three major theoretical frameworks commonly labeled the individual (cost-benefit analysis, monetary and psychic benefits, and human capital characteristics), the structural (intervening opportunities, uneven economic development, and regional restructuring), and the family/household perspectives (accessibility to land, and participation in local wage labor markets). These three different theoretical approaches have been applied to different analytic levels of data. 3.2.1 Individual Perspectives of Migration Theories Conventionally, population movement has been conceptualized as the geographic mobility of workers who are responding to imbalances in the spatial distribution of 61 factors of production (Guest, 1989; Harris and Todaro, 1970; Sjaastad, 1962; Todaro, 1969, 1.976, 1980; Wood, 1981).15 The individual perspective of migration theories presupposes that workers seek out employment opportunities where their returns will be greatest. Therefore, Shaw (1975) argues that human migration, in a formal sense, could be considered as a case of the microeconomic theory of consumer choice. Human migration is an aggregate process. Migration flows are the cumulative results of individual decisions based on a rational evaluation of the benefits to be gained from and the costs entailed in moving (Wood, 1981). The cost-benefit model of microeconomics has played a crucial role in migration research. Since researchers view migration as the outcome of a rational evaluation of the costs and benefits of movement (see Massey, 1990a; Sjaastad, 1962; Todaro, 1976, 1980), the expected net return to migration has methodologically been used as an indicator to predict if a potential migrant would choose to move or to stay. If the expected net return is positive, potential migrants would choose to move; if it is negative, potential migrants would choose to stay; and if it is zero, potential migrants are indifferent about The assumptions underlying this approach are based upon a concept of dual economy consisting of areas characterized by zero or very low productivity because of surplus labor, and areas characterized by high labor wages because of the scarcity of labor. The resulting differential in wages stimulates workers from the low-wage areas to migrate to the high-wage areas. Furthermore, the labor movement changes the supply of, and demand for, labor in both sending and receiving areas. The redistribution of labor force theoretically adjusts wage rates of workers in both the origin and the destination. The increased supply of the loss of workers from the sending area where provides low-wage creates upward wage pressure there. Neoclassical economists, therefore, believe that migration is an equilibrating mechanism that brings about wage equality in the two sectors by way of shifting human resources from areas where their social marginal products are often assumed to be zero toward areas where these marginal products are not only positive, but also rapidly growing as a result of capital accumulation and technological progress (Lewis, 1954; Massey et al., 1993; Rains and Fei, 1961). 62 either migrating or staying (Bowles, 1970; DaVanzo, 1981; Guest, 1989; Harris and Todaro, 1970; Sjaastad, 1962; Todaro, 1969, 1976, 1980; Wood, 1981). The individual cost-benefit model, therefore, concerns the computation of the expected costs and benefits of migration (see Speare, 1971). Migrants may capture increased wages associated with their greater labor productivity, but they also must undertake certain investments, including monetary and non-monetary costs (Sjaastad, 1962). The former include expenses incurred by migrants in the course of moving, such as costs of transportation of themselves and disposal of movable and iminoVable property necessitated by a shift in residence. The non-monetary costs include the earnings foregone while traveling, searching for, and learning a new job. In addition, there are psychic costs involved in migration but which are difficult to measure. The returns of migration can also be broken down into monetary and non-monetary components. Non- monetary returns include changes in “psychic benefits” as a result of locational preferences. Sjaastad (1962) argues that the expected net return of migration is the sum of monetary returns and psychic benefits accrued from migration minus the monetary and psychic costs of migration. Nevertheless, the monetary returns of migration are not the real income to be earned in the receiving sector, but the “expected” income. Expanding upon Sjaastad’s notion, Todaro (1969, 1976) proposes that the expected net return is a function 63 of urban-rural expected income difference and the likelihood of obtaining an urban job. '6 The possibility of potential migrants obtaining jobs in modern urban sectors is a crucial element in the decision-making process to migrate; indeed, it is more important than the wage differential. In most cases, there is a gap between rural and urban wages. Urban '6 Todaro’s (1976) mathematical equation of calculating the expected net returns of urban-rural migration is as me) = I[P(t) x Y. (t) — Y.(t)1>< a" x dz — C(O) t=o where, V(0) : the discounted present value of the expected net urban-rural income stream over the migrant’s time horizon, P(t) : the probability that a migrant will have secured an urban job at the average income level in period t, Yu(t) : the average real incomes of individuals employed in the urban economies, Y,(t) : the average real incomes of individuals employed in the rural econorrries, C(O) : the cost of migration, n : the number of time periods in the migrant’s planning horizon; and i : the discount rate reflecting the rrrigrant’s degree of time preference. Massey (1990) modifies the equation of Todaro. ER(0)= jip.(t)xp2(t)>0. N98 >0. nob; )0. 3-3 :0 8.: )0. ch can: a 147 4.5.3 Demographic Process -- Fertility The fertility rate for Taiwan a3 a whole sharply increased in the early postwar period (Freedman et al., 1994; Hermalin et al., 1994). Using governmental records, Hermalin (1994) shows that the national crude birth rate in 1949 -- the year the nationalist government moved to Taiwan -- was 42.4 per 1000 population. Based on their calculations, Freedman and his associates (1994) demonstrate that the crude birth rate (CBR) of Taiwan in 1956 was 44.8 per 1000 people. Since 1956, the crude birth rate has constantly declined. Freedman and his associates (1994) also point out, “[b]etween 1965 and 1990, the total fertility rate for Taiwan fell by 72% [, and] the crude birth rate fell ...[by] 63%” (p. 267). Information from a series of the Taiwan-Fukien Demographic Fact Book, Republic of China shows that the crude birth rate of Taiwan area in 1965 was 32.1 per 1000 population, and 24.4 in 1979 (see Table 4.6). The data collected by interviewing villagers reveal that the crude birth rates in Hsin-Hsing Village were 31.7 and 22.1 per 1,000 population in 1964 and 1978, respectively (see Table 4.6). These two figures reveal a significant decrease (about 30 percent) in the crude birth rate in Hsin-Hsing Village during the period between 1964 and 1978. 148 Table 4.6 A Comparisons of Crude Birth Rate, General Fertility Rate, and Child-Woman Ratio of Hsin-Hsing Wllage (HHV), Taiwan Area (TWN), Chang-Hua County (CHC), and Pu-Yan Rural-Township (PYT) : 1965 and 1979‘ 1965 1979 HHV TWN CHC PYT HHV TWN CHC PYT Crude Birth Rate 31.7 32.1 31.3 33.8 22.1 24.4 27.1 28.8 General Fertility Rate 144.1 151.8 146.5 156.2 85.3 94.5 108.4 116.9 Child-Woman Ratio 422.2 742.5 732.2 756.5 356.6 438.7 485.7 483.9 Because the data used for this research were collected in 1965 and 1979, the crude birth rate and general fertility rate are calculated based on the number of live births in 1964 and 1978. The way the crude birth rate and general fertility rates are calculated is imperfect. For the crude birth rate of 1964, it is calculated by the number of live births divided by the number of year-end population of 1964, and multiplied by 1000. The same procedure is applied to the calculation of the 1978 crude birth rate. The year-end population of 1964 is the total population in 1965 minus new births in 1965. Therefore, in this case, the year-end populations of 1964 and 1978 are 524 and 540. Methodologically, the general fertilty rate is the number of live births in a given year divided by the number of women ages 15—49, multiplied by 1000. In this research, general fertility rate is calculated as the number of live births in 1964 or 1978 divided by the number of women ages 15- 49 in 1965 or 1979, multiplied by 1000. Although the way the crude birth rate and general fertility rate calculated is imperfect, they still provide a general picture of fertility in Hsin-Hsing Village in the mid-19603 and the late 19703. Hsin-Hsing Village as the whole followed the national trend in terms of crude birth rate. Other demographic statistics, such as general fertility rate (OF R) and child-woman ratio, also indicate Hsin-Hsing Village followed the national fertility trend -— a declining fertility. However, compared to the fertility rates of its surrounding areas in the same township and same county, and at the national level, Hsin-Hsing Village had low fertility rates. As seen in Table 4.6, the general fertility rate at the national level in 1965 was 151.8 per 1000 population, but it was 144.1 in Hsin-Hsing. Comparing this rate to the regional statistics, Hsin-Hsing was low, while the statistics at the county level and rural- township level were 146.5 and 156.2 per 1000. In 1979, the general fertility rate of Hsin- Hsing Village was also lower than those for Pu-Yen rural-township (116.9 per 1000), Chang-Hua county (108.4 per 1000), and Taiwan as a whole (94.5 per 1000). 149 With limited data from the Statistical Yearbook of the Republic of China, 1983 as a basis for comparisons, the lower fertility rate in Hsin-Hsing was due to a smaller proportion of married women who were in reproductive ages, compared to Taiwan as a whole. The governmental data show that in 1966, Taiwan had 2.6 million women aged 15-44. Sixty- five percent of the women in the same age group were married.47 In 1965, Hsin-Hsing Village had 118 women ages 15-44, and 52.5 percent of this group of women were married. For Taiwan as a whole, in 1980, there were 2.5 million married women at ages 15-44, accounting for 58.9 percent of the women in the same age groups. The data of Hsin-Hsing Village show that in 1979, 53.4 percent (62 out of 116) of women at ages 15- 44 were married. Although data at the county or rural township were not available, the limited data demonstrate that the low fertility rate of Hsin-Hsing Village could be related to a small proportion of married women at the childbearing age. In terms of the general fertility rate, migrant women had higher GFRs than non-migrant women. The data in the absolute numbers show that in 1964, there were 17 live births among the Hsin—Hsing population. Two births were given by migrant women while the , other 15 mothers were non-migrant villagers. In 1978, most mothers (eight) of new-bom babies were residing in urban cities. Only four women giving births in 1978 were non- migrants. Taking the sizes of migrant and non-migrant population into account, in 1964, the general fertility rate for the non-migrant women was 140.2 per 1000, while that for migrant women was 181.8 births per 1000 women ages 15-49. The high general fertility 47 See Supplementary Table 4. Single and Married Population 15 Years of Age and Over, by Sex and Age, in Thornton and Lin’s Statistical Yearbook of the Republic of China, 1983. 150 rate of migrant women in 1965 was due to, first, the small number of married migrant women, and second, the predominance of migrant women who were single. In the late 19703, due to the participation of young couples in migration or the increase of conjugal units in cities, the general fertility rate increased to 190.5 per 1000 women ages 15-49. On the other hand, the GFR of non-migrant women dramatically dropped to 51.3 births per 1000. Besides the general fertility rate, the child-woman ratios for Hsin-Hsing Village as a whole were lower than those for the rural-township, county and the national levels in both research periods. While in 197 9 the village’s child-woman ratio was lower but close to the statistics at the rural-township, county and the national levels, the village’s 1965 child-woman ratio was much lower. than the statistics at all three other levels. The low child-woman ratio was directly related to the low number of live births in the early 19603, which led to a low proportion of children under 5 years of age. At the national level, in 1965 children under age 5 accounted for 15.7 percent of the population. In the same time, children under age 5 only accounted for 10.6 percent of the Hsin-Hsing population. In 1979, the gap in the proportion of population accounted for by children under 5 narrowed. Only 8.5 percent of Hsin-Hsing’s population was made up of children under age 5; the proportion was 11.2 percent for Taiwan as a whole. While the decline in fertility was caused by the initiation of the national family planning program, it was also a product of a delay in marriage. Illegitimate birth was not morally acceptable in the 19603 and 19703 and, in Taiwan, women would not have given birth 151 until they were married.48 The census data in 1966 show that in Taiwan 8.6 percent of women ages 15-19 and 59.5 percent of women ages 20-24 were married, while 92.9 percent of women ages 20-29 were married (see Lin, 1994). In 1980, 5.3 percent, 41.5 percent, and 82.7 percent of women at the age groups 15-19, 20-24, and 25-29 were married (see Lin, 1994). Nationwide industrialization is likely to lead to a delay in marriage. With the increase of occupational opportunities in the industrial and service sectors in Taiwan, single women were likely to participate in labor markets to contribute the economy of their natal families as long as possible. Consequently, women were very likely to delay marriage. In contrast to 1965, in which one birth occurred among teenagers in the village, no teen mothers were found in Hsin-Hsing Village in 1979. Besides starting birthing later in the ‘ late 19703, women stopped giving birth earlier. While the 1965 survey data show some births by women age 35 and older, there were no births by women older than 35 years old in 1979. A delay of marriage postpones the start of women giving births. and shortens the period of reproduction. All the data show that villagers’ reproductive behaviors had been changing during the mid-19603 and the late 19703. The changes were possibly related to the change of the village’s economic structure. The shifting from an agricultural-based economy to an industrial-based created more job opportunities for not only men, but also women. To 48 Pre-marital sex was practiced but in most instances pregnancy led to marriage. 152 contribute to the economy of their natal families, single women delay marriage, and consequently, postpone their first births. In addition, because of an increase of young married women residing in the urban cities with their husbands, in the late 19703, the migrant population had a high fertility rate than the non-migrant one. 4.6 Socio-economic Infrastructure 4.6. 1 Education The educational attainment of Hsin-Hsing villagers improved from an average of 2.89 years of schooling in 1965 to 4.40 years in 1979 (see Table 4.7). This increase reflects changes in governmental educational policies in the 19503 and 19603, especially, in 1968 when compulsory education was increased to the ninth grade, as well as the villagers’ improved economic condition. Education was provided to people in Taiwan for free, although students paid a small amount of fees for tuition, materials and lunches. Students from poor families, however, were eligible for discounts and/or scholarships. Villagers in the late 19703 generally received more education than their counterparts in the mid-19603. Among villagers at age six and over, the illiterate accounted for 38.1 percent in 1965 and 26.5 percent. For villagers age 25 and over, the proportion accounted for by the illiterate decreased from 60.3 percent in 1965 to 41.2 percent in 1979. Their average educational attainment was 2.32 years in 1965 and 4.07 years in 1979. 153 For Taiwan as a whole, the proportion of the school-age population in elementary school increased from 4.7 percent in 1905 to 57.6 percent in 1940.49 In 1964, 96.8 percent of children ages 6-11 were attending elementary schools, 43.5 percent of the population aged 12-14 was enrolled in junior high schools, and 23.7 percent of people of ages 15-17 was enrolled in senior high schools.50 Only 8.1 percent of people ages 18-21 attended colleges. By 1979, the proportions of population attending primary, junior high, and senior high schools, and colleges increased. Almost all children (99.7 percent) ages 6-11 were enrolled in primary schools. Eighty-six percent of teenagers ages 12-14 attended junior high schools, and 52.8 percent of the population ages 15-17 were enrolled in senior high schools. The proportion of people ages 18-21 in colleges almost tripled, increasing to 23.9 percent in 1979. Generally speaking, from the beginning of the 20th century to the late 19703, the educational environment improved continually. More and more people attended schools and stayed inschool longer. According to a Taiwanese governmental report published in 1981, at the national level, 72.9 percent of Taiwanese aged six and over either were attending schools or had some education, and 23.1 percent of the population was illiterate in 1965.51 In 1979, the illiterate accounted for only 10.3 percent of the Taiwanese population. Meanwhile, the educated accounted for 86.6 percent, who either were in schools or had some education. 49 See Table 3.2, Selected Demographic and Socioeconomic Indicators, 1905-1940, in Social Change and the Family in Taiwan (1994: 52). so See Table 3.2, School Attendance Rates by Level of Schooling and Sex, Selected Years, 1949-1988, in Social Change and the Family in Taiwan (1994: 68). ' 5‘ See Table 248, Population by Levels of Education, in Taiwan Statistical Data Book (1981). 154 In Hsin-Hsing, the improvement of educational attainment of young people was significant, especially for those who were ages 15-34. In 1965, the average years of education for those 15-24 of age were 5.36 years, and 3.70. for those aged 25-34. The data collected in 1979 reveal that the average years of education for Hsin-Hsing villagers increased to 8.75 years for those aged 15-24, and to 7.03 years for ages 25-34. Besides demonstrating an increase of educational attainment of the villagers between 1965 and 1979, these figures show that young villagers generally had more education than their older counterparts. Table 4.7 Average Years of Educational Attainment of Hsin-Hsingjilfigers, 1965 and 1979 - 1965 1979 Age Group Male Female Total Male Female Total Under 15 2.15 1.75 1.94 2.21 2.12 2.17 15-24 6.89 3.91 5.36 9.13 8.25 8.75 25-34 4.92 2.14 3.70 8.32 5.86 7.03 35-44 4.50 1.48 2.96 6.04 2.48 4.19 45-54 2.63 .88 1.77 5.00 1.65 3.37 55-64 2.73 .00 1.30 2.35 .57 1.46 65 and over .23 .00 .10 1.42 .00 .63 25 and over 3.49 (3.9) 1.11 (2.4) 2.32 (3.5) 5.40 (4.0) 2.83 (3.6) 4.07 (4.0) All ages 3.80 (3.6) 2.01 (2.9) 2.89 (3.4) 5.07 (4.1) 3.64 (3.8) 4.40 (41» Note: Standard deviations are in the parentheses. In addition to the inequality of educational attainment between the younger and older generations, educational inequality also can be found between men and women. Although both the male and female population in Hsin-Hsing generally increased their overall educational level, boys received more education than their female counterparts. The gap 155 between males and females was wider among older than younger villagers. Also, the difference was wider in 1979 than in 1965. In 1965, among villagers who were 25 years and older, 41.3 percent of males attended school for six years or longer, while only 14.3 percent of female villagers did so. In 1979, the proportion of villagers with at least six years education were 64.7 percent among males and 38.6 percent among female villagers. Opportunities to attend school for male and female villagers increased, but the inequality between men and women persisted. In 1979 among 64.7 percent of males with at least six years education, more than 35 percent of them at least had finished a middle-school education. In contrast, only 14.3 percent of women finished a middle-school education among those who attended schools for six years 01' more. B. Gallin (1966: 196), describing the village in the late 19503, argues that “[some] girls had either not attended [schools] at all that year [1957], or had dropped out during the course of the school year.” He (1966: 196) further argues that “[t]his is probably attributable to one or both of two factors: the impoverished condition of some families and the negative attitude of some parents toward the necessity of schooling, particularly for girls” (p. 196). In contrast, the gap in the educational levels between males and females of the young generation narrowed between 1965 and 1979. In other words, in 1979 the difference between the educational attainment of young males and young females was smaller than that of 1965. This change may reflect a change of parental attitudes toward the necessity of schooling for girls. The other contribution to this could 156 be the implementation of a new government educational policy in 1968 extending free education to nine years of schooling. In general, migrants had more education than non-migrant villagers. Comparing the average years of educational attainment of villagers in the labor force, in 1965 while the non-migrants had an average of 2.56 years, migrant villagers had an average of 5.38 years. In 1979, the gap between non-migrants and migrants was narrowed (4.00 years for non-migrants and 5.41 years for migrants). The similar pattern is also found among both male and female villagers. In 1965, the difference of average educational years between non-migrant and migrant men was about 2.1 years (3.43 years versus 5.51 years). In 1979, the difference decreased to about 1.3 years (4.70 years versus 5.98 years). For female villagers, the change was dramatic. In 1965, while non-migrant female villagers had only 1.82 years of education, their migrant counterparts had 5.00 years. The gap in education between non-migrant and migrant female villagers decreased to about 1.5 years (3.22 years versus 4.73 years) by 1979. However, the difference in the years of education between migrant and non-migrant villagers was due to the uneven distribution of the illiterate. Most illiterate villagers resided in the village. In 1965, among villagers aged 25 and over, 32 percent of migrants were illiterate, while 64.5 percent of non-migrant villagers were illiterate. In 1979, among the villagers in the same age group, while only four migrant villagers (6.3 percent) were illiterate, 96 non-migrant villagers (53.6 percent) were illiterate. 157 4.6.2 Occupation In postwar Taiwan, agriculture was the primary economic base. In the first two decades of the postwar period, more than 50 percent of the working population aged 12 and over was participating in agricultural production.52 In 1965, 53.7 percent of workers age 12 and over were working in the agricultural sector. Only 12 percent of the working population participated in industrial production. Meanwhile, people working in commerce, transportation, personal services, professions, government services, and other sectors accounted for 34.3 percent of the working population age 12 and over. In 1979, the proportion of working population participating in industrial production for the first time exceeded that of people working in agriculture.53 At the national level, people participating in agricultural and industrial productions respectively accounted for 29.7 percent and 29.8 percent of the working population age 15 years and over.54 Although there were no data available to compare the occupational patterns of urban and rural residents directly, we, however, should not assume that the occupational patterns of urban and rural population were the same. As Tsai (1981) reports, rural industrial development was started in the early 19503. During the early stage, “[t]he total area of land used for industrial zones in rural counties [was] comprised of 28.4% of the total land within the planned zones” (Tsai, 1981: 20). In 1960 for Taiwan a whole there were 18,791 factories. There were 5,282 factories located in five large cities, accounting 52 See Table 2-5A, Employment (Age of 12 and Over), in Taiwan Statistical Data Book (1981). Please note that in 1965, the governmental statistical data were only available for age 12 and above. 53 See Table 2-SB, Employment (Age of 15 and Over), in Taiwan Statistical Data Book (1981). 5" Since 1967, only workers aged 15 and over have been included into employment surveys. 158 for 28.1 percent of factories in Taiwan, and 4,395 in four metropolitan counties, which accounted for 23.4 percent. Other rural counties in Taiwan contained 9,114 factories, which accounted for 48.5 percent of factories in Taiwan. In the 19603 and 19703, rural industry emerged. A great number of new factories were established in rural areas. By 1979, 20,966 factories were found in mral counties, which accounted for 35.9 percent of all factories in Taiwan. There were 23,921 (40.9 percent) and 13,548 factories (23.2 percent) located in four metropolitan counties and five large cities, respectively. Consequently, although we do not know the exact proportions of population working in agriculture, industry, or service sectors in urban and rural areas, we can assume that more urban labor force participated in the industry sector than rural population. 4.6.2.1 Non-Migrant Villagers (1965) Hsin-Hsing data on occupations were collected by asking villagers two questions. The first question is “what does a family member do most of the time?” The second one is “does he/she do anything else?” The primary economic source of livelihood for Hsin-I Hsing villagers was agriculture in 1965. About 58.7 percent (148 out of 252) of non- migrant villagers in the labor force, who were residing. in the village, took jobs relevant to agricultural production, especially male villagers (see Table 4.8). While 47.3 percent (95 out of 201) of male non-migrants were not in the labor force, 106 male villagers who were in the labor force accounted for 52.7 percent of non-migrant males. Seventy-nine male villagers primarily worked as self-employed farmers accounting for 39.3 percent of non-migrant male villagers. The same group of male villagers also accounted‘for 74.5 percent (79 out of 106) of non-migrant male villagers who were in the labor force in 1965. Only a few non-migrant male villagers (17) held occupations other than as self- 159 cultivating farmers. Two male villagers sold their labor working in agriculture production while seven male villagers worked as off-farm wage workers. Without industrial occupational opportunities, four of these seven male villagers worked at governmental offices, while one was hired to work at a pork stand and one guarded a fish pond. Only one male villager worked in a small factory. The remaining seven male villagers operated their own businesses in the village. These seven male entrepreneurs accounted for the same proportion as those who were off-farm wage workers. 160 Table 4.8 Hsin-Hsing Villagers’ Occupations by Migration Status and Labor Force: 1965 and 1979 1965 1979 Male Female Male Female N % N % N % N % Non-migrants Not in labor force Sub-total 95 47.3 98 40.2 95 46.6 73 40.1 In labor force Housekeeper 73 29.9 1 0.5 39 21.4 Family worker I 0.4 l 0.5 6 3.3 Self-employed farmer 79 39.3 55 22.5 35 17.2 25 ’ 13.7 Farm laborer 2 l .0 12 4.9 Off-farm worker 7 3.5 3 1.2 33 16.2 35 19.2 Self-employed (off-farm worker) 7 3.5 l 0.4 20 9.8 4 2.2 Military 8 4.0 17 8.3 Unemployed 3 l .5 l 0.4 2 1 .0 Sub-total 106 52.7 146 59.8 109 53.4 109 59.9 Total 201 100.0 244 100.0 204 100.0 182 100.0 Migrants Not in labor force Sub-total 8 14.0 7 30.4 38 45.2 29 41.4 In labor force Housekeeper 3 1 3 .0 13 1 8.6 Family worker I 1.2 5 7.1 Off-farm worker 38 66.7 10 43.5 21 25.0 21 30.0 Self-employed (off-farm worker) 3 5.3 22 26.2 2 2.9 Military 2 3.5 2 2.4 Apprentice 4 7.0 2 8.7 Unemployed 2 3.5 Unknown 1 4.4 Sub-total 49 86.0 16 69.6 46 54.8 41 58.6 Total 57 100.0 23 100.0 84 100.0 70 100.0 Among 146 female non-migrant villagers who were in labor force, most reported housekeeping as their primary occupation, followed by those who reported themselves to be self-employed farmers. In 1965, those who were not in labor force accounted for 40.2 percent (98 women) of non-migrant female villagers, while 146 women who were in the labor force (out of 244 women) accounted for 59.8 percent of non-migrant female villagers. Seventy-three of non-mi grant female villagers who were in labor force reported 161 being housekeepers at home as their primary occupations, and 55 (22.5 percent) non- migrant women worked as self-employed farmers in 1965. The 55 self-cultivating farmers also accounted for 37.7 percent (55 out of 146) of the female labor force within the village. In addition, one woman operated her own business. Nevertheless, some of women who reported working as housekeepers were also doing other jobs. For example, two women reported being farm laborers as their secondary occupations, while two women did piece work at home, one was a dress-maker, and four farmed their family land. Combining both male and female villagers, there were eight villagers operating their own businesses in the village in 1965. Because the rural industrialization had not started influencing the village’s economic structure, most of these enterprises were service-based at this time. While five (including four men and one woman) entrepreneurs were selling fruits, vegetables, fish, or pork, and were like venders, one man operated a grocery store in the village. The other two villagers who owned their own businesses included one doctor and one man who operated a carpenter workshop. In addition, in 1965, 148 (including 81 males and 67 females) non-migrant villagers worked in the agricultural sector. This group of villagers accounted for 58.7 percent (148 out of 252) of non-migrant villagers who were in the labor force. Although this proportion does not include those women who reported “housekeeping” as their primary occupation and who also worked on family land, it was still higher than the national statistics (53.7 percent), mentioned previously. This reflects the fact that in 1965 162 agriculture was the primary source of family income for people residing in the village. Because off-farm occupational opportunities in the surrounding area were limited, family land was the primary means of production. Therefore, people in the village relied more on agriculture than any other income sources, although remittances from family members residing and working outside of the village were critical to their survival.55 Wang and Apthorpe (1974: 32) point out in their research that “half of the first 25 households interviewed in depth said that they were in receipt of income from outside family members. On average this means 33.5 percent of the total income of households is not derived from farming”. 4.6.2.2 Migrant Villagers (1965) In 1965, migration carried a strong economic function. People migrated to seek occupations elsewhere. Most Hsin-Hsing migrants (86.0 percent of male migrants and 69.6 percent of female migrants) participated in wage labor markets. Similarly to those residing in Hsin-Hsing, migrant villagers who were not in labor force were primarily students or children too young to attend schools.56 However, while most villagers residing in Hsin-Hsing worked in agricultural production, most migrant villagers sold their labor for wages. Among 49 male migrants who were in labor force, 38 worked for 55 Wang and Apthorpe (1974:78) suggest that remittances were not necessarily in “the form of cash.” They ranged from visible to invisible. For the visible, in some cases they could have been agricultural machinery (Wang and Apthorpe, 1974). For the invisible, some “brothers living away from the farm may have made over at any rate the use of their land to a brother living at home perhaps for no rent if he pays the land and irrigation tax” (Wang and Apthorpe, 1974: 78). 56 Those who reported themselves housekeepers who were usually female are counted in the labor force. Although they did not directly contribute to a family monetary income, they indirectly contributed to the family economy. 163 wages at destination, while only three male migrants were self-employed. A great proportion of the 38 hired male villagers did service work, such as pulling carts and operating pedicabs, delivering goods, and working as store clerks. According to the records, there was only one male migrant villager working as a manufacturing worker in 1965. Among migrant villagers, males and females had different occupational patterns in 1965. While a great proportion of male migrants (20 out of 38) worked in the service sector, 50 percent of female migrants (5 out of 10 women) who were working as off-farm workers participated in industrial production as factory workers. These women were relatively young. Three of them were teenagers and two were in their early 203. In addition, they were all single. The other five female labor migrants included two teenagers, two in their 203, and one 60-year old woman. Two of them worked as cooks, one worked in the Central Market, one was a paid housekeeper, and the other worked as a clerk. Among these five women, two were married and three were single. To sum up, a comparison of the occupations of people residing in Hsin-Hsing and those who migrated out to seek occupational opportunities shows that most of villagers (including migrants and non-migrants) were not participating in industrial production in 1965. There were only seven villagers (including one non-mi grant male villager, one migrant male villager, and five migrant female villagers) in the labor force, who worked in industrial production. Most non-migrant villagers participated in farming, while a great proportion of migrant villagers participated in the service sector. 164 These findings reveal that in 1965 migrants were more likely to take occupations in the service sector, while non-migrants were more likely to work on farms. Very few migrants or non-migrant villagers worked in industrial production for wages. While no industrial production opportunities were available for non-migrant villagers in Hsin-Hsing, social networks influenced the migrant villagers’ occupation-seeking behaviors. Studies suggest that job seekers’ heavily rely upon social networks (Campbell and Marsden, 1990; Holzer, 1987 and 1988; Rees, 1966). The influence of social networks on Hsin-Hsing migrants’ job seeking behaviors is also evident. In the mid-19603, industrial production occupations were available in cities, but 20 male migrants, who accounted for more than 50 percent of male migrant off-farm workers, were working in the service sector. Strikingly, 15 male migrants in the service sector were working at the same place -- the large, wholesale Central Market located in Taipei. These 15 male migrant villagers accounted for three-quarters of male migrants working in the service sector. They were either working as cart drivers who delivered vegetables or as clerks who sold vegetables. 4.6.2.3 Non-Migrant Villagers (1979) By 1979, the occupational structure of the residents of Hsin-Hsing had changed dramatically. Although 27.5 percent of non-migrant villagers in the labor force still worked in farming, agricultural production was no longer the primary economic source of family income. As the data in Table 4.8 show, there were more male non—migrant villagers working as off-farm laborers (33) or self-employed (20) than those who were self-cultivating farmers (35). The 33 male villagers who sold their labor for wages included: (1) eleven villagers who worked in factories, (2) six who worked as white- 165 collar workers, including those people working in banks and government offices, (3) twelve working in the service sector including one deliverer, three drivers, seven construction workers, and one baker, and. (4) four male villagers hired as carpenters for wages. In sum, one-third of male non-migrant villagers who reported they were off-farm workers participated in industrial production. Of the remainder, 15.2 percent were white- collar workers and 36.3 percent of them participated in the service sector. The 35 male non-migrant villagers working as self-employed farmers accounted for 32.1 percent of male non-migrants in the labor force. There was also a great increase in the number of women working off-farm. While in 1965, only 1.2 percent of non-migrant female villagers worked off-farm, 19.2 percent of non-migrant female workers reported themselves to be off-farm workers in 1979. If we merely take the women participating in the labor force into account, there were only 2.1 percent (3 out 146) of non-migrant female laborers working off-farm for wages in 1965. In 1979, the proportion dramatically increased to 32.1 percent (35 out of 109). The occupations that were more likely to be taken up by this group of people included working in factories and doing piece work at home. There were two female non-migrants working in the service sector. Meanwhile, between 1965 and 1979, there was a significant decrease in the proportion of females reporting themselves as housekeepers. Among 39 women who reported housekeeper as their primary occupation, ten also reported working a second job in 1979. They were working as cooks, construction workers, or farming family land. The increase in women reported taking a second occupation could be attributed first, to the availability of more occupational opportunities 166 for villagers in 1979 than in 1965, and second, to more males participating in off-farm labor markets, thus making farming family land a responsibility shared by women. In addition, among villagers residing in Hsin-Hsing Village, those who worked in agricultural production were generally older than those Who worked as off-farm laborers (see Table 4.9). This pattern was found in both 1965 and 1979. In 1965, while the average age of male self-employed farmers was 38.2, that of male off-farm workers was 28.8. A similar pattern was also found among female villagers (28.8 versus 28.3). In 1979, the difference in the average ages between self-employed farmers and off-farm workers increased. The average ages of male self-employed farmers and off-farm workers were 54.7 and 34.0, respectively (see Table 4.9). Among female villagers, they Were 44.3 for self-employed farmers and 32.3 for off-farm workers (see Table 4.9). These Statistics reveal that agriculture has been geriatrified; family farming was usually taken care of by the old population. In contrast, young villagers preferred to participate in off- farm labor markets. 167 Table 4. 9 Average Age and Educated Years of Hsin-Hsing Villagers by Selected Occupations, and M :gration Status: 1965 and 1979 1965 1979 Male Female Male Female Age Educ. Age Educ. Age Educ. Age Educ. Non-migrants Housekeeper -- - -- 37.6 1.7 80.0 0.0 49.8 1.8 Family worker -- -- 22.0 6.0 29.0 9.0 30.0 6.0 Self-employed farmer 38.2 3.7 28.8 2.1 54.7 2.7 44.3 1.8 Farm laborer 41.5 2.5 27.2 2.2 -- -- -- -- Off-farm worker 28.8 6.6 28.3 2.0 34.0 6.7 32.3 4.9 Self-employed (off-farm worker) 35.7 5.6 47.0 0.0 36.9 6.4 37.8 4.5 Migrants Housekeeper -- -- 22.7 4.0 33.2 6.0 Family worker -- -- -- -- 21.0 9.0 30.2 6.0 Off-farm worker 26.5 6.0 23.9 4.9 32.4 8.7 24.4 7.5 Self-employed (off-farm worker) 30.3 2.7 -- -- 30.7 7.8 26.5 6.0 With the emergence of rural industry in the surrounding area of Hsin-Hsing, there was an increase in the number of villagers participating in off-farm labor markets, and in entrepreneurs in the village. (According to the 1979 data, there were twenty male and four female entrepreneurs residing in the village (see Tables 4.8 and 4.10). One woman operated a workshop, which was a satellite factory of a big manufacturer, and one man owned a spring factory (see Table 4.10). Both of these entrepreneurs hired‘factory workers. Three male villagers operated machine, motorcycle shops in 1979, which had not been seen in 1965. Three villagers were operating construction businesses, including two men and one woman. With the growth in entrepreneurship, vender-type small-service businesses decreased from four in 1965 to one in 1979. 168 Table 4.10 Entrepreneurship amorlgresident vflgers, Hsin-Hsin : 1965 and 1979 1965 1979 Entrepreneurship Male Female Male Female Spring manufacturer 1 Motorcycle shop 2 Machine shop 1 Workshop (carpentry, suitcase manufactory, satellite factories) 2 1 Village store 1 3 1 Drug store 1 Repair shop 2 Construction services 2 1 Rice mill and rice drying business 2 Doctor 1 Barber shop 1 Other merchant (fruit, vegetable, fish, pork, sawdust, 3 1 4‘ construction materials) - Total 6 1 20 4 In 1979, there were four male villagers selling vegetables, sawdust, and construction materials. However, there was only one village running a vender-type business. 4.6.2.4 Migrant Villagers (1979) Among 46 male migrant villagers, 21 (45.6 percent) sold their labor for wages, and 22 (47.8 percent) operated their own business at their destination in 1979. While the number of villagers selling their labor for wages decreased between 1965 and 1979, the number ' of male villagers operating their own businesses in cities increased from three to 22 male migrant villagers. Meanwhile, although more male migrants worked in manufacturing in 1979 than in 1965, the number was still small. Eight male migrants worked in factories in 1979, representing 38.1 percent of male migrants who sold their labor for wages. Not surprisingly, as more male villagers operated businesses in cities and conjugal migrant units increased (which will be discussed in Chapter VI), the number of migrant women who worked as family workers increased. In 1965, no women worked as a family worker in cities. In contrast, five married women, who accounted for 12.2 percent of 169 female migrants in the labor force, helped to operate family businesses in 1979. In addition, the number of female migrants selling their labor for wages increased from 10 in 1965 to 21 in 1979. Among these 21 off-farm female workers, only eight (38.1 percent) were married, and only two married women worked in manufacturing. The eight women who worked in factories were mainly single. The national statistics show that those who participated in agricultural production accounted for 46.5 percent of the employed population in 1965 (Taiwan Statistical Data Book, 1981). The proportion decreased to 21.5 percent in 1979 (Taiwan Statistical Data Book, 1981). Following the national trend, the occupation patterns of non-migrant villagers reveal the increase of occupational opportunities in the industrial and service sectors in Hsin-Hsing Village between 1965 and 1979. The newly developed occupation opportunities in the industrial and service sectors absorbed the labor force released from the agricultural sector or just entering the labor market. More female villagers sold their labor for wages in 1979 than in 1965. They primarily worked in non-agricultural sectors. Meanwhile, agriculture was geriatrified. While younger villagers participated in off-farm labor markets, older villagers had to take responsibility for farming family land. In addition, more and more villagers were self- employed. The number of self-employed villagers increased among both migrants and non-migrants. Villagers, especially male, chose to work off-farm and created occupational opportunities for themselves. They were relatively younger than those who worked in the agricultural production. Furthermore, although in 1965 most migrant 170 villagers who participated in labor markets were male, the number of women who were migrant workers in cities increased in 1979. Among female migrant workers, while very few married migrant women worked in manufacturing, most factory female workers were single. As shown in Table 4.9, not only was the average age different between people holding different occupations, so also was the average year of educational attainment. Among non-migrant villagers, those who participated in agricultural production had fewer years of education than those working off-farm. In 1965, among non-mi grant men, while self- employed farmers had an average of 3.7 years of education, off-farm workers had 6.6 years and entrepreneurs had 5.6 years. In 1979, the average years of education for male 'off-farm workers and male entrepreneurs respectively increased to 6.7 and 6.4, and that for male self-employed farmers decreased to 2.7 years. The data also show that in 1979, among female non-migrants, those who worked off-farm had more education than self- employed farmers. Among the migrants, entrepreneurs had less education than those who worked as off- farm worker or helped operate family businesses. In 1965, while male entrepreneurs had an average of 2.7 years of education, those selling their labor for wages had 6.0 years of education. In 1979, while male migrant off-farm workers had 8.7 years of education, migrant entrepreneurs had 7.8 years. A similar pattern was found among female migrants. While those migrant women working for wages had 7.5 years 'of education, those women operating businesses in urban cities had 6.0 years of education. In sum, the statistics on 171 the average years of education reveal that those who sold their labor for wages had more education than those who held other types of occupations. In addition, these statistics might imply that occupations recruiting workers required more education than did participating in agricultural production or being entrepreneurs. 4.7 Summary With very few exceptions, Hsin-Hsing Village, in general, followed the national trends in several ways. First, although median ages were different between Taiwan as a whole and Hsin-Hsing Village, the age structures were simila -- the young accounted for a large portion of the population. For example, in 1965 both Taiwan as a whole and Hsin-Hsing had more than 50 percent of their populations younger than 20 years old. The population information in 1979 shows that the populations in Hsin-Hsing and Taiwan as a whole were getting older. Their median ages increased, and more old people could be found in 1979 than in 1965 in the population. However, the migrantand non-migration populations had different age structures. While in the mid-19603 migrants were primarily married males and the young generation participated in migration in the late 19703, the mean age of migrants in 1965 was older than that of non-migrants, and in 1979 non- migrants had a higher mean age than migrants. In 1979, young people in Hsin-Hsing and Taiwan as a whole did not account for the same proportions of the population as they did in 1965. This was primarily related to the decline in the fertility rate. Hsin-Hsing Village followed the national trend. Fertility rates decreased dramatically. In addition, women in the village had a shortened reproductive period. In 1979, female villagers at the reproduction ages started their first birth later and 172 stopped reproduction earlier than did women in 1965. This reproduction behavior of the female villagers followed the national pattern. As Freedman et a1. (1994) points out: “Taiwan’s fertility decline began first among women at older ages and later among women at younger ages,” implying that for Taiwan as a whole, women’s reproductive period shortened. In terms of education, more and more children attended schools and they stayed in school longer in 1979 than in 1965. This can be attributed to changes in governmental education policies in the 19503 and 19603, especially the extension of the compulsory education to the ninth grade in 1968, and rising incomes. The change in educational policies influenced the young population more than the old. For example, in 1979, Hsin-Hsing villagers ages 15-24 had 8.75 years of education, which means that almost every villager in this group finished his/her junior high school education. The average years of education attainment for villagers at ages 25-34 increased from 3.7 years in 1965 to 7.0 years in 1979. Besides education, Hsin-Hsing Village was also influenced by national economic plans. As the development of rural industry was a part of national economic plans, more and more off-farm occupation opportunities were created by the newly established industries and factories in the rural hinterland in the 19703. Hsin—Hsing villagers, following the national trend, shifted from working primarily in agricultural production to work in the industrial and service sectors. At the national level, more than a half of workers worked in agricultural production in 1965, but the proportion dropped to less than 30 percent in 173 1979. In Hsin-Hsing Village, the proportion of family members in the labor force working as self-cultivating farmers dramatically declined from 43.7 percent in 1965 to 23.1 percent in 1979. Meanwhile, the proportion of family members holding paid occupations in the businesses in and surrounding Hsin-Hsing Village increased from 22.5 percent in 1965 to 33.9 percent in 1979. 174 CHAPTER V THE DETERMINANTS OF RURAL-URBAN MIGRATION FROM HSIN-HSING VILLAGE Chapter IV provided a profile of the research village, in terms of its geographical location, demographic infrastructure, and family, social, and economic infrastructure. As the data showed, between 1965 and 1979, the number of Hsin-Hsing villagers living outside of the village almost doubled. During the same time period, for Hsin-Hsing Village as a whole, the number of villagers participating in wage labor markets increased, while the number of self-employed farmers dramatically decreased. Between 1965 and 1979, Hsin-Hsing Village experienced a slight increase in population from 537 to 545, respectively. The number of people actually residing in the village, however, decreased dramatically (from 454 to 389). In fact, while there were 83 migrants (15% of Hsin-Hsing’s total population) documented in 1965, by 1979, the number of the villagers who resided outside of Hsin-Hsing Village increased to 156, accounting for 28.6 percent of the total population. Hsin-Hsing Village followed national trends in economic development. The postwar national economy was developed from a base of agriculture to one primarily dependent on industrial production (B. Gallin and R. S. Gallin 19823; Gallin and Ferguson, 1988; Ho, 1978), and rural industrial development occurred in the 19703 (Tsai, 1981; Lu, 1981). The growth of industries in a spatially decentralized manner enabled an increasing number of farm households to combine farming with part-time or full-time employment 175 urn—- in non-farm activities. Hsin-Hsing villagers shifted from working primarily in agricultural production to employment in the industrial and service sectors in the late 19703. The number of self-employed farmers (134), accounting for 53.2 percent of non- migrant villagers in the labor force in 1965, decreased to 60, accounting for 27.5 percent in 1979. The proportion of villagers holding off-farm jobs for pay increased from 26.2 percent (83 out of 317) of the village labor force in 1965 to 53.8 percent (164 out of 305) in 1979 (see Table 4.8). In Taiwan, agriculture accounted for 23.6 percent of the national gross domestic production in 1965 and decreased to 8.6 percent in 1979 (see Table 2.1). The significance of agriculture declined in Hsin-Hsing Village over time as well. And, as we saw above, the proportion of Hsin-Hsing villagers in the local labor force who reported they were self-cultivating farmers dramatically declined between 1965 and 1979. At the same time, farming was geriauified. The mean age of farmers increased from 34.2 years in 1965 to 50.7 years in 1979. Among male farmers, the mean age increased from 38.1 years old in 1965 to 54.7 in 1979, while that of female farmers increased from 28.8 years old in 1965 to 45.0 in 1979. In addition, the total area of land available for cultivation (including both land owned and rented-fborrowed-in) for farmers in the village decreased from 43 chia in 1965 to 39.2 chia in 1979. This chapter examines the relationship between: (1) change in land access and migration, and (2) change in rural economic structure and migration in both 1965 and 1979. Specifically, the issues discussed include: 1. whether labor migration was related to family type, 176 2. whether labor migration was related to a family’s accessibility to land, and 3. whether labor migration was related to villagers’ local labor market participation. 5.1 Family Type and Migration 5.1.1 Theoretical Position and Hypothesis The structure of the Taiwanese family takes one of three forms: conjugal, stem, and joint.57 These forms represent not only different compositions but different sizes as well. In general, a joint family has more family members than other family forms, while a conjugal family has the fewest members. The stem family falls in between in size. A family, like a person, experiences both birth and growth. As a family moves from the simple conjugal type toward the complex joint form, its size increases. An increase in family size leads to additional family members joining the labor force as well as an increase in a family’s consumption needs. Therefore, in rural areas, the different family types not only represent differences in size, but also reflect varying needs for land to meet a growing family’s need for consumption and labor input. For large families, sending family members to seek job opportunities in cities may be necessary as well as possible when rural economic opportunities are inadequate. Having additional family members increases consumption needs, in addition to a possible increasing need for labor input opportunities. Larger families include the labor reserves necessary to cover excess farming work generated by absent migrant family member(s). 177 In contrast, migration is less likely for conjugal families, which typically consist of a married couple and their unmarried children, although in Hsin-Hsing a number of simple families include older parent(s).58 Therefore, the first hypothesis examined in this research is: H1: Stem and/or joint families are more likely to have migrant family members than _ conjugal families. I 5. 1 .2 Measurement Migration considered in this analysis is labor migration, which is defined as the movement of people of working age away from Hsin-Hsing Village. A labor migrant is a villager in the urban labor force, who was residing outside of the village during the two research periods. Therefore, at the family level, labor migration in this research is measured as the number of family members in the urban labor force residing outside of Hsin-Hsing Village. .To examine the first hypothesis, analysis of variance (AN OVA) is used; this statistical technique “detect[s] evidence of any difference among a set of group means” (Agresti and Finlay, 1986:398). First, the mean numbers of labor migrants between types of family are tested for significant differences. Second, complex family types (i.e., stem and joint) are examined to determine if they have more family members in the urban labor force than the simple family type (i.e., conjugal). 57 See Chapter III for definitions of the three types of Taiwanese families. 58 Both the 1965 and 1979 data showed there was one conjugal family consisting of a married couple and the mother of the household head. 178 5.1.3 Analysis and Discussion In 1965 there were no significant differences in the number of labor migrants according to family type, while in 1979 there were significant differences in the number of labor migrants between different types of family. The data for both years generally show that larger families had more family members in the urban labor force than did conjugal families (see Table 5.1). Specifically, in 1965 stem and joint families had an average of 0.87 labor migrants who were in the labor force and residing outside of the village,lwhile conjugal families had 0.64. Nevertheless, the F -test demonstrates that the difference in the number of labor migrants between conjugal and complex families is not statistically significant at or=0.05. Therefore, the first research hypothesis is not supported by the 1965 data. Table 5.1 Analysis of Variance for Labor Migration by Family Types, Hsin- Hsigg: 1965 and 1979 Mean N F 1965 Conjugal family 0.64 44 1.06 Stem-Joint family i 0.87 38 1979 Conjugal family 0.42 36 50.60 Stem family 0.67 . 24 Joint family 4.31 13 m significant at level of .01 " significant at level of .05 significant at level of .10 1 Two joint families are included into this category for the analysis. As can be seen in Table 5.1, the F-test shows that in 1979 family type and labor migration had a significant relationship. The difference in the number of family members in the urban labor force between different types of domestic units is statistically 179 significant at 0t=0.01. The first research hypothesis is supported by the 1979 data. On average, however, the number of labor migrants in conjugal and stem families was very close (see Table 5.1). While conjugal families had an average of 0.42 labor migrants, stem families had 0.67. Joint families with an average of 4.31 labor migrants was much higher than the averages of conjugal and stem families. Post hoc comparisons are done to determine if larger family types had more labor migrants than small families in 197 9. In Table 5.2, we can see that the mean difference in the number of family members in the urban labor force between conjugal and stem families is not significant at or=0.05. Joint families had more family members in the labor force residing in urban cities than the other two types of family structures. Specifically, in 1979 joint families had 3.89 more labor migrants than conjugal families and 3.64 more than stem families. This finding indicates that the statistical significance shown in Table 5.1 for the 197 9 data was primarily due to joint families for either one or both of two reasons. First, large families have more members available to deploy to cities. Second, individual units within joint families, which already had migrant family members settled in cities, expanded and with the addition of children and, perhaps, as the youngsters, aged, additional wives and children as members. Without sufficient data, this research is not able to identify if the family members of joint families in urban cities were new migrants or the members of long-time migrant families. Regardless, joint families supplied many more labor migrants than any other type of family structure. 180 Table 5.2 Comparison of the number of labor migrants between conjugal, stem, andpint families in Hsin-HsirLgL1979 Mean difference Conj ugal families vs. stem families _.25 Conjugal families vs. joint families -3.89 ... Stem families vs. joint families -3.64 m ... significant at level of .01 .. significant at level of .05 ' significant at level of .10 The relationship between family type and labor migration was not significant in 1965 while it was significant in 1979. In the early years, some villagers adopted labor migration as a strategy to maintain family sustenance. Theoretically, facing uncertainty in urban cities, villagers were not sure of the benefits of migration (Massey 1990b; Taylor 1986). Therefore, male villagers usually migrated to cities alone and left their wives and/or families in the village (B. Gallin, 1978; B. Gallin and R. Gallin, 1974). In addition, a great proportion of migrant villagers maintained their landholdings in Hsin-Hsing for security. B. Gallin and R. Gallin (1974: 344) point out that in 1965. about three-quarters of Hsin-Hsing migrants to Taipei “still owned or rented some land in the village.” This finding suggests that regardless of family type, Hsin-Hsing villagers probably adopted similar strategies to overcome economic hardship. Nevertheless, small families with a limited labor force, part of which had to take care of family land, included few members who could migrate to seek job in cities in order to supplement the meager profits gleaned from the land. Put another way, small families had fewer available members to send to urban cities than did complex families. During the mid-19603, however, larger families did not send significantly more family members to seek job opportunities in urban cities, although they had more family 181 members in the labor force than small families. This might be the product of two phenomena. First, at that time, most villagers grew rice in the first and second crops and production required intensive human labor. In the absence of modern agricultural machines or technology, “[t]he entire rice-growing operation [was] performed by hand labor” (B. Gallin, 1966:49). (This operation included germinating and planting seedlings, flooding the rice paddy, transplanting seedlings to land, application of fertilizer, irrigating the crop, weeding, and finally rice harvesting.) Because rice production was dependent on a tremendous amount of human labor, labor migration in the 19603 was not generally considered an option unless “the family economic situation was extremely serious” (B. Gallin and R. Gallin, 1974: 344). Second, unmarried women and men rarely were deployed to cities in the 19603, because of the dearth of opportunities for jobs there for them. Therefore, families usually had only one or at most two family members who sought occupational opportunities in the cities. In fact, in 1965, there were only slight insignificant differences between conjugal and stem/joint families in terms of the number of family members in the labor force residing outside of Hsin-Hsing Village (0.64 versus 0.87). In 1979, more family members in the labor force resided outside of the village than in 1965. As shown in Table 4.8, 87 villagers (28.5% or 87 out of 305) who were in the labor force migrated or lived outside of the village in 1979. The comparable figures for 1965 were 65 (20.5% or 65 out of 317). By 1979 complex families, especially joint families, had more family members in the labor force residing in the cities than did simple families 182 (see Table 5.1). These differential labor migration patterns are primarily attributable to two structural factors: economic and family. Three aspects related to the economic structure contributed to the difference. First, agricultural production was not as important in 197 9 as it was in 1965. In the 19603, agricultural production was the primary source of family income. By the end of the 19703, off-farm employment had replaced agriculture as the primary source of family income for villagers residing in Hsin-Hsing (B. Gallin and R. S. Gallin, 1982a). If there were no economic opportunities locally, participating in off-farm labor markets would have been impossible in the late 19703. Therefore, the second factor responsible for differential labor migration patterns between families was that rural industrial development created off-farm jobs for villagers. As B. Gallin and R. S. Gallin (1982a) point out, in the late 19703, many firms, ranging from large labor-intensive factories to small satellite factories or family workshops, were established in Hsin-Hsing and its surrounding area, in addition to a govemment-sponsored industrial park erected in the next township.59 The third factor accounting for the differential, then, was the I “reorganization” of farming practices through (a) the implementation of the land consolidation program in the late 19603 and (b) the mechanization and chemicalization of Hsin-Hsing farming in the late 19703. The land consolidation program led to the possibility of utilizing modern agricultural technology, thereby releasing some human labor from agricultural production (B. Gallin, 1974; B. Gallin and R. S. Gallin, 1982a). 183 'F'_ _- Thus, compared to the late 19603, agriculture was mechanized rather than labor-intensive in the 19703. Although the conditions associated with the economic structure theoretically held across family type, family structure shaped the way the labor of working-age family members was deployed. In terms of labor migration, small families probably would have had fewer family members in the labor force residing in urban cities in 1979 than large families, especially joint families. Conjugal families, which, in general, consist of a couple and their unmarried child(ren), usually have two persons in the labor force. Although farming was not as labor intensive or necessary in 1979 as it was in the 19603, villagers with land continued to devote some time to farming.60 The increase in non-farm jobs combined with innovations in agricultural technology allowed villagers with land to take care of family land as well as to hold off-farm occupations in Hsin-Hsing Village and its surrounding area. This was especially true for those family members from conjugal families. For example, in 1979, there were 29 non-migrant villagers in the labor. force who worked for monetary income and held a secondary occupation as a self-employed farmer. Among these 29 villagers, 21 were men and eight were women living in 24 Hsin-Hsing families, 5" Although rural industrial development generated only a portion of the occupational opportunities available for villagers in the late 19703, it led to the creation of occupational opportunities in other sectors, such as the personal and service sectors. 6° Gallin and Gallin (1982:218) point out four reasons. First, land was a source of family income. Second, family land was a source of food. Third, farming families had to grow rice to pay for land taxes. Four, additional taxes would be imposed if family land was not cultivated. 184 of which six families included labor migrants. These villagers were either self-employed entrepreneurs, or worked as off-farm workers or family workers as their primary occupations. Sixteen (55.2%) lived in conjugal families while seven (24.1%) and six (20.7%) were members of stem and joint families, respectively. The proportion accounted for by those family members in conjugal families was larger than that of the combination (44.8%) of those in stem and joint families. Therefore, it is clear that in the late 19703, conjugal families were more likely than other types of families to increase their family incomes by participating in local off-farm labor markets and in devoting their spare time to taking care of family land than were other types of families. In contrast, complex families were more likely to include urban labor migrants than rural off-farm workers at this time. In point of fact, the data show that labor migration was a preferred option for complex families, especially joint families. In 1979, the average family size for conjugal, stem, and joint families was 5.1, 7.3, and 14.3, respectively (see Table 4.1). Among joint families, 54.7 percent (58 out of 106) of family members in the labor force worked outside of the village, while the figures for conjugal and families were only 12.7 percent (14 out of 110) and 18.2 percent (16 out of 88), respectively. In short, complex families included more labor migrants than simple families. Nevertheless, migrant families continued to farm. Because complex families, especially joint families, in general had more surplus labor than smaller families, when modern agricultural machines and chemical inputs reduced the need for a large agricultural labor 185 force, farming was left to older family members and to part-time workers, i.e., non- migrant family members who worked off-farm and took care of the family land. For example, within 13 joint families with labor migrants, 19 villagers participated in agricultural production. They included ten villagers (52.6%) who primarily worked as self-employed farmers and nine villagers (47.4%) who primarily worked as off-farm workers and tended the land as their secondary occupation (see Table 5.3). Among the ten villagers working as self-employed farmers, nine (90%) were age 50 and over. Among the nine villagers who worked as self-employed farmers as their secondary occupation, six (66.7%) were age 50 and above while only three (33.3%) were younger than 50 years old. Table 5.3 Characteristics of Hsin-Hsing Villagers Working on Family Land within Conjugal, Stern, and Joint Families with Labor Mi rants, 1979 Con'u al Stem Joint Total M F M F M F M F Villagers working on family land as a primary occupation Under 50 l 2 0 3 0 l l 6 Age 50 and over 4 3 3 0 7 2 14 5 Subtotal 5 5 3 3 7 3 15 ll Villagers working on family land as a secondary occupation Under 50 2 l 0 0 3 2 4 Age 50 and over 0 1 1 4 2 5 3 __Subtotal 2 l 1 4 5 7 7 The same analysis is conducted for the eight stern families with labor migrants. Six members (75%) of these families primarily worked on family land, while two members 186 (25%) worked on family land as their secondary occupation (see Table 5.3). Among the six villagers working on family land as their primary occupation, three were men age 50 and over, and three were women under age 50. The two villagers working on family land as their secondary occupation, were both age 50 and over, and one was a man and the other was a woman. In point of fact, within the 29 families (13 joint, 8 stem, and 8 conjugal) with family members in the urban labor force, 14 villagers, accounting for 35.0 percent of the 40 villagers working on family land, were primarily working off-farm and working on-farm as their secondary occupation (see Table 5.3). Twenty-seven villagers (67.5%) who participated in agriculture (as either a primary or secondary occupation) were age 50 and over. This reflects the fact that agriculture was geriatrified. If we take only the members of stem and joint families into account, 20 villagers participating in agricultural production were age 50 and over, and they accounted for 74.1 percent (20 out of 27) of the agricultural labor force of stem and joint families in the village. This proportion was much higher than that for conjugal families (53.8% or 7 out of 13). In contrast, 46.2 percent (6 out of 13) of the villagers farming family land in conjugal families with labor migrants were under age 50. Those who worked family land as their second occupation were all under age 50 (see Table 5.3). This finding reflects the fact that in complex families, farming was more likely to be left to older non-migrant villagers than it was in conjugal families. Younger villagers living in conjugal families were likely to work off- farm for wages as their primary occupation, and farm family land as their secondary occupation. 187 The division of labor by age was accompanied by some change in the division of labor by gender. There were slightly more male than female farmers who worked the land either as a primary or secondary occupation; 22 (55.0%) were men and 18 (45.0%) were women. Of these 18 women, eight were age 50 and over, and they accounted for 20.0 percent (8 out of 40) of the villagers participating in agricultural production. The remaining ten (25.0%) were under age 50. Of the 22 male farmers, 19 (47.5%) were age 50 and over. Only three male farmers (7.5%) were younger than age 50. This reflects the fact that agricultural production in the village in the late 19703 was primarily performed by older villagers. In addition, younger farmers were likely to be female. This reflects the fact that young resident women on occasion took over men’s farming work, thereby releasing male villagers for off-farm employment. The foregoing discussion shows that in the late 19703, there was no significant difference in the number of family members in the urban labor force between stern and conjugal families. The association between family type and the number of family members in the labor force residing outside of the village was primarily related to the increase in the number of joint families between 1965 and 1979. The fact that joint families were more likely to adopt labor migration or to have family members in the labor force residing in urban cities than other types of families in 1979 is attributable to the structures of the local economy and families. First, complex families had more family members to deploy; and second, joint families had more surplus labor than smaller families. These two factors intersect with the other three factors related to the economic structure: (1) the importance 188 of agricultural production was replaced by off-farm economic activities; (2) more off- farrn occupational opportunities for resident villagers existed in the late 19703 than the mid-19603; and (3) the land consolidation program of the late 19603, and the introduction of modern agricultural technology in the late 19703, obviated the need for a large farm labor force. These five factors also explain why smaller families, especially conjugal families had fewer family members in the labor force residing outside of the village. In sum, the first hypothesis (H1) is not supported by the 1965 data. The relationship between family type and labor migration in 1965 was not significant. As is evident in Table 5.1, the difference in family members resorting to labor migration between conjugal and stem/joint families was insignificant. In the early years the insignificant relationship between the number of family members in the labor force residing outside of the village and family type can be attributed to the fact that intensive human labor was necessary for farming land in Hsin-Hsing. Regardless of family type, most families could have only one or two labor migrants in the mid-19603, and the labor force which remained behind had to apply itself to agricultural production. The first hypothesis, however, is supported by the 1979 data. As shown in Table 5.2, joint families had significantly more labor migrants or family members in the labor force residing outside of the village than conjugal and stem families in 19.79. The “modernization” of agriculture and the development of rural industry allowed non- migrants, to maintain off-farm occupations as well as take care of family land. While the members of conjugal families were more likely to remain in Hsin-Hsing Village than 189 those who were members of larger families, joint families included more family members who held jobs or sought occupational opportunities in cities. Joint families used migration to solve the problems of surplus labor and insufficient agricultural production generated by family land. 5.2 Accessibility to Land and Migration 5.2.1 Theoretical Position and Hypothesis In the absence of off-farm economic activities, land is the most valuable resource and means of production in rural areas. The amount of land a family has accessible influences not only job opportunities available for the unit’s members, but also agricultural production for a family’s consumption. Adequate land provides family members an outlet for their labor power and generates sufficient agricultural products to meet the consumption needs of a family. If family land does not generate adequate agricultural products for consumption, a family must have its family members sell their labor for wages, thereby providing money to purchase resources to satisfy this need. Labor migration is a strategy to solve both the problem of insufficient agricultural production and the problem of surplus family labor (Grigg, 1980; Guest, 1989; Wood, 1981). Because agriculture was the base of the economy in Hsin-Hsing in 1965, accessibility to land was extremely important for the villagers. Theoretically, migration was one strategy to achieve a fit between their consumption needs and the labor power at their disposal (Boyd, 1989; Grigg, 1980; Guest, 1989; Wood, 1981). Once a family could not produce enough for consumption, the family was likely to resort to migration for some of its 190 members. Labor migration was also an option for families with insufficient land to absorb its labor. Therefore, the resulting research hypotheses are: H2: The smaller the landholding per family member in the labor force, the more labor migrants the family has. H3: The smaller the landholding per family member, the more labor migrants the family has. 5.2.2 Measurement I"'_"". While the second hypothesis (H2) examines the relationship between landholding for labor input and labor migration, the third hypothesis (H3) examines the relationship between landholding for family consumption and labor migration. The overall family farming land is the sum of the size of self-cultivated land and that of rented-/borrowed-in land. Landholding for labor input is operationallzed as the amount of overall family land divided by the number of family members in the labor force. This measurement, named landholding per working family member, indicates the average size of farming land per family member in the labor force. The numerator for this measurement is the amount of land a family has accessible, while the denominator is the number of family members in the labor force, including both resident and non-resident villagers. This measurement does not indicate whether a family has adequate family land for each member to input his/her labor. It is assumed, however, that a family with a larger average size of land per family member in the labor force provides more occupational opportunities to its I members than those with a smaller average size of land. 191 Landholding for family consumption is operationalized as the amount of overall family land divided by the total number of family members who are or are not in the labor force. Therefore, the numerator for this measurement is the amount of land a family has accessible, and the denominator is the family size, which counts both resident and non- resident villagers. This variable is called landholding per family member. Assuming family members have similar consumption needs, a larger family needs more farming land to generate agricultural products for consumption because such a unit has more family members to feed than a smaller one. Although this measurement does not indicate if the land available is sufficient, it assumes that a family with a greater amount of land per family member is more able to satisfy the unit’s consumption needs than a family with smaller amount of land per family member. In a broad sense, labor migration and the demand for land are inter-locked. On theone hand, a high demand for land would lead to a high possibility of labor migration if little land were available to accommodate a family’s need for labor input or to produce food for its member’s consumption. On the other hand, labor migration would be likely to relieve the pressure on land because labor migration reduces the number of people who need to input their labor and the amount of food needed to feed them. Because this research focuses on how family land accessibility influences labor migration, however, a family’s land accessibility is treated as an independent variable. In contrast, labor migration is treated as an outcome of a family’s land accessibility. “Family’s land accessibility” is assumed to precede the occurrence of labor migration, and represents the 192 size of land available for family members in the labor force to input their labor. With little land available, families would have to have family members who are in the labor force seek job opportunities outside of the village to input their labor. “Family’s land accessibility” also represents the amount of agricultural products generated for consumption. With little land, families may have inadequate agricultural production to satisfy their members’ consumption needs. Some family members in the labor force, therefore, might have to seek occupational opportunities outside of the village to maintain and supplement the family economy, thereby providing funds to purchase the required food. Family’s land accessibility is hypothesized to contribute to labor migration. “Family land accessibility” in this research is not designed to measure population pressure on the land, nor is it treated as an outcome of labor migration. Therefore, this research does not focus on or measure how labor migration relieves pressure on land. In other words, when the landholding for labor input and the landholding for family consumption are operationalized, family members in the labor force residing outside of the village are included in the denominators for measuring landholding per working family member and landholding per family member. 5.2.3 Analysis and Discussion To examine the second and third hypotheses, the Pearson’s correlation coefficient is adopted, which is also called the standardized regression coefficient and indicates the direction and strength of the association between two variables (Agresti and Finlay, 1986; McTavish and Loether, 1988). The Pearson’s correlation coefficients displayed in Table 193 5.4 demonstrate that in 1965 labor migration was significantly associated with landholding per working family member, but not with landholding per family member. Statistically, one standard deviation increase in the landholding per working family member was associated with 0.22 standard deviation decrease in the number of labor migrants. One standard deviation increase in landholding per family member decreased 0.15 standard deviation in the number of family members in the labor force residing in urban cities. These negative figures suggest that the more land a family has accessible, the fewer family members in the labor force who reside outside of the village. A similar pattern is found in the 1979 data. A one standard deviation increase in landholding per working family member and in landholding per family member was associated with 0.19 and 0.13 standard deviation decreases in the number of labor migrants, respectively. Nevertheless, the associations are not statistically significant. Table 5.4 Correlation between Accessibility to Land and Labor Migration In Hsin-Hsinfi1965 and 1979 1965 1979 r N r N Landholding per working family member -.22" 81 -. 19 72 Landholding per family member -. 15 81 -. l 3 72 "rsignificant at level of .01 Wiigni‘ficant at level of .05 ' sigrlficant at level of .10 Examining the second (H2) and third (H3) research hypotheses, there are almost no significant relationships between family landholding and labor migration in 1965. The only exception is the negative association found between landholding per working family member and number of labor migrants. The relationship was significant at or=0.05. This 194 significantly negative correlation supports the second research hypothesis (H2) which is, “the smaller the landholding per family member in the labor force, the more labor migrants the family has.” In 1965, the adoption of labor migration was significantly related to the amount of land available per family member in the labor force. Hsin-Hsing families with larger land parcels per working member had fewer labor migrants than those with smaller land parcels per working family member (see Table 5.4). This means that larger land parcels had a negative effect on migration to cities to seek work, while smaller land parcels per working family member had a positive effect on movement to cities to seek jobs. I Note, however, that all Pearson’s correlation coefficients were in the same direction and were similar in terms of strength; they range from -.13 to -.22. The differences among the four different Pearson’s correlation coefficients are slight. The significant association between landholding per working family member and the numberof family labor migrants in 1965 could be a result of the combination of a slightly larger correlation coefficient and a larger sample size in that year than in 1979.61 Nevertheless, labor migration was weakly associated with landholding per working family member in 1965, although the relationship is statistically significant. ‘When the Pearson correlation coefficient is -0.22, the R-square is about 0.048, which indicates that landholding per 6‘ For this analysis, “family” is the analytic unit. The data consist of 81 families in 1965 and 72 families in 1979. 195 working family member only explains 4.8 percent of the variation of labor migration in 1965. In sum, the 1965 data demonstrate a significant association between landholding for labor input and labor migration, but with a very slight difference from the insignificant association between landholding per family member and labor migration. An increase in landholdings per working family member reduced the number of family members in the labor force residing outside of the village. Nevertheless, in terms of the strength of association, the significant Pearson correlation coefficient is not much different from the insignificant one in 1979. In addition, the proportion of the variance in the number of _ family members in the labor force residing outside of the village explained by landholding per working family member in 1965 was very small. Therefore, a family’s accessibility to land seemed to have a weak relationship with the labor migration of Hsin- Hsing villagers in the 19603. The perspective viewing migration as a family sustenance strategy argues that migration acts as a mechanism to maintain the balance of adequate labor input opportunities and adequate production for family consumption (Grigg, 1980; Guest, 1989; Wood, 1981). Certainly this perspective strongly assumes that there is a negative relationship between family landholding and migration from rural areas. This pattern is not obvious in the 1979 data. There is no significant association between a family’s accessibility to land and labor migration. The number of labor migrants a family had was not significantly associated with the land a family had available for either labor input or family 196 consumption. Therefore, the 1979 data do not support either the second or the third research hypotheses. This seems to suggest that, in 1979, family differences in numbers of family members in the labor force residing outside of the village were not necessarily related to differences in the amount of land accessible to domestic units. The weak and/or insignificant associations between family’s land accessibility and the number of family members in the labor force residing outside of the village, however, do not necessarily imply that a family’s access to land had no influence on labor migration. Statistically, the Pearson’s correlation coefficients show that the different numbers of family members in the urban labor force were not significantly associated with differences in land accessibility. In this research, family’s land accessibility represents the differences in the amount of land available to individual families in Hsin-Hsing. It does not tell us whether a family had adequate or inadequate family land. The variation in family’s land accessibility between families was very small. Therefore, the insignificant associations show only that differences in family land accessibility could not explain why Hsin-Hsing families had different numbers of labor migrants. Nevertheless, the occurrence of labor migration in the village still was related to “inadequate family land accessibility.” In other words, families with less land available might not necessarily have had more family members in the urban labor force than those with more land available. Land in the village, however, was ubiquitously inadequate. The movement to cities of family members in the labor force was prevalent in 1965, regardless of the quantity of family land available as a comparison of Hsin-Hsing’s holdings to that of Taiwan as a whole show. Nationally, each family had an average of 0.95 chia of cultivated land, while 197 Hsin-Hsing families had only 0.53 chia (see Section 4.4.1 on the agriculture sector in Chapter IV). 5.3 Local Labor Market Participation and Migration 5.3.1 Theoretical Position and Hypothesis Structuralists argue that migration decisions are not made in a vacuum (Amin, 1974; Goldscheider, 1987; Massey, 1990a). Individual migration decisions are strongly influenced by factors in the larger environment. The previous section of this chapter merely discussed the relationship between migration and accessibility to land. Other structural factors influencing migration must also be examined. To solve the problems of a surplus labor force and inadequate agricultural production for agrarian families, the adoption of out-migration and a search for job opportunities in local labor markets are two additional strategies for most families. The opportunity to participate in the local wage labor market is one of the most important structural factors influencing migration decisions (Wood, 1981). In rural areas, families with inadequate land for labor input or family consumption must sell their labor for wages to maintain the family’s sustenance. The adoption of labor migration is not necessary until local employment opportunities are exhausted in the home area (Wood, 1981). This view would suggest that a negative relationship exists between finding employment in the local labor market and labor migration. Those who are unable to find work in the local labor market would be more likely to migrate than those who are able to find work in the local labor market. The research hypothesis is: 198 H4: Families with lower participation rates in local labor markets are more likely to have family members migrating than are families with higher local labor market participation rates. 5.3.2 Measurement This hypothesis suggests a negative relationship between participation" in local labor markets and the adoption of labor migration. Specifically, this hypothesis proposes that a family with a lower rate of its family members participating in local labor markets would have more family members out-migrating to seek job opportunities than would a family with a higher local labor market participation rate. The local labor market participation rate is measured in two ways: a family’s local labor- force participation rate and the proportion of family members working locally for pay (see Chapter III). A family’s local labor-force participation rate is the number of family members in the labor force holding occupations in the Hsin-Hsing area divided by the total number of family members in the labor force multiplied by 100. The denominator includes family members in the labor force holding paid or unpaid jobs in the Hsin-Hsing area and migrant family members working elsewhere. The proportion of family members working locally for wages is the number of family members working locally for pay divided by the total number of family members multiplied by 100. The group of family members working locally for wages excludes self-employed farmers and housekeepers. 62 Theoretically, there is a mutual relationship between labor migration and local labor market participation. Labor migrants could send remittances home, thereby reducing the necessity for family members to seek paid jobs locally. However, this research primarily examines how local labor market participation influences labor migration. In other words, in this research, family members’ participation in local labor markets is measured to study how families adopted migration as a strategy continue to the next page... 199 Measuring families’ participation in local labor markets demonstrates the input of their labor power and the sale of their labor to generate monetary income for family consumption. 5.3.3 Analysis and Discussion As shown in Table 5.5, labor migration was significantly and negatively associated with a family’s local labor-force participation rate at 0t=0.01. A one standard deviation increase in a family’s local labor-force participation rate was associated with a 0.70 standard deviation decrease in number [of family members in the labor force residing outside of the village in 1965. There was no significant relationship between the number of family members in the labor force residing outside of the Hsin-Hsing village and the proportion of family members working locally for paid wages in 1965. The Pearson’s correlation coefficients shown in Table 5.5 also demonstrate the significant and negative association between labor migration and local labor market participation for 1979. In 1979, a one standard deviation increase in a family’s local labor-force participation rate was associated with a 0.83 standard deviation decrease in the number of labor migrants. Further, a one standard deviation increase in the proportion of family members working locally for paid wages led to a 0.35 standard deviation decrease in the number of family members in the labor force residing outside of the village. ¥ in response to the likelihood of family members participating in local wage labor markets, which is primarily determined by the local economic structure. 200 Table 5.5 Correlation between Local Wage Labor Markets and Labor Migration in Hsin- Hsing: 1965 and 1979 1965 1979 r N r N F amily's local labor-force participation rate --.70 ... 82 -.83 ... 73 Proportion of family members working locally for paid -.05 82 -.35 ... 73 wajes ... significant at level of .01 .. significant at level of .05 ' significant at level of . 10 The fourth research hypothesis is supported by the 1965 data when a family’s participation in the local labor market is measured as a family’s local labor-force participation rate. The 1965 data do not support the research hypothesis if a family’s participation in the local labor market is measured as the proportion of family members working locally for wages. Nevertheless, regardless of how a family’s participation in the local labor market is measured, the fourth research hypothesis -- families with lower participation in local labor markets are more likely to have family members migrating than are families with higher local labor market participation rates -- is supported by the 1979 data. In general, labor migration was negatively associated with obtaining jobs locally. The higher a family’s local labor-force participation rate, the fewer the number of family members in the labor force residing outside of the village. In 1979, the number of family members in the labor force residing in urban cities was also associated with the proportion of family members working locally for wages. The higher the proportion of family members working locally for pay, the fewer the number of people deployed as labor migrants. 201 Nevertheless, the data in Table 5.5 raise a question: Why was labor migration not associated with the proportion of family members working locally for wages in 1965? The answer to this question lays in the fact that during the 19603, the primary economy in Hsin-Hsing Village was agriculture, and only a few non-agricultural jobs were available locally. Most villagers worked only on their family land. In Table 5.6, we see that of the 252 villagers in the labor force residing in Hsin-Hsing in 1965, 134 villagers, accounting for 53.2 percent of non-migrants in the labor force, were working as self-employed farmers. Seventy-three female villagers (50.0%) reported their primary occupation as housekeepers. Among this group of women, six reported working on-farm as their secondary occupation, while three reported working off-farm as their secondary occupation. Only 33 villagers (13.1%) were self-employed off-farm workers, workers in family enterprises, farm laborers, and off-farm workers laboring for income.63 It appears that the opportunity to participate in the local labor market to generate a monetary income for family consumption was small in the 19603. The few paid wage opportunities available led to the limited number of people earning a monetary income locally in 1965. ‘ 63 Although family workers in 1965 rarely received wages for their labor, their families benefited economically from their labor input. In other words, family workers contributed to an increase in their family’s monetary income through their unpaid labor. Thus, they are considered to bring monetary income to their families indirectly, and, accordingly, treated as paid workers. _ 202 Table 5.6 Labor Force and Local Labor Market Participation in Hsin-Hsing by Family Types: 1965 and 1979 Male Female Total N % N % N % 1965 Self-employed farmer 79 74.5 55 37.7 134 53.2 Housekeeper 0 0.0 73 50.0 73 29.0 Self-employed (off-farm worker) 7 6.6 l 0.7 8 3.2 Family worker 0 0.0 l 0.7 l 0.4 People working for wages. 9 8.5 15 10.3 24 9.5 Military 8 7.5 -- -- 8 3.2 Unemployed 3 2.8 l 0.7 4 l .6 People in Labor Force 106 100.0 146 100.0 252 100.0 1979 Self-employed farmer 35 32.1 25 22.9 60 27.5 Housekeeper l 0.9 39 35 .8 40 l 8.3 Self-employed (off-farm worker) 20 18.3 4 3.7 24 l 1.0 Family worker I 0.9 6 5.5 7 3.2 People working for wages. 33 30.3 35 32.1 68 31.2 Military 17 15.6 - -- 17 7.8 Unemployed 2 l .8 0 0.0 2 0.9 People in Labor Force 109 100.0 109 100.0 218 100.0 People working for wages include farm laborers and off-farm workers. Theoretically, labor migration should be negatively associated with local work opportunities. The weak and insignificant association between labor migration and the proportion of family members working for wages evidenced in the 1965 data, however, should not be viewed as local employment having no bearing on a family’s migration decision. The weak and insignificant association in the 19603 was due to the rural economic structure in which agriculture production occupied villagers’ daily life and local off-farm wage occupations were not readily available. In the 19603, most people in the labor force stayed in the village to farm their family land because farming demanded a great amount of human labor. Only 49 male and 16 female villagers from Hsin-Hsing, accounting for 20.5 percent of villagers in the labor force, worked in cities in response to 203 poor employment opportunities at home, and they tended to be oscillating migrants.64 Most did not permanently settle in their destinations. Rather, they, especially men, moved back and forth between their destinations and Hsin-Hsing in accord with the exigencies of the agricultural cycle in the village. Rural industrial development in the 19703 created more paid work for Hsin-Hsing villagers locally than had been available in 1965. In 1979, the number of villagers who resided in Hsin-Hsing and worked for monetary incomes increased from 32 (12.7%) in 1965 to 92 (45%). Of these 92 villagers, 24 (11.7%) were self-employed workers and 68 (33.3%) sold their labor for wages (see Table 5.6). Concurrently, agricultural production became less crucial as the source of family income than it had been in the 19603. Agricultural profits simply could not compare with the income that could be earned from off-farm employment (B. Gallin and R. S. Gallin, 1982a). Along with the change in the economic structure of the village, the number of villagers who reported themselves as self-employed farmers dramatically decreased from 134 (53.2%) in 1965 to 60 (27.5%) in 1979 (see Table 5.6). By the late 19703, improvements in agricultural technology made work off-farm possible for Hsin-Hsing villagers. Some farming work that had been dependent on human labor in the 19603 (e. g., preparing the land and transplanting seedlings) was done with modern agricultural machines in the late 19703. Tube wells were dug in fields and diesel engines 6" As B. Gallin and R. Gallin (1974: 344) point out, “men move initially without their wives or families and send part of their eamingsback to their families in the village. At the outset, they return often to continue to the next page... 204 and electronic motors were used for irrigating fields. The widespread adoption of herbicides and pesticides reduced the need for human labor (B. Gallin and R. S. Gallin, 1982a).65 But, as pointed out previously, totally abandoning the family farming land was still impractical. The creation of off-farm occupations in the late 19703 made it possible for villagers to earn an income by participating in the non-agricultural labor market in addition to caring for their family land. But, in general, farming was left to older or retired family members, while the younger generation participated in local off-farm labor markets (R. S. Gallin, 1984). Therefore, the large number of villagers participating in labor markets in Hsin-Hsing and its surrounding area in 1979 not only reflects the high proportion of family members working locally, but also the high proportion of family members working locally to generate monetary income. The increase in local opportunities led to a significant and negative association between the number of family members in the labor force residing outside of the village and a family’s local labor market participation in the late 19703. In summary, there were different associations between labor migration and the proportion of family members participating in local wage labor markets in 1965 and in 1979. While the association was insignificant in 1965, it was statistically significant at 0t=0.01 in 1979. In contrast, data in both time periods reveal strong negative relationships between the village, to plant and harvest crops or to observe festivals or rituals.” ‘5 Meanwhile, the traditional system of exchange labor for harvesting was abandoned, and it was replaced by hiring groups of professional laborers to reap the crop. 205 migration and the proportion of family members in the labor force holding either paid or unpaid jobs in the local area. The opportunity to hold jobs in Hsin-Hsing by family members in the labor force reduced the number of family members moving to cities to seek wage-eaming opportunities. The industrial development surrounding the Hsin-Hsing area possibly changed the villagers’ migration behaviors. During the 19603, migration was adopted by those families which were able to or had to send family members to work in the cities. Rural industrialization created more wage employment opportunities in the local area for villagers by the 19703. As mentioned previously, seeking employment opportunities locally or in cities was an option for villagers to cope with the problems of surplus labor and inadequate family production. The newly created waged job opportunities in the 19703 not only provided villagers more paid employment opportunities than were available in 1965, but also the opportunity to increase their family incomes. In 1979, the number of family members relying on labor migration was significantly and negatively associated with the proportion of family members in the labor force obtaining jobs in the Hsin-Hsing area as well as the proportion of family members obtaining paid jobs in local labor markets. 5.4 The Best Predictor of Labor Migration Finally, multiple regression is applied to examine which factor(s) is/are the best predictor(s) of the number of family members in the labor force residing outside of the village. Statistically, multiple regression is appropriate under a situation where “there are usually several independent variables that have an effect on any dependent variable, and 206 those variables are usually correlated/inter-related among themselves” (Agresti and Finlay, 1986: 316). In this research, multiple regression is more appropriate than a bivariate analysis, such as Pearson’s correlation coefficient, to reveal which factor(s) is/are better predictor(s) for ganging the number of family members in the urban labor force. In addition, the multiple regression method allows the analysis of partial relationships between labor migration and one of the influential factors, controlling for other influential factors. Further, the standardized partial regression slopes provide a means of comparing the effect of independent variables (McClendon, 1994). Table 5.7 shows that, in both time periods, a family’s local labor-force participation rate was a strong predictor of the number of family members in the labor force residing outside of the village, especially in 1965. The standardized regression coefficient (or standardized partial regression slope) indicates that in 1965 one standard deviation increase in“family’s local labor-force participation rate” led to a decrease of 0.79 standard deviation in the number of family members in the labor force residing outside of the village, controlling for other factors such as family type, proportion of family members working locally for wages, landholding per working family member, and landholding per family member. In other words, controlling for other factors, there is a negative association between “family’s local labor-force participation rate” and labor migration; the increase in “family’s local labor-force participation rate” caused the decrease in the number of family members in the labor force residing outside of the village. In 1979, the influence of a family’s local labor-force participation rate on labor migration was weaker than it was in 1965. Nevertheless, it remanded strong. Controlling 207 for other factors, one standard deviation increase in “family’s local labor-force participation rate” in 1979 caused a decrease of 0.57 standard deviation in the number of family members in the labor force residing in urban cities. Table 5.7 Multiple Regression for Labor Migration, Family Type, Family’s Local Labor Force Participation Rate, Proportion of Family Members Working Locally for Paid Wages, Landholding Per Working Family Member, and Landholding Per Family Member for Hsin-Hsirm Village: 1965 and 1979 1965 1979 B Beta B Beta Family type Joint (reference). Stem (reference)+ -2. 17 -.54 ... Conjugal -33 -. l6 “ -223 -.60 Family's local labor-force participation rate -.04 -.79 ... -.04 -.57 ... Proportion of family members working locally for paid .00 .02 -.00 -.02 wages Landholding per working family member -.73 -.09 -.55 -.05 Landholding per family member .50 .04 1.01 .05 Constant 3.88 6.19 R-smrare .66 .81 N= 79 72 For the analysis of the 1979 data, joint families are used as the reference group. 4. For the analysis of the 1965 data, stem families are used as the reference group. In the analysis of the 1965 data, two joint families are excluded due to a small number of cases. However, these two joint families are not combined with 36 stem families. The analysis with the combination of stem and joint families produces a very different result from the one shown in this table. Stern and joint families in the 1965 data seemed to have different characteristics. It would be inappropriate to combine them as a category for the multiple reggssion analysis. Additionally, the results of the multiple regression show that family type was also a significant predictor of labor migration, especially in 1979. The multiple regression model for 1965 includes 79 conjugal and stem families while that for 1979 consists of 72 families. Compared to stem families, in 1965, conjugal families had fewer family members in the labor force residing outside of the village. Controlling for other factors, in 208 1965, conjugal families had 0.33 family labor migrants fewer than stem families. The influence of family type on the number of family members in the labor force residing in urban cities increased in 1979. In 1979, conjugal and stem families had significantly fewer numbers of family members who were in the labor force and residing outside of the village than did joint families. The multiple coefficient of determination (R-square) for the 1965 data is 0.66, which means that 66 percent of the total variation in “the number of family members in the labor force residing outside of the village” is explained by the simultaneous predictive power of all independent variables, including family type, family land accessibility, and participation in the local labor market. In 1979, the proportion of the total variance in the number of family members in the urban labor force accounted for by the combination of independent variables increased to 81 percent. Interestingly, the multiple regression analysis demonstrates results different fi'om those of the bivariate analysis, especially the association between family type and the number of family members in the labor force residing outside of the village. The multiple regression analysis examines the relationships between labor migration and one of the influential factors, controlling for other independent variables. Family type is significantly associated with labor migration when other influential factors are held constant. In this case, among families with the same access to land and local labor participation rates, larger families had more family members in the labor force residing outside of the village than did smaller families. 209 In sum, the results of the multiple regression analysis reveal high multiple coefficients of determination which demonstrate that the variations in the number of family members in the urban labor force were accounted for by the combination of the following independent variables: family type, family land accessibility, and family’s participation rate in the local labor market. Further, family type and local labor market participation rate were more significantly associated with the number of family‘members in the labor force residing outside of the Hsin-Hsing area than was family’s land accessibility regardless of its measurement. In other words, in both time periods, while family type and family’s participation rate in the local labor market were significant associated with labor migration, there was no significant association between family’s land accessibility and labor migration, even when other independent variables were held constant. In addition, family type was not simply associated with the number of family members in the urban labor force, because of its family size or structure. Larger families did not necessarily have more labor migrants than smaller families. The association between family type and labor migration was also influenced by factors related to the local economic structure, such as the availability of paid work in the local labor market. 5.5 Summary This chapter demonstrated how family type, family access to land, and participation in the local labor market were associated with the labor migration of Hsin-Hsing villagers in 1965 and 1979. The associations between labor migration and the factors affecting it changed over time according to Hsin-Hsing’s economic structure. 210 In both 1965 and 1979, the associations between labor migration and the villagers’ local labor market participation were determined by the local economic structure. In the 19603, agriculture was the primary means of production in the Hsin-Hsing area. The amount of farm land available to input labor was responsible for the villagers’ local labor market participation. By the end of the 19703, a rural industrial zone established near the village as well as local industrial development, influenced the work patterns of Hsin-Hsing villagers. More non-agricultural job opportunities became available to the villagers and, in 1979, villagers participating in local non-farm wage labor markets accounted for a great proportion (45.4% or 99 out of 21 8) of the labor force working locally. More villagers worked locally for monetary income in 1979 (45.4%) than in 1965 (5.2%). The land consolidation program of the late 19603 and development of new agricultural technologies increased the likelihood that villagers were both self-employed farmers and paid off-farm workers by the late 19703. In the 19603, farming was labor-intensive, and there were few off-farm employment occupations available in the Hsin-Hsing area via which to diversify family income sources or to supplement the family economy. Villagers had to go out of the local area to seek jobs. When more off-farm jobs became available in the late 19703, villagers were able to diversify family income sources and increase the unit’s income by taking waged jobs locally. They could simultaneously work on their family land because innovations in agricultural production reduced the need for intensive human labor. 211 In sum, the analyses in this chapter show that family’s local labor-force participation rate, which is measured as the number of family members in the labor force holding occupations in the Hsin-Hsing area divided by the total number of family members in the labor force multiplied by 100, was more influential in labor migration than other factors. The results of the multiple regression analysis show that when other factors are controlled, a family’s local labor-force participation rate had the most influence on labor migration in 1965. It was also very influential in 1979. In 1965, the bivariate analysis shows that family type and the number of family members in the labor force residing in urban cities were not significantly associated with each other. The results of the multiple regression, however, show that in both research periods, family type had significant associations with a family’s labor migration whenother factors are controlled. The results demonstrate that the simple type of families had fewer family members who were labor migrants than families of the complex type, controlling for other independent factors. In addition, the influence of family type on the number of family members in the labor force residing outside of the village increased over time. With the exception of the relationship between family land accessibility and labor migration, the results basically support the theoretical framework. In addition, the results demonstrate that migration is a complex process. Explanations based solely on the bivariate relationship analyses are too simple and incomplete to reveal how and why migration is used to improve a family’s economy. In Chapter III, I discussed how migration is theoretically related to each of the individual factors considered in this 212 chapter. Their relationships are conditional, however. For example, migration theoretically should be positively related to family type. Nevertheless, labor migration is not necessary or possible, when few family members are in the labor force. In an agrarian society, farming is the primary source of family income. A large number of family members in the labor force leads to a high demand for land to input labor and to produce resources for family consumption. In addition to the structure of families, the adoption of migration is strongly influenced by the local economic structure within which villagers are embedded. When a local economic structure is able to provide local people the job opportunities they desire, out- migration will decline. In sum, the influencing factors of rural-to-urban migration are inter-locked. Research frameworks should not be based on the bivariate relationships, but rather need to be multi-variate. This chapter has primarily examined the Hsin-Hsing data at the family level. That is, the “family” in this chapter is mainly characterized in terms of its type, land accessibility, and participation in local labor markets. The “family,” however, can also be characterized in terms of other characteristics, such as family members’ genders and ages. In the next chapter, this dissertation will move forward to discuss migration at the individual level. The data analyses and interpretations of labor migration will incorporate villagers’ genders, ages, and education levels to discuss how family power dynamics are implicated in the process of migration. 213 CHAPTER VI MIGRATION AND FAMILY POWER DYNAMICS IN HSIN-HSING VILLAGE Chapter V examined the associations between labor migration and influential factors such as family type, family landholdings, and local labor market participation. The analyses were at the family level. This chapter will focus on the individual level, examining the associations between migration and family power dynamics in Hsin-Hsing Village. I begin with a brief review of the individual perspective of migration theory, arguing that migration decisions are affected by the characteristics of potential migrants -- characteristics that are implicated in power dynamics within the family. I then summarize the different characteristics of Hsin-Hsing people residing in urban cities and living in the village. In the sections that follow, I provide a discussion of family power dynamics during the process of migration by examining who migrated and who did not, and exploring the association between employment status and migration. Family power dynamics are analyzed (1) in terms of relations among women and (2) in terms of male- female relations. 6.1 Overview of the Individual Perspective of Migration Theory The individual perspective within migration theory, especially the individual cost-benefit model, suggests that migration is the outcome of a rational evaluation of the costs and benefits of movement (see Massey, 1990; Sjaastad, 1962; Todaro, 1976, 1980). The expected net return of migration has been used methodologically as an indictor to predict if a potential migrant will choose to move or to stay. Potential migrants will choose to 214 move if the expected net return of migration is positive; if it is negative, potential migrants will choose to stay (Bowles, 1970; DaVanzo, 1981; Guest, 1989; Harris and Todaro, 1970; Sjaastad, 1962; Todaro, 1969, 1976, 1980; Wood, 1981). If the expected net return of migration is zero, potential migrants are indifferent about either migrating or staying. Todaro (1969, 1976) proposes that the expected net return is a function of expectations about urban-rural income differences and the likelihood of obtaining an urban job. The possibility of potential movers obtaining jobs in modern urban sectors is a crucial element in the decision-making process to migrate. Because of the importance of obtaining an urban job in the decision-making process, the individual perspective emphasizes the different characteristics (e. g., age and gender) and human capital (e. g., education) of migrants, in accord with the neoclassical microeconomic theory linking migrants’ characteristics with the probability of obtaining a job. The individual perspective of migration theory has been applied prevalently in empirically tests (see Browning, 1969; Chang, 1979; Chiang, 1978; Ladinsky, 1967; Li, 1974; Liao, 1977; Liu, 1993; Long, 1973, 1992; Speare, 1974; Tsai, 1978; Yin, 1978; Zachariah, 1966). Results suggest that personal characteristics determine the decision of migration. However, the empirical research on migration in Taiwan usually draws different conclusions. As shown in Chapter IH, researchers (e. g., Chang, 1979; Chiang, 1978; Li, 1974; Liao, 1977; Liu, 1993; Speare, 1974; Tsai, 1978; Yin, 1978) identify different personal characteristics of movers involved in the migration process on Taiwan. 215 These different conclusions suggest that while the migration decision may be an outcome of a rational evaluation of the “expected returns” of movement, this emphasis is insufficient to fully explain the migration decision-making process. Migration may also be a consequence of the dynamics inherent in a family power hierarchy, which is shaped by the interaction of age and gender. While personal characteristics determine the likelihood of migration, the fact is that migration reflects a family power hierarchy, ' which is based on personal characteristics such as age and gender. As Wolf (1991) argues, “household[/family] strategies necessarily embody relationships of power, domination, and subordination if a strategy is formulated by the decision maker(s) and successfully executed by those for whom decisions are made” (p. 32). In this research, the individual perspective of migration provides a means to inspect migration and to identify characteristics of migrants and non-migrants in Hsin-Hsing Village in the mid- 19603 and the late 19703. I move beyond this focus, however, by emphasizing the intersection of age and gender, and exploring how family power dynamics are implicated in the decision-making process of migration and the employment status of migrants and non-migrants. 6.1.1 Characteristics of Villagers in Urban Cities As reported in Chapter IV, the number of Hsin-Hsing villagers residing in urban cities increased from 79 in 1965 to 157 in 1979. During this same time period, among villagers residing outside of Hsin-Hsing Village, the proportion accounted for by women rose from 29.1 percent (23 out of 79) to 45.5 percent (71 out of 157). In both research periods, there were more men than women residing outside of Hsin-Hsing Village. Additionally, the 216 proportion of villagers in urban cities accounted for by those in the labor force66 declined from 82.3 percent (65 out of 79) in 1965 to 56.5 percent (87 out of 154) in 1979 (see Table 4.8).67 The migrant and non-migrant populations had different age structures. In 1965, the mean age of migrants was 26.3 years while that of non-migrants was 25.1 years. In 1979, due to the addition of a young generation to the group of villagers residing in urban cities, the mean age of migrants decreased to 20.5 years. In contrast, the mean age of rural villagers increased to 29.9 years. This increase reflected the geriatrification of the rural labor force. As the 1979 data showed in Chapter IV, the mean age of villagers in the rural labor force was 40.8 years, while that of their counterparts in the urban labor force was only 30.0 years. Further, as the data in Table 6.1 reveal, in the mid-19603, Hsin-Hsing’s migrants were more likely to be in the urban labor force than were those in the late 19703. The 1965 data reveal that 79 villagers resided outside of the village, and that most of these migrants (78.5% or 62 out of 79) were of working age (i.e., 15-64 years). Male villagers accounted for more than 70 percent (46 out of 62) of villagers in the urban labor force. By contrast, 66 Theoretically, everyone over a certain age is capable of getting a job. In this research, people in the labor force include those who are capable of participating in the wage labor market. This group of people are usually ages 15-64. However, as members of the labor force, they can be paid or un-paid workers or unemployed. In this research, people in the wage or paid labor force are those who participate in the labor market for income. Therefore, people in the labor force are not necessarily in the wage or paid labor force. For example, women who report themselves as housekeepers are in the labor force, but they are not in the labor force for income. The reason I make this distinction is to acknowledge that women’s domestic work, though unpaid, is labor. 217 women migrants accounted for less than 30 percent (16 out of 62) of the total. On average, women migrants were younger (22.3 years) than were their male counterparts (26.1 years) (see Table 6.2). While in 1979 more than 40 percent (19 out of 46) of male migrants were married, only about 30 percent (5 out of 16) of their female counterparts were married. 67 Because of missing data, the number of Hsin-Hsing villagers residing in urban cities is sometimes reported as 157 and sometimes reported as 154. 218 Table 6.1 Villagers’ Migration Status by Gender, Marital Status, Hsin-HsingVillage: 1965 and 1979 Male Female Un- Sub- Un- Sub- Migration Status Married married“ total Married marriedt total Total 1965 Non-migrant not in labor 7 88 95 3 95 98 193 force (3.6%) (45.6%) (49.2%) (1.6%) (49.2%) (50.8%) (100.0%) Non-migrant in labor 72 34 106 89 57 146 252 force (28.6%) (13.5%) (42.1%) (35.3%) (22.6%) (57.9%) (100.0%) Non-labor migrant (no 0 10 10 0 7 7 l7 ‘ occupation) * (58.8”) (58.8%) (41.2%)‘ (41.2%) (100.0%) Labor migrant 19 27 46 5 l l 16 62 (30.6%) (43.5%) (74.2%) (8.1”) (17.7%) (25.8”) (100.0%) Total 98 159 257 97 170 267 524 (gift) (30.3%) (49.0%) (18.5%) (32.4%) (51.0%) (100.0%) 1979 Non-migrant not in labor 7 88 95 3 70 73 168 force (42“) (52.4%) (56.5%) (1.8%) (41.7%) (43.5%) (100.0%) Non-mi grant in labor 71 3 8 109 79 30 109 218 force (32.6%) (17.4%) (50.0%) (36.2”) (13.8%) (50.0%) (100.0%) Non-labor migrant (no 0 38 38 0 29 29 67 occupation) * (56.7”) (56.7%) (43.3%) (43.3%) (100.0%) Labor migrant 28 18 46 26 15 41 87 (32.2%) (20.7%) (52.9%) (29.9%) (17.2%) (47.1%) (100.0%) Total 106 182 288 108 144 252 540 (19.6%) (33.7%) (53.3%) (20.0%) Q67”) £167”) (100.0%) The category of “unmarried” includes those who were single, divorced, separated, and widowed. These seven women include three retired women at their 603 and 703, one student in her 203, and three preschoolers. 219 Miggants who did not have a job could be retired, students, or too young be in thelabor force. Table 6.2 Mean Ages of Villagers by Gender, Marital and Migration Status, Hsin-Hsing Village: 1965 and 1979 Male Female Un- Sub- Un- Sub- Migration Status Married marriedl total Married marriedT total Total 1965 Non-migrant not in labor 72.4 9.0 13.7 67.0 1 1.4 13.1 13.4 force Non-migrant in labor 41.8 24.3 36.2 39.9 22.1 33.0 34.3 force Non-labor migrant (no -- 12.8 12.8 -- 35.4‘ 35.4 21.6 occupation) + Labor migrant 35.1 19.5 26.1 40.4 18.2 22.3 25.1 Total 42.7 14.3 25.1 40.3 16.4 25.1 25.1 1979 Non-migrant not in labor 69.4 9.8 14.2 73.3 15.8 18.2 15.9 force Non-migrant in labor 47.2 27.4 40.3 45.3 31.0 41.4 40.8 force Non—labor migrant (no -- 7.6 7.6 -- 7.9 7.9 7.8 occupation) + _ Labor migrant 34.4 25 .5 30.9 31.0 23.1 28.1 29.6 Total 45.3 14.6 25.9 42.8 18.2 28.6 27.2 T '_’-- u The category of “unmarried” includes those who were single, divorced, separated, and widowed. These seven women include three retired women at their 603 and 703, one student in her 203, and three preschoolers. Migrants who did not have a job could be retired, students, or too young to be in the labor force. 0 In the late 19703, villagers who were too young (i.e., under the age of 15 years old) to be in the labor force or who were students accounted for a large proportion of Hsin-Hsing people residing in urban cities, making up 43.5 percent (67 out of 154) of migrant villagers (see Table 6.1). Those who resided in cities in 1979, but were not in the labor force, were primarily youngsters, as their mean age (7.8 years) reveals (see Table 6.2). Those who were in the urban labor force made up 56.5 percent (87 out of 154) of villagers residing outside of the village. At the same time, the gender difference among villagers in the urban labor force was slight. While therewere 46 men in the urban labor force (52.9% or 46 out of 87), 41 women accounted for 47.1 percent (41 out of 87) of 220 urban villagers who were of working age. In addition, most villagers in the urban labor force were married. Twenty-eight men (60.9% or 28 out of 46) and 26 women (63.4% or 26 out of 41) were married. As mentioned in Chapter IV, migrants, in general, had more education than non-migrant villagers. A comparison of the average years of educational attainment of villagers in the labor force shows that, in 1965, villagers in the urban labor force had an average of 5.38 years of education while villagers in the rural labor force had an average of 2.56 years of schooling. The implementation of a new government educational policy in 1968, however, narrowed the gap between migrants and non-migrants. In 1979, while villagers in the urban labor force had an average of 5.41 years of education, their rural counterparts had an average of 4.00 years, increasing from 2.56 years in 1965 (see Section 4.61). These statistics reveal that Hsin-Hsing villagers residing in urban cities had different characteristics in the mid-19603 and the late 19703. These different characteristics reflect their position in the labor force, educational attainment, and gender and marital status. Villagers in the labor force accounted for a larger proportion of migrants in urban cities in the mid-19603 than in the late 19703 (82.3% vs. 56.5%). Among female villagers in cities, the proportion accounted for by single women decreased from 68.8 percent (1 1 out of 16) to 36.6 percent (15 out of 41) between 1965 and 1979. Married women and youngsters increased dramatically among Hsin-Hsing villagers residing in urban cities in the late 19703. 221 Associated with the increase in married women and youngsters, the family structures of people residing in urban cities changed as well. In the mid-19603, solo migration was prevalent. Villagers usually lived alone at destination.68 The total number of migrant conjugal units was only 11. In 1979, there were 27 migrant conjugal units living outside of Hsin-Hsing Village, in which 114 villagers clustered. The increa3e in the number of migrant conjugal units reflected the settlement of early migrants, which led to family growth/maturation in cities. These findings on migrants’ characteristics are not exactly the same as the findings of past research on Taiwan (see Chang, 1979; Chiang, 1978; Li, 1974; Liao, 1977; Liu, 1993; Speare, 1974; Tsai, 1978; Yin, 1978). The different conclusions might reflect the dissimilar labor needs of different economic structures in the 19603 and 19703, and the unlike structures of families, which determine how many family members and who among them migrates. In Chapter V, I discussed the associations between family structure, local and national economic structures, and migration. In this chapter, I move beyond the traditional “cost-benefit” argument and consider the role of power dynamics in the migration process. To this end, I examine the association between migration and family power dynamics. Specifically, this chapter will provide a discussion of family power dynamics during the process of migration, by examining (1) who migrated and who did not, and (2) the association between employment status and migration. This discussion will focus on male-female power relations and on power relations among 68 Some men lived with co-villagers or kin (Gallin and Gallin 1974). 222 women. The purpose of this discussion is to encourage an expansion of the individual perspective of migration by considering why personal characteristics are implicated in the decision-making process of migration. 6.2 Theoretical Position and Hypotheses The individual perspective of migration theory suggests that personal characteristics determine decision-making about moving. Within the Taiwanese family framework, however, migration can be viewed not only as an outcome of personal characteristics, but also as a product of family power dynamics. Traditionally, age and gender shaped the authoritarian hierarchy, which guided inter-relationships among family members and the behavior of individuals in Taiwan. These power dynamics reflected Taiwan’s patrilineal kinship structure (R. S. Gallin, 1985). While migration is a rational reaction to the outcome of a “cost-benefit” analysis of movement, I argue that migration is also a consequence of the dynamics inherent in a family power hierarchy, which is shaped by the interaction of age and gender. Different migration behaviors thus are related to the way the intersection of age and gender position people within this hierarchy. Research has showed the association of migration with gender and age, respectively. Roos (1983) argues that gender differences between men and women lead to different possibilities of participating in labor markets. Other researchers (see Chang, 1979; Chiang, 1978; Li, 1974; Liao, 1977; Liu, 1993; Speare, 1974; Tsai, 1978; Yin, 1978) suggest that age determines the likelihood of migration to seek waged job opportunities. However, it is the intersection of gender and age that influences the decision to migrate as well as the employment status of a migrant at 223 destination. An individual’s gender as well as his/her age determines whether or not a villager migrates and what he/she does at destination. In other words, his/her gender and age simultaneously determine whether a migrant villager participates in the urban wage labor market. Men and women in rural Taiwanese families usually were, and continue to be, treated differently. Women generally had a lower status than men did. Women’s low family status was traditionally reinforced by patrilocal rules of residence. As R. S. Gallin (1984:385) explains, traditionally, when a woman married, she left her natal home to live as a member of her husband’s family. . .. [Therefore,] parents considered daughters a liability, household members who drained family resources as children and who withdrew their assets (domestic labor and earning power) when they married. Sons, in contrast, contributed steadily to the family’s economic security during its growth and expansion and provided a source of support for their parents in old age. As a consequence, “parents strongly preferred male children” (R. S. Gallin, 1984: 385). Taiwanese parents usually transferred all the family estate to their sons and, directly and indirectly, persuaded their daughters to give up their inheritance in favor of their brothers (Tang 1985). Due to their inferior status in the family, women, especially daughters who were seen as victims of the patriarchal kinship system by feminist researchers (R. S. Gallin, 1985; Gates, 1987; Greenhalgh, 1985, Wold, 1972), usually did not participate in migration 224 decisions or initiate the decision to migrate.69 Female migration to cities usually was not based on choice but rather on the will of other family members, particularly fathers. On the one hand, because parents were protecting their daughters’ presumed innocence, young women were not allowed to migrate without parental permission (Huang 1984). On the other hand, because Taiwanese parents traditionally thought daughters drained family resources and withdrew their labor power from their families when they married, they felt that they were not able to pay back the money the older generation invested in them before they married out (Greenhalgh, 1985). As a consequence, Taiwanese parents deployed their daughters into the labor market as early as possible, thereby enabling them to begin repaying their debt by improving and/or maintaining the family economy. In contrast, they kept their son(s) in school to increase their earning ability, thereby investing in the parents’ future, i.e., their life in old age. In addition to gender, the different status of members within a family was determined by age. Traditionally, the eldest man in the family had the highest status in Hsin-Hsing Village as well as in Taiwan as a whole. Moreover, older family members usually had a higher status than younger members in their families. Although the relationships among female members were very dynamic and complicated (see R. S. Gallin 1986), in general, mothers and mothers-in-law traditionally had a higher status than their daughters and daughters-in-law. Before their marriage, women were subordinated by their parents, 69 Although this research emphasizes the inferior status of unmarried daughters in the mid-19603 and the late 19703, later research shows that they, as well as married women, may benefit from participating in wage labor markets (see R. S. Gallin, 2001). Being an income earner can change a woman’s position in the family power hierarchy. 225 especially fathers. Once she married a woman was subordinated by her parents-in-law, especially her mother-in-law; “a women came to her husband’s home as a submissive, exploitable bride” (R. S. Gallin, 1986:38). A daughter-in-law traditionally was expected to assume the responsibilities of her mother-in-law, taking care of the house and its members. Mothers-in-law were allowed to enjoy the leisure time provided by this" division of labor. When agriculture was the primary source of the family economy, as it was in Hsin-Hsing in the mid-19603, a daughter-in-law was also expected to participate in agricultural production. When, in 1965, men had to seek wage employment opportunities in urban cities to improve and/or to maintain the family economy, daughters-in-law had to stay in the village, participating in agricultural production and taking care of their mothers-in-law and the domestic unit.70 Therefore, due to their subordinate status in relation to their mothers-in—law, daughters-in-law were less likely to be migrants than other family members. If married couples resided together in urban cities, the men usually participated in the waged labor market. In contrast, the women’tended to work without pay at jobs such as housekeepers or unpaid family workers. Both men and women had internalized the belief that “domestic work was women’s work” (R. S. Gallin, 1995:125) and that it was “natural for a man to earn money” (R. S. Gallin, 19952125). As a consequence, while married migrant men worked outside to improve and/or to maintain their family 7° Men traditionally held major responsibility for the land. When they migrated to cities to seek waged job opportunities, married women assumed major responsibility for this work in the absence of their husbands (R. S. Gallin, 1984). 226 economy, married migrant women were likely to stay at home to manage their homes and to take care of family members. In sum, while parents, especially fathers, sent their unmarried daughters to cities to seek waged employment to increase the family income, they kept their sons in school, thereby investing in their sons’ earning ability and thus the older generation’s security in old age. Unmarried women were theoretically a component of a family’s sustenance or mobility strategies but their input into these efforts was temporary. Married women also were a part of these strategies; because they were expected to assume the domestic responsibilities of their mothers-in-law, however, they were unlikely to participate in the urban wage labor force. When opportunities for waged jobs in the rural labor market were insufficient, family members in the labor force had to seek job opportunities outside of the rural area. Because men were socialized to earn money to support and/or maintain their families, they were likely to participate in the waged labor market. Therefore, unmarried women and married men were more likely to participate in the waged urban labor force than unmarried men and married women. Assuming married people are older than the unmarried, I hypothesize that: H5: Younger women are more likely to migrate to cities and to work for pay than are older women. H6: Among those who participate in urban wage labor markets, male migrants are older than female migrants. 227 In general, these research hypotheses focus on how age and gender are related to migration and to occupation at destination. Research hypothesis five (H5) examines the relationship between age and migration and employment status among female villagers. In other words, this research hypothesis examines if younger or older women are more likely (I) to move to urban cities and (2) to take employment for wages at destination.7| Because the hypothesis deals with two different phenomena -- i.e., (1) migration to cities, and (2) occupation at destination, I examine this research hypothesis in two stages. First, I explore if an increase in age reduces the likelihood of a woman being a migrant in urban cities among women of working age.72 Second, I explore (1) who is more likely to work for pay among female migrants of working age, and (2) whether female migrants ages 15-64 are more likely to work for pay than are their non-migrant counterparts. Research hypothesis six (H6) emphasizes the age differentials between male and female villager migrants participating in the urban wage labor market. Examining and discussing these research hypotheses will throw light on how power dynamics within families influenced migration and the employment patterns of rural migrants in Taiwan. In the absence of quantified data on the operation of power in the family, measures such as those used will allow inferences to be made about the exercise of power within the family, particularly given the theoretical discussion above. 7‘ Not every woman of working age is necessarily in the labor force. For example, a female high school student is working age, but in this research she is not considered to be in the labor force. 72 Students are excluded. 228 6.3 Measurement To examine research hypothesis five (H5), the first stage is to apply logistic regression to explore the likelihood of being a migrant among female villagers ages 15-64, who are in the labor force. A logistic regression model provides a good way to examine how various factors influence a binary outcome. Moreover, as an inferential statistical tool, logistic regression analysis can also be adopted as a tool to illustrate changes in the likelihood of migration along with increases in age. In the analysis, the independent variable is a villager’s age, which is measured as the difference between a person’s year of birth and the research year. Thus, for example, a woman born in 1925 will be 40 years old in 1965 and 54 years old in 1979. The dependent variable is the occurrence of migration, which is dichotomous. Please note that this analysis does not include students. Among the group of villagers included in the logistic regression analysis, those who resided in urban cities are coded “l,” and those who resided in Hsin-Hsing are coded “0” for the dependent variable.73 To examine whether age influenced the probability of female villagers being migrants, this analysis includes migrant and non-migrant women in the labor force.74 As the hypothesis proposes, younger women are more likely to migrate and to work for wages at destination than are older women. In the first stage of the analysis, I expect to see that, with an increase in women’s age, the likelihood of residing outside of the village 73 Female students are not coded “l ,” because they are not in the labor force. 74 All women in the labor force are included for the purpose of comparison. 229 becomes smaller. In other words, the proposed association between the likelihood of migration and age among female villagers is negative. In the second stage of the analysis of research hypothesis five (H5), contingency tables classify female villagers of working age by three characteristics: migration status, occupation, and living arrangement. Migration status includes two “categories: migrants and non-migrants. Occupation includes five categories: housekeeper, unpaid family worker, self-employed farmer, off-farm worker, and self-employed business owner. “Living arrangement” is categorized in two ways: living with a mother-in-law and living with preschooler(s). Each of these two categories is coded either ‘yes” or “no.” Contingency tables are created to examine women’s characteristics in terms of their occupation and living arrangement by their migration status. Analysis of variance (ANOVA) is used to examine research hypothesis six (H6). In the examination of H6, ANOVA is applied to detect if there is a significant difference in terms of age between men and women migrants participating in wage labor markets. As mentioned above, “age” is measured as the difference between the research year and the birth year. The analysis for this hypothesis includes male and female migrants working for pay in cities. Since the hypothesis proposes that male villagers in the urban wage labor force are older than are their female counterparts, -I expect to find the mean ages of male villagers in the urban wage labor market to be higher than are those of their female counterparts. 230 6.4 Analysis and Discussion 6.4.1 Age (and Migration among Female Villagers The fifth research hypothesis (H5) proposes that younger female villagers are more likely to move to urban cities and to work for pay than are older women. The intentions behind this research hypothesis include examining how age influences women’s migration and employment status, and demonstrating how age is related to family dynamics among women during the process of migration. Because the hypothesis deals with two different phenomena: migration to cities, and occupation at destination, in the analysis and discussion of this research hypothesis, I take a two-stage analysis. First, I examine if age is related to the likelihood of women of working age being migrants. If it is, the question becomes: How is female migration related to power dynamms among women and other L family members in the domestic unit? To answer this question, I discuss how migration is related to the employment status of women in the labor force. In other words, I examine whether or not women residing in urban cities were more likely to participate in the wage labor market than were women who lived in Hsin-Hsing Village. Furthermore, the results of logistic regression analysis are converted to predict the probabilities of migration among villagers of working age. While Table 6.3 demonstrates the statistical results of the logistic regression analysis, Figures 6.1A and 6.1B show the probabilities of being migrants along with increases in ages. The logistic regression results indicate not only that age was related to migration, but also that gender was related to likelihood of migration (see Table 6.3 and Figures 6.1A and 6. 13). Female villagers had much lower probabilities of being migrants than did their 231 male counterparts in the mid-19603. This finding is consistent with the statistics presented in Chapter IV and in section 6.1.1 of this chapter, which showed that more male than female villagers resided in urban cities, especially in the mid-19603. Age was significantly associated with the probability of migration in the mid-19603 and in the late 19703 (see Table 6.3). The association between age and migration was negative. With the increase in ages, villagers in the labor force became less likely to reside in urban cities. The logistic regression coefficients are converted into the probabilities of being migrants (see Figures 6.1A and 6.1B). These two figures illustrate that with an increase in age, the probabilities of movement to urban cities decreased for both male and female villagers, who were ages 15-64 in both the mid-19603 and the late 19703 Table 6.3 . Logistic Regression of Labor Migration, Hsin-Hsing: 1965 and 1979 Male and Female B Exp(B) Sig. 1965 Gender (male) -1 .48 .23 .00 Age -.07 .93 .00 Constant l .24 3.46 .09 -2 Log likelihood 247.28 N= 286 1979 Gender (male) -. 14 .87 .60 Age -.06 .94 .00 Constant 1.26 3.54 .00 -2 Log likelihood 321.04 N= 291 232 Figure 6.1A Estimated Probabilities of Migration Hsin-Hsing Villagers Age 15-64, 1965 0.9 I- - k 0.8 1. 0.7 f ‘7 ’ e... l ‘ \OK“ 1 l “as 0.6 1 ‘x‘ 1 l “.“\‘\ . 0.5 - _ “e- l 1 - - \ 'u 0.4 ‘ 7 R “.‘ j ‘ ‘ “Q. i ~ . _ 03 ‘l . ‘ ‘ “‘ 1 3 ‘ 7 " ., I“\o..‘> l 0.2 ' ‘ ‘ - - 7‘ ‘4‘ l 7 ' ‘ - _ 7.72% l 0.1 ““““““““ N ‘ .1 o 7i? —‘_—""“— —‘—— T__"——___ w—" "— — T.— T — if ‘ - - :l 15 2o 25 30 35 4o 45 50 55 60 Age in 1965 - +—« Male — - - -- Female Figure 6.18 Estimated Probabilities of Migration Hsin-Hsing Villagers Age 15-64, 1979 097—— ,___,#__ M ~———§A 0.8 l lt‘twct . . . -~.\“ 0 7 l h‘ ‘t‘ .4. - :fi‘w... o 6 l ' " :‘fo‘ - “-. 1 7 ‘ 15“..- 0.5 r ‘ - :1. ‘ ‘ (“ma 0 4 ‘ ‘ "1‘94. '0 . : ‘ 1 ‘ ‘h‘t ‘ 0.3 ' ‘ ‘rt‘t‘e e l ‘ ' “tax 0 2 ‘l . ‘ ‘mttw 0.1 1 15 2o 25 30 35 4o 45 50 55 60 Age in 1979 ___.. , Male ————— Female 233 6.4.1.1 Migration among Women To test research hypothesis five (H5), gender is controlled for in the logistic regression model. The results shown in Table 6.3 reveal that research hypothesis five (H5) is supported by both the 1965 and the 1979 data. In other words, the information in Table 6.3 demonstrates that among female villagers, the negative association between the likelihood of being a migrant and age is significant. The 1965 data show that a one-year increase in age resulted in a seven-percent-decrease in the log odds of being a migrant. A significantly negative association between age and being a migrant is also found in the 1979 data. A one-year increase in age resulted in a six-percent-decrease in the log odds of being a migrant. The results of the logistic regression for female migration suggest that women’s ages were negatively related to the likelihood of their being in the urban labor force. In both time periods, among female villagers of working age, an increase in age decreased the likelihood of residing in urban cities. The individual perspective of migration theory suggests that the negative association between migration and age among women reflected the fact that, in the Taiwanese economic environment of the mid-19603 and the late 19703, job opportunities in cities attracted more younger than older women. However, this explanation does not indicate how migration is related to family power dynamics. In the following sections, I would like to move beyond the limits of the individual perspective on migration by discussing how different employment statuses among female migrants, and between migrant and non-migrant women in the labor force, suggest the association between family power dynamics and migration. 234 _.‘ The fact that younger women in the labor force were more likely to reside in urban cities than their older counterparts was shown in Chapter IV. There we saw that, first, younger women accounted for a larger proportion of women in the urban labor force than did older women in both 1965 and 1979. Second, we saw that the mean age of women in the urban labor force was lower than that of women in the rural labor force during this same time period (see Table 4.9). Further, in the mid-19603, unmarried women accounted for a large proportion (68.8% or 11 out 16) of female villagers of working age residing in urban cities. I begin to explain these differences by discussing why daughters’ position in the family made them more likely to be migrants than sons. Then, I explore why married women were unlikely to migrate in search of paid work. Showing the different positions younger and older women held in the power hierarchy of the family helps illuminate why younger women were more likely to work for wages in cities than were older women. The fact that female migrants were single (see Table 6.1) reflects the fact that young women, rather than young men, were deployed to cities as a means to maintain or improve their family economies. Compared to their female counterparts, single men in cities accounted for 57.8 percent of male villagers in the urban labor force in 1965. This proportion was lower than that accounted for by single women in the urban labor force (i.e., 68.8%) in 1965. I argue that the high proportion of young female villagers in the urban labor force, compared to their male counterparts, reflected the implementation of household heads’ power. Traditionally, patrilineal kinship determined young women’s 235 position in the family power hierarchy (R. S. Gallin, 1985; Gates, 1987; Greenhalgh, 1985, Wold, 1972), land they were treated as temporary family members. In the mid- 19603, withdrawing daughters from school and deploying them into the wage labor market as early as possible were strategies used to improve and maintain the family economy. Married women’s position within the family power hierarchy and the agriculture-based economic structure were barriers to their participation in the urban wage labor force. When opportunities for waged jobs in the rural labor market were insufficient, family members in the labor force had to seek job opportunities outside of the rural area. Because men were socialized to earn money to support and/or maintain their families, they were likely to participate in the waged labor market. Married women were expected to stay behind to assume the domestic responsibilities of their mothers-in-law, to care for the land in the absence of their husbands, and to service the desires of their mothers-in- law. They, therefore, were unlikely to migrate to cities. The combination of the likelihood of the migration of single women and the unlikelihood of married women engaging in migration explains the different moving patterns of younger and older women in the mid- ] 9603. In short, gender in combination with age defined a woman’s position in the power hierarchy of the family. Young women migrated to earn incomes that sustained the family. Older women stayed at home to serve their mothers-in-law and sustain a component of the family’s economic base--the land. 236 In the late 19703, deploying daughters into the urban wage labor market was still a practice used by Hsin-Hsing parents to improve or maintain their family economies. Although, the development of rural industry in the Hsin-Hsing area absorbed many young women who might otherwise have been sent to cities in search of employment, the number of unmarried female villagers residing outside of the village increased from 18 in 1965 to 44 in 1979 (see Table 6.1). Fifteen single women participated in the wage labor market in urban cities. These 15 single women accounted for 36.6 percent (15 out of 41) of the women who were living in cities and of working age, a percentage lower than that (68.8% or 11 out of 16) in the mid-19603 (see Table 6.1). Nevertheless, the decreased proportion of single women in the urban wage labor force did not represent a cessation of Hsin-Hsing parents’ practice of sending daughters to seek waged opportunities in urban cities. In the late 19703, rural Taiwanese parents still sent their unmarried daughters to seek waged employment opportunities in urban cities. Among the 15 single women in the urban wage labor market, 11 (73.3% or 11 out of 15) were either living alone or residing with their married siblings in urban cities. More married women resided in urban cities with their husbands in the late 19703 than in the mid-19603. In terms of absolute numbers, they increased from five in 1965 to 26 in 1979. These 26 married women residing in urban cities accounted for 63.4 percent (26 out of 41) of migrant women of working age (see Table 6.1). The increase in the number of married female migrants probably reflected the maturation of migrant families. Within the village, the proportion of women of working ages accounted for by married women 237 increased from 60.1 percent (89 out of 146) in 1965 to 72.5 percent (79 out of 109) in 1979 (see Table 6.1). This might have reflected the fact that rural industrialization made it possible for married women in the village to combine productive and reproduction work. In sum, both logistic regression analysis and descriptive statistics reveal that younger women were more likely to be members of the urban labor. force than older women in both the mid-19603 and the late 19703. This reflects their positions in the authoritarian family hierarchy. In the absence of husbands and the presence of domineering mothers- in-law, married women had to stay in the village to take care of the land and to assume older women’s work. In the late 19703, rural industrialization created paid off-farm jobs for married women, although it did not obviate the need for them to retain their obligations of take care of the domestic unit. In the absence of a long-term value for the security of the family and in the presence of sons with potential continuing value, daughters had to migrate to cities in search of work. The practice of sending young single women to the urban cities to seek employment opportunities to maintain and/or improve the family economy indicates the traditional implementation of household head’s power over their daughters. 6.4.1.2 Employment Status among Married Female Migrants While the analysis in the first stage demonstrates the association between age and the likelihood of migration among women of working age, it does not necessarily demonstrate the reasons for migration among women. In the second stage of the analysis, I, therefore, focus on the employment status of migrant women. 238 ‘1‘..- When married women resided in urban cities, what did they do? Did they participate in the urban wage labor market? The association between age and migration among women and how family power dynamics influence the decision to migrate were shown in the previous section. Family power dynamics, however, not only influenced women’s migration, but also their employment status in urban cities. To address the question posed at the beginning of this paragraph, then, I discuss how family power dynamics influence the employment status of female villagers residing in urban cities. Participation in the urban wage labor market may reflect family power dynamics, which determine who is employed for income and who is not. As Shihadeh (1991:432) argues, “[t]he most powerful determinant of employment returns among wives was not their economic and demographic background characteristics but whether or not they played a subsidiary role in the family migration.” Following traditional gender norms, married women were responsible for domestic work and the care of other family members (R. S. Gallin, 1995). The responsibility for child-care is the primary barrier preventing married women from participating in labor markets for income (Friedl, 1967; Presser & Baldwin, 1980; Salaff, 1981; Saraceno, 1984). Indeed, Craig (1981) and O’Connor (1988) argue that getting their children out of the house is crucial for young mothers who wish to participate in wage labor markets. Therefore, reducing their child-care responsibilities is expected to increase the possibility of married women participating in wage labor markets. As Connelly (1992) argues, availability of no-cost child care leads to a relief 239 from child-care responsibilities, and increases married women’s participation in the wage labor force. There were thus two factors that influenced the likelihood that Hsin-Hsing women would or would not participate in wage labor markets: (1) living with young children, especially preschoolers, and (2) living with older female members, especially mothers-in-law. Children who needed care generally were those who were too young to participate in the formal educational system. Day-care options were under-developed in Taiwan in the 19603 and 19703. Children younger than six years old, the age for entering elementary school in Taiwan, thus became the major barrier to married women’s participation in wage labor markets. A negative relationship consequently existed between the likelihood of a women with preschoolers participating in wage labor markets. Those who lived with preschoolers were less likely to work for pay than those who did not. Following the same logic, unmarried women were more likely to join wage labor markets than were married women with preschoolers (Chattopadhyay 1998). 6.4.1.3 Employment Status of Migrant and Non-Migrant Women In 1965, among migrants from Hsin-Hsing village, there were 31 migrants living in 11 conjugal units in cities (see Table 6.4). Compared with Hsin-Hsing villagers residing in cities in the late 19703, in 1965 the proportion of villagers migrating with other members in the same conjugal units was relatively low, only 37.8 percent (31 out of 82). Among these 11 conjugal units outside of the village in 1965, seven conjugal units were constituted by unmarried siblings. Only four conjugal units in cities were made up of married women residing with their husbands. Due to the early migrants’ eventual 240 settlement in urban cities, migrant conjugal units increased. In the late 19703, there were 114 Hsin-Hsing villagers clustered in 27 migrant conjugal units in urban cities. The growth/maturation of those early migrant families led to a large number of married men residing in urban cities with their wives and children in the late 19703. Table 6.4 Hsin-Hsing’s Migrants in Conjugal Units and Migrant CorLuggl Units, 1965 and 1979 - 1965 1979 Total number of migrants 82 153 Migrants in conjugal units 31 1 l4 Migrant conjugal units 11 27 The increase of migrant conjugal units led to the increase in the proportion of children and married women residing outside of Hsin-Hsing Village. This increase in the number of married women raised the proportion of female Hsin-Hsing villagers in the urban labor force. This increase, however, did not boost the proportion of women participating in the urban wage labor force. The fact is that married women with preschoolers tended either to work at home in cities, taking care of these young children, or to hold jobs (such as work in a family business or industrial outwork) that provided them with the flexibility to take care of young children and to work at the same time. Although taking care of children is the major barrier preventing married women fiom participating in wage labor markets, if this difficulty can be overcome, joining the wage labor market theoretically becomes possible. Therefore, the question becomes: Who can 241 take care of preschool age children other than their mothers? The answer is that older family members, usually patriline grandmothers, assume this responsibility. A mother-in- law also can share a woman’s responsibility for care of the home. Therefore, living with an older female family member could be positively associated with the probability of a woman’s participating in the wage labor market. But, were rural married women more likely to participate in the wage labor market than their urban counterparts? Was employment status related to residential arrangements? Table 6.5 demonstrates that among women living with preschoolers in 1979, those who resided in cities were less likely to participate in the labor market for income than were those who resided in Hsin-Hsing Village. The difference, however, is slight. Among married female villagers residing in cities, who had preschooler(s), only 29.4 percent (5 out of 17) were employed and earning income. At the same time, only 32.0 percent (8 out of 25) of women in the rural labor force, who were mothers of preschooler(s), participated in the wage labor market. The very slight difference between women with preschoolers in cities and in the village suggests that taking care of preschoolers did not cause the different likelihood of participating in the labor market for income between married women residing in cities and in Hsin-Hsing. In other words, living with preschoolers had a similar influence on participation in the wage labor market for women in cities and those in the village. 242 Table 6.5 Women Classified by Occupation, Migration Status, and Presence or Absence of Preschool Children, Hsin-HsingVillage, 1979 Non-migrant Mi ant no pre- with pre- no pre- with pre- Occupation schoolers schoolers schoolers schoolers Single Women Housekeeper/housewife 2 -- -- 1 Family worker 4 -- -- -- Self-employed farmer -- -- -- -- Off-farm worker 15 -- 13 -- Self-employed (business owner) -- -- 1 -- N= 21 -- 14 1 1“" Married Women Housekeeper/housewife 28 9 3 9 Family worker l l 2 3 Self-employed farmer l8 7 -- -- Off-farm worker 13 7 4 4~ Self-employed (business owner) 3 1 -- 1 N= 63 25 9 17 Although living with preschoolers was not a necessary condition to explain why married women in the urban labor force did not work for income, mothers residing in the village with preschoolers who worked for pay might have received support from other female family members. In the late 19703, there were 43 married women (25 in Hsin-Hsing and 18 living outside of the village) in the labor force who resided with their preschool child(ren) (see Table 6.6). Among this group of women, eight rural and five urban women participated in the labor market for income (see Table 6.6). They accounted for 32.0 percent (8 out of 25 rural women) and 27.8 percent (5 out of 18 urban women) of those living with preschoolers. The difference is slight. However, 55.6 percent (10 out of 18) of urban female villagers who had child(ren) of preschool age were housekeepers. Among rural female villagers, the proportion of women working as housekeepers was only 36.0 percent (9 out of 25), a much lower percentage than that of urban women. 243 Table 6.6 Occupations of Married Women Living with Preschoolers Classified by Migration Status, and Presence and Absence of Mother-in-law, Hsin-Hsing: 1979 Non-mi ant married women Migrant married women Living Living Living Living with without with without mother-in- mother-in- mother-in- mother-in- Occupation law law Subtotal law law Subtotal Housekeeper 5 4 9 0 10 10 Family worker 0 l 1 0 3 3 Self-employed farmer 3 4 7 -- -- -- Off-farm worker 6 l 7 0 4 4 Self-employed 1 O 1 0 l 1 Total 15 10 25 18 18 For women with child(ren) of preschool age, the probability of being an income-earner was determined by the accessibility of social support. Could older female family members have been the primary source of this support for women in the paid labor force. 4‘) Did living with such relatives increase the likelihood of participating in the labor market for income? In the late 19703, non-mi grant women were more likely to reside with older female family members than were migrant women. In Hsin-Hsing, 15 women with preschool age child(ren) lived with their mothers-in-law, while no urban female villagers lived with mothers-in-law (see Table 6.6). Six rural women who lived with mothers-in- law were off-farm workers. They accounted for 85.7 percent (6 out of 7) of rural women who were working off-farm and living with preschool child(ren). The fact that these daughters-in-law were able to work outside of their houses implied that their mothers-in- law assumed responsibility for domestic work and took care of grandchildren as well, thereby releasing the younger women for paid employment. This datum suggests that 244 with rural industrialization, the traditional power dynamics between mothers- and daughters-in-law changed, and mothers-in-law gave up their traditional prerogative of a life of leisure in old age (R. S. Gallin, 1986). Married women in the village, however, did not always benefit from living with older female family members. The presence of a mother-in-law did not necessarily release a daughter-in-law from domestic work. Traditionally, when daughters-in-law married into families, they assumed their mothers’-in-law responsibilities including their work in agricultural production. Following this logic, living with a mother-in-law did not necessarily increase the probability that a married woman would participate in the labor market to earn income. Table 6.7 Occupational Matrix of Mothers-in-law and Daughters-in-law, Hsin-Hsing: 1979 Daughters-in-law Self- M th . 1 House- Off-farm employed Self- 0 ers-rn— aw keeper worker farmer employed Total Retired 3 2 4 O 9 Housekeeper 3 5 4 l l 3 Self-employed farmer 2 3 0 1 6 Total 8 10 8 2 28 Note: Columns are the occupations of daughters-in-law, and rows are the occupations of mothers-in-law. “3” in the first cell of the first column indicates that there were three daughters-in-law working as housekeepers, while their mothers-in-law were retired. Table 6.7 demonstrates an occupational matrix for 28 daughters-in-law residing with their mothers-in-law in Hsin-Hsing Village. On the one hand, the data in this table show that 245 it when mothers-in-law shared or assumed responsibility for taking care of the house, daughters-in-law were employed for monetary income. As R. S. Gallin (1984:391) writes, “the existence of a supportive family structure in which mothers-in-law took over some of the younger women’s tasks had a direct impact on women’s access to wage work.” For example, among 13 daughters-in-law living with mothers—in—law [who took care of the home, five (38.5%) worked for wages, while one (7.7%) was self-employed. On the other hand, the data in Table 6.7 demonstrate that not all married women benefited from living with their mothers-in-law. Nine daughters-in-law living with their mothers-in—law assumed primary responsibility for the work previously done by older women. Three (33.3%) of them were housekeepers and four (44.4%) worked on family land, while only two (22.2%) worked off-farm for income. This analysis suggests that married women within the Taiwanese family continued to remain subordinate to their mothers-in-law. Their employment status not only depended on their position in the family power hierarchy, but also influenced their position in this hierarchy. The division of labor between mothers-in-law and daughters-in law, on the one hand, influenced the probabilities of married women being waged laborers. On the other hand, it reflected the power dynamics between mothers-in-law and daughters-in-law. As R. S. Gallin (1986:42) argues, [w]hen villagers were tied to the land, children were dependent on farms controlled by parents for a livelihood, and the old held sway over the young. Under those conditions, daughters-in-law had few resources to serve as a base from which to defy the authority of their mothers-in-law. 246 it Nevertheless, change in the village’s economic structure had the potential to affect the balance of power between mothers-in-law and daughters-in-law. When most villagers depended on off-farm employment to sustain the family economy, some mothers-in-law gradually lost their authority. As R. S. Gallin (1986:43) argues, “[w]ith rural industrialization, most income was derived from off-farm employment, parents were dependent on children in whom the major income power rested, and daughters-in-law had achieved a new bargaining position with which to resist the traditional authority of their mothers-in- law.” Some married women residing in the village received support from their mothers-in-law, and they thus participated in the rural wage labor market. Other rural married women, in contrast, continued to assume the responsibilities of their mothers-in-law, releasing the older women from the drudgery of reproduction and production. This difference was a product of a family’s position in the class structure. Although this research could not measure class, R. S. Gallin (1994) argues that family power dynamics were affected by class. As she points out, poor older women in the village had no authority to command the labor of their daughters-in-law, and they had to work to secure their own future. Some financially secure older women, however, were able to maintain their life in old age and direct their daughters-in-law because their husbands owned productive property upon which the younger generation depended. In both instances, the division of labor between women of different generations reflected family power dynamics between mothers-in-law and daughters-in-law. But class intervened to influence the direction of the balance of power. 247 In sum, the association between age and working for pay among migrant women was negative. This negative relationship might have reflected the fact that urban employers were more likely to hire younger than older women. However, the data collected in the mid-19603 and the late 19703 show that while some migrant women were employed for income, others, especially married women with preschoolers, stayed at home as housekeepers. Unmarried and younger women in the urban labor force were more likely if- to work for income than their older and married counterparts. Because of the patrilineal -r-.-—'r' kinship system, unmarried daughters were seen as temporary family members. Sending them to wage labor markets as early as possible made them start repaying their debt to their parents sooner rather than later. Nevertheless, in some cases these working daughters were able to change their position in their family power hierarchy by contributing to the family economy (R. S. Gallin, 2001 ). Some rural women were able to work for pay while others were not, and this difference in occupational status was implicated in their mothers-in-law ability to control them and their labor. Urban women, in contrast, were unlikely to work for pay, and none lived with mothers-in-law. While the absence of older women may not have been a determining factor in urban daughters-in-law’s occupational status, it does mean that mothers-in-law lost control over the younger women. Nevertheless, while older women’s authority over » them was mitigated, these women did not change their position in the traditional male- female hierarchy. They continued to be subordinate to their husbands. Following traditional gender norms, and in the absence of a mother-in-law to help in domestic work, 248 married women in urban cities were unlikely to work outside of their houses.” Without accessibility to an income, married women in urban cities had no base “from which to change absolute financial dependence on their husbands -- a dimension of the conjugal relationship they judged oppressive” (R. S. Gallin, 1995:129). 6.4.2 Age Differentials between Male and Female Villagers in the Urban Wage Labor Market To demonstrate the association of migration with family power dynamics, I examine the relationship between age and gender among villagers participating in the urban wage labor market. 1 hypothesize that among villagers participating in urban wage labor markets, men are older than females. Analysis of Variance (ANOVA) is applied to examine the difference in age between men and women who were participating in urban labor markets for pay. The results of the ANOVA shown in Table 6.8 reveal that research hypothesis six (H6) is not supported by the 1965 data. It is, however, supported by the 1979 data. In the mid- 19603, the age differential between men and women in the urban wage labor market was not statistically significant, while the age difference was significant in the late 19703. In the mid-19603, there were 60 Hsin-Hsing villagers (including 13 women and 47 men) working for pay in cities. In the late 19703, 69 villagers (including 22 women and 47 men) participated in the urban wage labor force. The mean age of men participating in the _ urban labor force for wages was 26.1 while that of their female counterparts was 22.2 in 75 Working outside of the home in urban cities was possible and necessary for married women when their continue to the next page... 249 1965. Despite the fact that men had a higher mean age than women, the difference was not statistically significant. In 1979, the mean age of men working in the urban labor market for pay increased to 31.8. At the same time, the. mean age of women working for income in urban cities increased to 24.6 from,22.2 in 1965. The difference in age between men and women participating in the urban labor force for income became statistically significant at (1:001. Table 6.8 Analysis of Variance for the Ages of Villagers in the Urban Wage Labor Market b Gender, Hsin-Hsing: 1965 and 1979 Sum of Mean Squares DF Square F 1965 Between Groups 154.6 1 154.6 1.37 Within Groups 6547.5 58 112.9 Total 6702.2 59 1979 Between Groups 781.3 1 781.3 13.82 Within Groups 3787.2 67 56.5 Total 4569.2 68 m significant at level of .01 ' significant at level of .05 ' sigrlificant at level of .10 While the individual perspective of migration theory suggests that participation in the wage labor market is determined by personal characteristics, it is possible that the likelihood of j oining the labor market for pay can be a product of other factors such as family and economic structures. In Chapters IV and V, I discussed how family structure and local and national economic development influenced rural villagers’ migration. husbands did not earn enough to support their families. 250 Although migration might have been related to personal characteristics, moving to seek work also could have reflected power dynamics within a family. Although some researchers (e.g., Browning, 1969; Chang, 1979; Chiang, 1978; Ladinsky, 1967; Li, 1974; Liao, 1977; Liu, 1993; Long, 1973 and 1992; Speare, 1974; Tsai, 1978; Yin, 1978; Zachariah, 1966) suggest that personal characteristics determine the likelihood of migration, the fact is that migration is determined by the family power hierarchy, which is based on personal characteristics such as age and gender. As shown previously, female villagers accounted for 25.8 percent (16 out of 62) of Hsin- Hsing villagers in the urban labor force in 1965. Their proportion increased to 47.7 percent (41 out of 87) in 1979. The increase in the number and the proportion of migrants, however, did not lead to an increased proportion of women working for pay in the urban labor force. In Taiwanese society, men traditionally were expected to be more outwardly oriented than were women. According to traditional gender norms, “[m]en dominated the public domain, working outside the home ...[, while] women presided over the domestic sphere, managing the household, [and] servicing its members ...” (R. S. Gallin 1984:386). Accordingly, differences in social expectations for married men and women result in different probabilities for occupational attainment (O’Connor 1988; Quinn 1977). The age differentials between male and female villagers in the urban wage labor market can be discussed not only by examining the different positions of and social expectations for married men and married women, but also by interrogating those of and for unmarried 251 men and unmarried women. On the one hand, young unmarried women do not have the same home and child-care demands as married women. They thus are more likely to pursue non-traditional work options and to optimize income-producing activities than are married women (Chattopadhyay 1998; Treiman and Terrell 1975). On the other hand, Taiwanese parents preferred to deploy unmarried daughters rather than unmarried sons to urban wage labor markets to improve and/or maintain their family economies. Sending daughters to the urban wage labor market as early as possible means parents had to withdraw them from school early. Women’s unequal opportunity in comparison to men to gain education explains why men in the wage labor market were older than their female counterparts. The unbalanced positions of men and women in the family power hierarchy was reflected in their unequal educational opportunities. In Chapter IV, I showed that, in general, men had more education than women had in Hsin-Hsing Village. Table 6.9 shows that male adolescents were more likely to stay in school for education than their female counterparts did in the mid-19603 and the late 19703. In other words, while young single men were pursuing higher education, young single women were likely to participate in the wage labor market. Table 6.9 Students and Non-students by Gender, Hsin-Hsing: 1965 and 1979 1965 1979 Male Female Total Male Female Total Non-student 24 44 68 3 7 10 Student 17 ll 28 19 12 31 Total 41 55 96 22 19 41 252 In Table 6.9, the analysis of the 1965 data includes 96 male and female villagers ages 12- 18. The analysis based on the 1979 data includes 41 villagers ages 15-18. The data in Table 6.9 indicate that among villagers ages 12-18 in 1965 and ages 15-18 in 1979, female villagers were less likely than their male counterparts to stay in school. In 1965, while 20.0 percent (11 out of 55) of female adolescents stayed in school, 80.0 percent (44 out of 55) of single women ages 12-18 participated in the wage labor force. By contrast, in this same year 41.5 percent (17 out of 41) of young men stayed in school, while only 58.5 percent (24 out of 41) participated in the wage labor force. These statistics show that in the mid-19603, male adolescents were more likely than female adolescents to stay in school for education. A similar pattern is found in the 1979 data. While 86.4 percent (12 out of 19) of male adolescents stayed in school, only 63.2 percent (31 out of 41) of their female counterparts continued their studies. The statistics in Table 6.9 reflect the unequal opportunities young female and male villagers had to attend school. They, further, reflect the different expectations parents held for sons and daughters. Sending sons to school is a long-term investment by parents for the elders’ life in old age. Sending daughters to cities to seek employment opportunities for income as early as possible is a short-term family strategy adopted by parents to improve and/or maintain the family economy. In sum, this section shows that in the late-19703, among villagers who resided in urban cities and worked for income, men were older than women. This significant finding 253 reflects several social phenomena in the village, which led to the greater likelihood that young unmarried women and married men would participate in the labor market for pay than were unmanied men and married women. First, fathers dominated their daughters and held sway over their life chances. Young single women were deployed to seek work as a means to achieve the maintenance and/or improvement of the family economy. Their brothers, in contrast, stayed in school for education. Second, within the conjugal units in urban cities, married women with no access to an income continued to be subordinate to their husbands. As R. S. Gallin (1995:129) has argued, “[w]omen’s control of their earnings has the potential to erode traditional ideology and its norms of behavior and to provide women with the resources necessary to create an autonomous space.” The different employment status among women in the urban labor force, in general, was affected by the intersection of their marital status and age. Married women usually were older than their single counterparts. Married women, especially those who resided with preschoolers in cities, usually were unable to pursue paid employment. These arguments provide a strong foundation to understand: (1) why younger female villagers in Hsin- Hsing were more likely to participate in the urban paid labor force than were older women in the mid-19603 and the late 19703, (2) why younger female migrants were more likely than older female migrants to work for income, and (3) why men who participated in the urban labor force for income were older than their female counterparts. Among those living in the village, some married women assumed the responsibility of their mothers-in-law, including farming family land and taking care of the home. Other 254 married women, however, participated in the rural wage labor market, while their mothers-in-law cared for their children and shared responsibility for the home and farm. Nevertheless, men and women are unequal in rural Taiwanese society. According to tradition, married men work to bring income into their families while married women assume responsibility for domestic work (R. S. Gallin 1985). This leads to an imbalance in the economic power of married men and women, directly determining their positions in the family power hierarchy. Unmarried women were traditionally treated as temporary family members. Once they were married, they left their natal families to live with the families of their husbands. In this sense, they became members of their husbands’ families which controlled their labor power. Following this logic, sending girls to school was not the ideal investment for parents who wished to ensure sustenance in their old age. Instead, parents preferred to send boys, who were theoretically responsible for their care as they aged, to school. Thus, male adolescent villagers were more likely to stay in school than were their female counterparts, and unmarried daughters were more likely to join the wage labor market than were their single brothers. Additionally, this explanation provides a base to understand the unequal opportunities for education between male and female adolescents in the village. 6.5 Summary The primary intention of this chapter was to demonstrate how family power dynamics are related to migration. The inconsistent conclusions of past research derived from the individual perspective of migration theory suggest that while it is correct that the 255 “ml migration decision is an outcome of a rational evaluation of the “expected returns” of movement, placing the emphasis on an individual’s “cost-benefit” analysis is insufficient to fully explain the migration process. This chapter moved beyond the traditional “cost- benefit” argument and considered the role of power dynamics in the migration process. Two research hypotheses were examined to identify how family power dynamics, which are shaped by age and gender, influence migration and employment status at destination. Among women in the labor force, younger females were more likely than their older counterparts to migrate to cities to seek job opportunities. Among migrants participating in the urban wage labor market, men were relatively older than women. Labor migration is affected by the economic structure. While most Hsin-Hsing villagers moved to cities to seek employment opportunities in the mid-19603, job opportunities derived from the development of rural industry kept a great number of them in the rural area in the late 19703. The individual perspective of migration theory suggests that personal characteristics determine decisions about migration. However, as I have argued, migration is also a product of the family power hierarchy. While young male adolescents stayed in school, young female were sent by their parents to work in cities to increase their family’s monetary income and to maintain and/or improve the family economy. When more conjugal units migrated to or developed in cities, there was no increase in the percentage of married women working for income. Those who worked for income in the labor force were primarily married men and single women. Because married people were usually older than the unmarried, among migrants working for income, men were older 256 than women. This, in addition, implied that, in cities, married men’s and married women’s positions were maintained in the traditional power hierarchy. In the absence of a job from which to derive money, married women had no base to challenge “absolute financial dependence on their husbands” (R. S. Gallin, 1995:129). In the mid-19603, when their husbands resided and worked for pay in urban cities, married women stayed in Hsin—Hsing Village to take care of their families and to farm the family land. The relationship between mothers-in-law and daughters-in-law in the village usually remained defined by tradition. In the late 19703, however, when the development of rural industry brought a great number of off-farm employment opportunities to the Hsin-Hsing area, daughters-in-law had more opportunities to join the wage labor force than they previously had. For some, the development of rural industry brought changes in the traditional power dynamics between mothers-in-laws and daughters-in-law. When daughters-in—law worked in factories, older women assumed responsibility for domestic work. Under the changing economic structure, the increased off-farm employment opportunities in Hsin-Hsing area made young couples dependent more on the off-farm wage labor market for income than on farming. On the one hand, young couples were not dependent on land, which was controlled by their parents. On the other hand, daughters-in-law had more “resources to serve as a base from which to defy the authority oftheir mothers-in-law” (R. s. Gallin, 1986: 42). In sum, this chapter demonstrates how family power dynamics were related to migration among female and male villagers and to their employment status. Age and gender define 257 interpersonal relationships between family members, and they determine the position of each family member within the power hierarchy of the family. The migration status and employment status of a person, indeed, reflect, his/her position within the authoritarian family hierarchy. 258 CHAPTER VII CONCLUSIONS AND DISCUSSION Voluntary migration usually is considered an economic activity. The most common conclusion of migration studies is that people migrate primarily for economic reasons, especially in the Third World (Pamwell, 1993). Researchers have argued that migration responds to spatial inequalities in expected earnings (Guest, 1989; Harris and Todaro, 1970; Sjaastad, 1962; Todaro, 1969, 1976, 1980; Wood, 1981). The gap between rural and urban wages leads to migration from rural areas with high wage rates in the urban modern sector inevitably leading to [high expected income returns from rural emigration (Harris and Todaro, 1970). Others argue, however, that the rural-urban wage differential is institutionalized or politically determined, ratherthan market determined (Harris and Todaro, 1970; Montgomery, 1981). Montgomery (1981), for example, emphasized the rural-push side and insisted that, in certain areas, agricultural markets were highly distorted by government policies that made rural incomes artificially low, thereby stimulating rural-urban migration. While structuralist views such as these consider migration to originate in institutional change that affects the relations of production in the sending and receiving sectors, the individual perspective suggests that migration is a human behavior in response to the spatial inequality in expected income and occupational opportunities between rural and urban areas. Within this individual view, the family perspective suggests that migration is a collective behavior involving discussion by family members as a group. Migration is 259 one of a series of household/family strategies by which family members actively strive to achieve a balance between the domestic unit’s consumption necessities,ethe labor power at its disposal, and the alternatives for generating monetary and non-monetary income (Boyd, 1989; Grigg, 1980; Guest, 1989; Wood, 1981). For instance, under conditions of structural change, an imbalance between these two key components -- family labor input and family consumption -- is likely to occur and a domestic unit will have to seek an alternative sustenance strategy to achieve anew balance. The strategies for achieving such a new balance include seeking occupational opportunities in the local area as well as in places away from home. The family perspective on migration provides a theoretical framework to explain migration as one strategy to maintain and/or improve a domestic unit’s economy. The major limitation of the family perspective of migration theory is that it assumes migration decisions are made collectively. This perspective does not take into account the power hierarchy within which decision making occurs in the family. For Taiwanese families, the authority for decision making is traditionally held by one or a few family members. To expand the knowledge base of migration theory, therefore, this dissertation focused not only on the examination of how factors at the family level influenced migration in Hsin-Hsing, Taiwan, in 1965 and 1979, but also on the implication of family power dynamics on migration. Additionally, this research explored how changes in the economic structure of the village affected the influence of selective factors on migration at the family level. In other words, this dissertation discussed the associations between 260 family migration and its factors under the different economic structures of the Taiwanese rural area in the mid-19603 and the late 19703. In the following sections, I summarize the research and discuss its significance and limitations. I conclude with some thoughts about the political economy of Taiwan at the end of millennium -- the context within which migration will or will not occur. It is diffiCult to conclude with substantive recommendations because the dissertation deals with times long since gone. The final section, however, will demonstrate how the changing political economy in Taiwan might influence the migration of rural population. 7.1 Summary of Dissertation 7.1.1 Research Intentions This research adopted three major migration perspectives: family, structural, and individual. The first intention of this research was to examine internal migration in. Taiwan by adopting an integrated research fiamework, which laid out the relationship between factors that might spur movement and the likelihood of a family adopting l migration. Specifically, this research examined how family type, landholdings available for family working members and for the members ' consumption needs, and family members ' participation in local wage labor markets influence migration. Second, this research, using this integrated approach, examined how change in economic structure influenced a rural family’s implementation of migration. In other words, this dissertation adopted an integrated framework at the family level to examine how migration serves as a family sustenance/mobility/survival strategy to cope with structural constraints. 261 Nevertheless, such strategies are not adopted within a vacuum. Thus this research incorporated the notion of power and explored how a family power hierarchy intervened in the decision-making process influencing who did or did not migrate. In sum, this research was organized to answer the following four questiOns. 1. How does the amount of cultivated land available relate to the decision of migration? 2. How does the development of rural industry influence the migration of the rural population? 3. How does family type influence the migration decision of family members? 4. How do family power dynamics relate to migration decisions and processes? 7.1.2 Research Findings In Chapter V, I showed that the associations between labor migration and the factors affecting it changed overtime according to Hsin-Hsing’s economic structure. In the 19603, agriculture was the primary means of production in the Hsin-Hsing area. The amount of land available to input labor was responsible for the villagers’ local labor market participation. By the end of the 19703, a rural industrial zone established near the village as well as local industrial development influenced the work patterns of Hsin- Hsing villagers. More non-agricultural job opportunities became available to the villagers and, in 1979, villagers participating in local non-farm wage labor markets accounted for a great proportion of the labor force working locally. More villagers worked locally for monetary income in 1979 than in 1965. 262 The land consolidation program of the late 19603 and development of new agricultural technologies increased the likelihood that villagers were both self-employed farmers and paid off-farm workers by the late 19703. In the 19603, farming was labor-intensive, and there were few off-farm employment occupations available in the Hsin-Hsing area via which to diversify family income sources or to supplement the family economy. Villagers had to go out of the local area to seek paid jobs. When more paid off-farm jobs became available in the late 19703, villagers were able to diversify family income sources and increase the unit’s income by taking waged jobs locally. At the same time, they could work on their family land because innovations in agricultural production reduced the need for intensive human labor. While descriptive statistics revealed the associations between migration and economic structures, advanced statistical analyses were applied to examine the relationships between migration and family factors. Family’s local labor-force participation rate was more influential in labor migration than other factors. The results of multiple regression analysis showed that when other factors were controlled for, a family’s local labor-force participation rate had the most influence on labor migration in 1965. It was also very influential in 1979.76 263 Although a bivariate analysis for the 1965 data showed that family type and the number of family members in the labor force residing in urban cities were not significantly associated with each other, a multiple regression analysis suggested that family type had significant associations with a family’s labor migration when other factors were controlled for. The results demonstrated that simple type of families had fewer family members who were labor migrants than families of the complex type, controlling for other independent factors. In addition, the influence of family type on the number of family members in the labor force residing outside of the village increased over time. The analyses shown in Chapter V basically supported the theoretical framework, with the exception of the relationship between family land accessibility and labor migration. The weak and/or insignificant associations between family’s land accessibility and the number of family members in the labor force residing outside of the village, however, do not necessarily imply that a family’s access to land had no influence onlabor migration. Rather, the insignificant associations show only that differences in family land accessibility could not explain why Hsin-Hsing families had different numbers of labor migrants. Nevertheless, the occurrence of labor migration in the village was related to “inadequate family land accessibility.” Families with less land available might not necessarily have had more members in the urban labor force than those with more land available. But because land in the village was ubiquitously inadequate, all families in the 76 As mentioned earlier in this chapter and some chapters in this dissertation, migration is a complex process. Advanced statistical techniques are needs in examining the statistical relationships between migration and other influential factors. 264 community included migrants who sought paid in cities. The movement to cities of family members in the labor force was prevalent in 1965, regardless of the quantity of family land available, as a comparison of Hsin-Hsing’s holdings to that of Taiwan as a whole show. Nationally, each family in 1965 had an average of 0.95 chia of cultivated land, while Hsin-Hsing families had only 0.53 chia (see Section 4.4.1 on the agriculture sector in Chapter IV). In addition to the examination of the relationships between migration and factors influencing it at the family level, this research, using data at the individual level, examined power dynamics within Taiwanese families in Hsin-Hsing Village. The statistical analyses showed, first, that among women ages 15-64, age was negatively related to the likelihood of migrating away from Hsin-Hsing Village. Those who were younger were more likely to migrate to urban cities than were those who were older. Second, among women of working age in urban cities, those who worked without pay were older than those who worked for monetary incomes. Third, in the late 19703, the difference in age between men and women working in urban cities for pay was statistically significant. Men were significantly older than their female counterparts. These results reflected several social phenomena in the village. First, fathers usually dominated their daughters, and young single women were deployed to seek paid work as ' a means to improve the family economy. Second, because of their position in the family power hierarchy and the division of labor it dictated, married women were less likely to work for wages than were other family members. In general, upon marriage, they 265 assumed the responsibilities of their mothers-in-law, including farming family land and taking care of the household. Even when they migrated, married women were assumed to hold prime responsibility for the care of preschoolers. Gender inequality was also shown in the imbalanced opportunities young female and male villagers had to attend school. The unequal status between female and male adolescents, however, demonstrated a traditional social norm in Taiwan, which was re- enforced by partilocal rules of residence. As R. S. Gallin (1984:385) explains, traditionally, when a woman married, she left her natal home to live as a member of her husband’s family. . .. [Therefore,] parents considered daughters a liability, household members who drained family resources as children and who withdrew their assets (domestic labor and earning power) when they _ married. Sons, in contrast, contributed steadily to the family’s economic security during its growth and expansion and provided a source of support for their parents in old age. Therefore, “parents strongly preferred male children” (R. S. Gallin, 1984: 385). Having sons stay in school to acquire education was generally considered by Taiwanese parents as an investment to ensure parents’ security in old age. The significant results of the analyses demonstrated these different expectations for sons and daughters. In sum, the findings revealed that migration is a complex process. Internal movement within a society is affected by structural and family factors. But the process of migration 266 also is determined by family power dynamics, which are shaped by personal characteristics such as gender and age. 7.2 Significance of Dissertation The significance of this dissertation lies in its contributions to migration theory. It moved beyond extant theory in two ways. First, this research applied an integrative research framework based on multiple perspectives of migration theory. Second, this research examined power dynamics within families during the process of migration. The theoretical approach of this research is different from other migration studies on Taiwan. Internal migration was not a major issue in Taiwan until the over-urbanization of a few major cities was recognized in the 19703, and research about internal movement here started in tat decade. Past migration research primarily adopted either one of two perspectives: individual or structural. Some research based on the individual perspective utilized individual characteristics to carry out simple linear regression analyses (see Liao, 1977; Speare, 1971). Others studies compared the demographic characteristics of movers according to different migration streams (see Chiang, 1978; Li, 1974; Speare, 1974). Still other research u3ed aggregated data to demonstrate the differences in migrants’ backgrounds, using gender, age, education, and occupation as the major explanatory variables (see Chang, 1979; Liu, 1993; Tsai, 1978; Yin, 1978). In addition, most research used two-way contingency tables to compare economic variables with migrant characteristics. 267 When the economy in Taiwan started booming in the late 19703 and the early 19803, the focus of migration research shifted from the individual perspective to the structural perspective. Migration studies at this time demonstrated that urban centers with more employment opportunities attracted more labor migrants than areas with fewer of j ob opporttmities. In addition, these studies explained why certain areas became new destinations (see Liu and Tsai, 1990; Tsai, 1981; Tsai, 1990). This research, however, over-emphasized the importance of economic factors. While they provided a broad framework for understanding the incidence of migration in relation to the industrialization and the economic development process, these studies ignored the social dimensions that contribute to migration. The first theoretical significance of this research then is that it proposes an integrated framework at (l) the family level to examine how migration serves as a family sustenance/mobility/survival strategy to cope with structural and economic constraints, and (2) the individual level to examine how power dynamics within Taiwanese families differentially affect the movement of members of the domestic unit. The detailed discussions in the dissertation strongly revealed that migration is a complex process. Explanations based solely on a single perspective are too simple and insufficient to reveal how and why migration is used to improve and/or maintain a family’s economy. Influencing factors at one level are inter-locked with factors at other levels. 268 The second theoretical significance of this dissertation is that it focuses on the influence of a social dimension, which contributes to migration. Specifically, this dissertation, as indicated, examined power dynamics within Taiwanese families during the migration process. Past migration research in Taiwan heavily emphasized the causes of migration and how economic structures were related to migration. Family power dynamics were never dealt with within the frameworks adopted to explain migration. Traditionally, the individual perspective of migration theory suggests that personal characteristics determine decisions about migration. As I have argued, migration also should be seen as a product of a family power hierarchy, which shapes the interaction and behavior of family members and thus their movement. In this dissertation, I analyzed family power dynamics in terms of relations (1) between men and women, and (2) among women. The results of the analyses showed how a family power hierarchy influenced men’s and women’s movement and participation in the urban labor force for wages. When more conjugal units migrated to or developed in cities, there was no increase in the percentage of married women working for income. Those who worked for income in the labor force were primarily married men and single women. Because married people were usually older than the unmarried, among migrants working for income, men were older than women. This, in addition, implied that, in cities, married men’s and married women’s positions were maintained in the traditional power hierarchy. In the absence of a job from which to derive money, married women had no base to challenge “absolute financial dependence on their husbands” (R. S. Gallin, 1995:129). 269 In addition, the findings highlighted the different economic roles of young male and female adolescents in and from Hsin-Hsing Village, which were derived from their positions in the family power hierarchy. Female adolescents were more likely to migrate than their male counterparts because young women were sent by their parents to work in cities to increase the domestic unit’s monetary income and to maintain and/or improve its economy. In contrast, male adolescents stayed in school to achieve more education, thereby increasing their earning ability and presumably securing the life of their parents in old age. These findings are theoretically important, because they demonstrate the inter-locked relationships between migration and family power dynamics. First, power within the family determines the likelihood ’of migration of family members. For example, the findings provide explanations for why young women were more likely to migrate to urban cities to seek employment opportunities than their male counterparts, and how the unequal status between male and female adolescents influence their migration and employment status. In addition, the traditional power dynamics within the family may be maintained during the process of migration. For instance, among migrant married couples, power dynamics within their conjugal units were likely to remain the traditional financial dominance-dependence relations, because women were less likely to work outside of their urban residences for monetary income than their husbands. This research also showed that in the mid-19603, when their husbands resided and worked for pay in urban cities, married women stayed in Hsin-Hsing Village to take care 270 r‘i.’ _. of their families and to farm the family land. The relationship between mothers-in—law and daughters-in-law in the village usually remained defined by tradition. In the late 19703, when the development of rural industry brought a great number of off-farm jobs to the Hsin-Hsing area, daughters-in—law had more opportunities to join the wage labor force than they previously had. For some, the development of rural industry brought changes in the traditional power dynamics between mothers-in-laws and daughters-in- law. When daughters-in-law worked in factories, older women assumed responsibility for domestic work. Under the changing economic structure, the increased off-farm employment opportunities in Hsin-Hsing area made young couples dependent more on the off-farm wage labor market for income than on farming. On the one hand, young couples were not dependent on land, which was controlled by their parents. On the other hand, daughters-in-law had more “resources to serve as a base from which to defy the authority of their mothers-in-law” (R. S. Gallin, 1986: 42). These findings, then, illustrate the theoretical relationships among economic development, migration, and family power dynamics. Although this dissertation suggested that the hierarchy of power within a family influences the process of migration, the findings also suggest that family power dynamics can be shaped by local economic development, which directly or indirectly influences the employment status and migration status of villagers. The growth of the local economy increased the employment opportunities available to villagers, thereby lowering the degree of out-migration from the community. The changes in villagers’ employment patterns further influenced the power dynamics within their families. 271 Although this research focused only on power within the family in Taiwan, it has implications for migration theory about other societies-The individual perspective of migration theory considers personal characteristics a form of human capital. People with different characteristics have different human capital. This perspective also assumes that those with certain personal characteristics (e. g., young male, and high educational level) are more likely to migrate than others because they can find employment in urban cities more easily than those who are less well endowed. This dissertation focused on personal characteristics such as gender and age to argue that migration was an outcome of a hierarchy of power within the family. Gender and age define a person’s position within a family power hierarchy. These characteristics also determine the interaction between and among family members. In the process of migration, which includes the decision-making process and the implementation of migration, family members may act together as a single unit to overcome economic and structural constraints. Nevertheless, family migration decisions are not usually made by domestic units. Rather, they are made by certain family members with power for those with less power. Migration is not necessarily based on the personal will of individual family members. The migration status and employment status of family members is the outcome of the dynamics of power, which are shaped by the gender and age of family members. In sum, the arguments about power presented in this dissertation provide an additional perspective to understand the process of migration. Migration may be an economic activity. But it is also a social phenomenon. To understand migration, the 272 social dimensions of the process are also needed. The notion of power within the family is one of those needed social dimensions. 7.3 Limitations of Dissertation Although this dissertation contributes to the theory of migration, it has some limitations, which were caused by (l) ignoring qualitative data, and (2) missing data. Because of these problems with the data -- intentional and unintentional -- three key variables were excluded from the research reported herein. First, the research was primarily based on quantitative data, and it was designed to examine the statistical relationships between migration and factors influencing the process. The data collected by Dr. Bernard Gallin and Dr. Rita S. Gallin include qualitative data as well as quantitative data, but this research was based only on the quantitative data. Ignoring the qualitative data represents a drawback of this research. Statistical analyses based on quantitative data are inadequate to understand the nuances of social facts. To discover the underlying meanings of and patterns in relationships, qualitative analyses are needed. Without utilizing the qualitative data, some discussions in this dissertation had to rely on arguments or examples presented in the published work of Dr. Rita S. Gallin and Dr. Bernard Gallin, which were primarily generated from their qualitative data. Their arguments and examples derived from the qualitative data provide meaningful contexts for this research to discuss and explain the numerical statistics. 273 ‘F5 .— Thus, although this research was designed to examine an integrated migration model, the exclusion of the qualitative data precluded the inclusion of some important influential factors of labor migration such as social networks. Social networks have theoretically and empirically proved important in migration decisions (see Dinerrnan, 1978; MacDonald and MacDonald, 1974; Massey, 1990a, 1990b; Massey et al., 1987; Mines and de Janvry, 1982; Mines, 1984; Mullan, 1989; Taylor 1986, Tilly and Brown, 1967). Interpersonal relationships are embedded in social networks. In Taiwan, social networks influence people’s daily life (B. Gallin, 1974). Social networks also include people who are not kin-related, such as schoolmates and sworn brothers. Through social networks that link migrants and non-migrants, information of employment opportunities in urban cities are transferred to people at hometowns. Settled migrants provide new migrants assistance, thereby lowering the cost of relocation and increasing the likelihood of obtaining jobs. Social networks, therefore, are expected to positively influence the adoption of family migration. Without utilizing the qualitative data developed in in-depth interviews, there were no appropriate quantitative data to measure social networks. This research, therefore, did not examine or discuss the influence of social networks on migration. Due to missing data, the second important factor which was excluded from this research, was social class or socioeconomic status. On the one hand, social class influences the decision of migration. Stark and his associates (1985) apply the concept of “relative deprivation” to explain migration behavior. They suggest that contrasting his/her situation with other people in the same area motivates a person’s migration decision. 274 Those who have low social economic status are more likely to resort to migration than those who have high social economic status (Stark, et a1. 1985). In addition, social class influences the interactions between family members. R. S. Gallin (1994) argues that family power dynamics were affected by class in Hsin-Hsing Village. As she points out, poor older women in the village had no authority to command the . i‘ labor of their daughters-in—law, and they had to work to secure their own future. Some financially secure older women, however, were able to maintain their life in old age and - direct their daughters-in-law because their husbands owned productive property upon which the younger generation depended. In both instances, the division of labor between women of different generations reflected power dynamics between mothers-in-law and daughters—in—law. But class intervened to influence the direction of the balance of power. This example reveals the importance of the social class/socioeconomic status of villagers for this research. However, quantifying or measuring social class among rural villagers is difficult. In an agrarian society, land usually is utilized to measure the social class of people, because land is the primary means of production. In the mid-19603, although farming was the primary economic source of Hsin-Hsing villagers, the amount of land available to individual families did not sufficiently represent the social class or socioeconomic status of villagers. Most villagers had very little land. There was no variance, in terms of the size of land, among villagers. In the late 19703, the economic structure of the Hsin-Hsing area changed. Farming was no longer the primary economic source of villagers’ livelihood. Land was not the primary means of production. Therefore, 275 the size of land available to Hsin-Hsing families could not measure the social class or socioeconomic status of villagers. How then can social class be measured? What are appropriate indicators to measure social class or the socioeconomic status of villagers? A conventional sociological definition of class is a group of people who have in common a specific power that shapes life chances (Weber, 1978:11). People with property and those without property are two basic categories of all class situations (Weber, 1978:H). A person’s socioeconomic status reflects the style of life, economic conditions, and income level of the person (Dahrendorf, 1959276). When land is not the primary means of production for villagers, it is inefficient to measure the socioeconomic status of villagers. Appropriate indicators for measuring the socioeconomic status of villagers should be those which can directly reflect their economic conditions and income level. In Hsin-Hsing, I suggest using “family income” to measure the socioeconomic status of villagers for two reasons. First, family income not only directly reflects villagers’ economic conditions and income level, but it also influences and/or determines the life style of villagers. Family income can come from a variety of sources. It can be derived from farming family land, working off-farm for wages, and/or operating a business. Family income is much better than the size of land owned for measuring the socioeconomic status of villagers. The size of land owned by villagers does not necessarily reflect their income levels, and, therefore, their life styles and chances. Second, although conjugal units within a complex family may be financially independent, 276 the family still can remain an undivided unit. Therefore, in a rural village, socioeconomic status is usually not discussed at the individual level, but at the family level. For example, a villager is building a mega-house in the village. When people in the village talk about the socioeconomic status of his family members, they generally refer to the family’s socioeconomic status. People do not consider income differentials among family members when discussing class. In the survey conducted in the 19603, Dr. Bernard Gallin and Dr. Rita S. Gallin collected data on villagers’ possessions (e.g., clocks, radios, sewing machines, television sets, bicycles, vehicles, washing machines, and refrigerators) and living conditions (e.g., materials of house wall, floor, and roof, and the ownership and types of toilet, bath, and kitchen). This information theoretically can be used to measure the social class/socioeconomic status of villagers, serving as a proxy for family income. Nevertheless, because of missing data, this dissertation could not rely on the survey data. Among 82 families in Hsin-Hsing Village in 1965, there were only about 30 families for which adequate information on possession of household appliances and living conditions was available. Given the more than 50 families without information on socioeconomic indicators, the variable of social class/socioeconomic status had to be excluded from this dissertation. The issue of missing data also happened in the 1979 survey data. The majority of families did not have adequate information to measure their socioeconomic status. 277 Power dynamics within the family were also not measured directly because of missing data. This research simply discussed the different characteristics of labor migrants, in _ terms of their age and gender, to make inferences about family power dynamics during the labor migration process. Although age and gender traditionally shape the power hierarchy within Taiwanese families, they may not necessarily indicate a person’s status within a family power hierarchy. They cannot directly reflect the dynamics of power within the family during the process of migration. For example, gender and age cannot tell “how” a migration decision is made or “who” makes the migration decision. Therefore, the measurement of family power dynamics during the process of migration decision making needs to be more sophisticated and direct than just using age and gender as proxies. For example, we, first, can identify indicators, which are appropriate for measuring family power dynamics. These indicators can be (1) who participates in the migration decision-making, (2) if migrants have a chance to make a decision about migration, and (3) how the migration decision is made. Methodologically, we can design a series of questions to collect information on these three indicators. No data were available for this type of operation in this dissertation. “Family power dynamics,” therefore, in this research was measured by using gender and age as proxies. In sum, the limitations of this research were caused by the sole utilization of quantitative data and missing data. Limiting this research to the quantitative data and the examination of statistical analyses caused some drawbacks in the research. First, the influencing factor -- social networks -- associated with migration were excluded. Second, this dissertation had to rely on the arguments and examples in the published work of Dr. Bernard Gallin 278 and Dr. Rita S. Gallin derived fi'om the qualitative data. Their arguments and examples, however, are extremely significant in providing meaningfiil contexts to understand migration in the rural village in Taiwan for this dissertation. In addition, their qualitative analyses reveal that research relying solely on statistical analyses is insufficient. Other limitations are because of the issues of missing data. First, because there were more than 50 families without adequate information to measure “social class/socioeconomic status” for the statistical analyses, this variable had to be excluded from the whole research. Second, the same issue led to the measurement of “family power dynamics” by proxy. Gender and age were used to make inferences about the meaning and practice of power within the family. Nevertheless, the future research needs to deal with these issues. First, the future research needs to utilize quantitative data to provide meaningful contexts to understand migration in Taiwan. Second, key factors such as social class/socioeconomic status and social networks needs to be collected, examined, and discussed. These variables are important in understanding migration in Taiwan. 7.4 Some Additional Thoughts It is most usual to end a dissertation by including the implications for policy that emerge from the study upon which the dissertation was based. In this instance, however, such an enterprise is difficult because over 20 years have passed since the second period 279 examined in this research, i.e., the late 19703. Many changes have occurred in Hsin- Hsing and its environs in particular and Taiwan in general since then, making suggestions for policy (based on the 1979 patterns) rather useless. Thus, in the space remaining, 1 document sOme of these changes and discuss how they may influence the migration of villagers. In May 2001, I spent one day in Hsin-Hsing Village. Newly built highways make traveling to the village? much easier than in the 19603 and 19703. A new east-west highway has an exit to the north of the village. A paved two-lane road runs along the north edge of the village, next to the river. Nowadays, automobiles and motorcycles are the two major means of transportation. The trip from Taipei now takes less than four hours rather than the six to eight hours required in the 19603. In addition to changes in transportation facilities, the rural infrastructure has changed dramatically. Two- or three-story buildings abut the road running between Lu-Kang and Hsi-Hu. Factories and retail stores have been erected on agricultural land, and there are few rice fields to be found next to the road. Within the village, there are also many two- or three-story buildings, many equipped with modern appliances such as washing machines, air-conditioners, and stereo systems. In contrast, there are a number of abandoned houses, symbols of the villagers who migrated and settled elsewhere. Farm land also has been converted to industrial use. The only farm land within the village is located on the east-end of the village. Most of villagers I saw were old, and during the day-time they were taking care of grandchildren or were chatting with their fiiends in the 280 grocery store in front of the village or at the temple in the village. The only young people I saw were working in small factories scattered in the village; other young villagers were working locally outside of the village during the day-time. According to my conversation with Dr.Bemard Gallin and Dr. Rita S. Gallin during this visit, I learned some things about the village. First, the oldsters still are the “farmers,” although most farming is now done by entrepreneurs who sell their labor power and work the land with machines. Second, villagers complain about the sluggish economy and several are living in China, having been sent there as managers by the companies for which they work. Third, there are a number of guest workers in the area, including Philippines working in the large mirror factor next to Hsin-Hsing and Thai and Indonesian men working in other nearby industries. Some of these immigrant workers rent vacant houses in the village, although they leave early in the morning and return after 9:00 PM. There also is a Philippine guest worker living in the village who has been hired for work in agricultural production. Fourth, several male villagers have married women from China and Vietnam, and they live in Hsin-Hsing where they have assumed the status of dutiful daughters-in—law.77 Finally, most of the people who now emigrate are those who are highly trained and educated. Those who are less endowed tend to stay in the village, either starting a small enterprise or working in local factories or businesses. These current phenomena serve as the guidelines for this section because they lead me to believe in the continuity of rural out-migration. 77 Under the current law, these women are not allowed to work outside their homes for pay. 281 The geriatrification of agricultural production is a continuing trend in Taiwan (see China Times, December 2, 1999).78 Today, most young villagers in Hsin-Hsing do not know how to farm (personal communication, Dr. Bernard Gallinand Dr. Rita S. Gallin, May 2001), and farming is primarily taken care of by old people. Old farmers hire entrepreneurs to perform rice farming during the first and second crops. In addition, villagers now; rarely grow vegetables during the third crop, which was traditionally a cash crop. Most let their land lie fallow during the third crop because, they complain, growing vegetables is too labor intensive and not profitable. The geriatrification of agricultural production in Hsin-Hsing Village reveals the declining role of agriculture in the national economy and is related to the unprofitability of this sector of the economy. The decline in the importance of agriculture makes it unlikely that young people will be attracted to farming as a source of income. The out-migration of rural young people, therefore, will probably continue. Once Taiwan joins the World Trade Organization (W TO), agricultural production will be even less important for the national economy than it currently is. During the past decade, the Taiwanese government has actively attempted to join the WTO.79 Joining the WTO 78 See also “Farmland Must Not Be the Victim of Conglomerates” at Taiwan Hc@dlincs (http://www.taiwanheadlines. gov.tw/ 19991202/1999120201 .html). 79 According to an on-line document (Taiwan in the WTO: An Economic and Policy Analysis, http://www.taipeiorg/un/wtoOZZB.htm). to establish its own “international space,” Taiwan has tried to continue to the next page... 282 means the reduction of international trading barriers and protective restrictions. On the one hand, Taiwan’s economy will have to be fully liberalized. On the other hand, the Taiwanese government will lose its ability to protect the agricultural sector as it has in the past (see China Times, January 3, 2000).80 Restrictions on the import of agricultural products, including sugar, rice, and tobacco, will have to been removed. Imported agricultural products will replace some produced domestically. Agricultural production will become less profitable than participating in off-farm labor markets, and, therefore, will be very unlikely to attract young villagers. In addition, because domestic agricultural production will be replaced by and/or compete with foreign agricultural production, farming will be less profitable than it is now.81 Changes in the rural economic structure do not happen within a vacuum. They are determined by the global economy. During the 19903, international trade became more competitive than it was, and a number of new industrial countries emerged to challenge “dragons” such as Taiwan. To survive, entrepreneurs in Taiwan need to upgrade their technology to produce better quality commodities and to lower the cost of production, thereby enabling them top competes with NICs that have cheaper labor and raw materials available. Following an international trend, some Taiwanese entrepreneurs have started moving their business to countries in Southeast Asia and China in search of cheaper labor establish an increased presence in international organizations, such as the Asia Pacific Economic Council (APEC), the World Health Organization (WHO), and the World Trade Organization (WTO). 80 See also “Taiwan’s Future Economic Challenge” at Taiwan He@dlines (http://www.taiwanheadlines.gov.tw/20000103/2000010301 .htrnl). 283 than that available on the island. When Taiwanese manufacturers move overseas, they also send some managers to operate their off-shore factories. It is unlikely that this situation will change and the out-migration of Hsin-Hsing people with managerial skills and experience will undoubtedly continue. To stem the flight of manufacturing in the early 19903, the government began to allow manufacturers to import immigrant workers fi'om Southeast Asia countries, especially Thailand, Malaysia, Vietnam, Indonesia, and the Philippines (see Sinorama, December, 1999).82 On the one hand, the arrival of immigrant workers into the rural area may mean that jobs usually held by villagers with a low education will be taken by immigrant workers. On the other hand, local manufacturers may not be able or willing to upgrade to produce competitive products and will be unable to absorb labor with high education in the village.83 When the local sector cannot provide‘employment opportunities which match the educational levels and income expectations of local people, more foreign workers, who are willing to work longer and harder and for less pay, will be imported by 8' Farming land, therefore, will not be able to serve as a safety net for migrants who might lose their jobs in urban centers and return to the rural area in search of work. 82 See also “Foreign Labor Changes the Face of Taiwan” at Taiwan He@dlines (http://th.gio.gov.tw/show.cfm?news_id=486). 83 To upgrade, local manufacturers require a great deal of capital for expensive equipment, appropriate infrastructure, well-educated employees, and research and development (R&D). Because local factories in traditional industries usually are small-scale, such upgrading is problematic. First, new equipment and appropriate infiastructure are expensive, and many manufacturers may not have the necessary capital to underwrite such investments. Second, the owners of these industries usually do not have enough knowledge of new technology and their costs will increase if they have to hire well- educated employees to compensate for their lack of know-how. Third, the owners of these small-scale factories are not will to invest in R&D because it is too risky and unpredictable. In the short-term, then, upgrading factories is costly. Some owners may go bankrupt before they even can recoup their investment or make a profit. For these reasons, owners of small-scale local factories are reluctant to upgrade to produce competitive products. 284 o‘- local manufacturers. It is predictable that those young villagers who have high educational levels will out-migrate to areas that offer them appropriate job opportunities. Local traditional “sunset” industries are likely to be maintained by those villagers with low education and/or immigrant workers. In addition to the “push” factors found in rural areas, there are some “pull” factors that have the potential to lead to the out-migration of villagers, especially young villagers with high education. The Taiwanese government has tried to emphasize high technology and the computer industry, and it has established new science-based industrial parks in a few areas over the past two decades to strengthen its share in the information technology market and expand the national economy (see Forbes, October 9, 2000).84 Manufacturers producing information technology-related products, such as semiconductor chips, require a great deal of capital to establish and operate a business. The factories of the computer industry are unlikely to expand their operation to all places in Taiwan, as was the case when factories were established in hinterland for rural industrial development in the 19703.85 Rather, factories attract many people with high education from all places in Taiwan, including THE ISLAND’S rural areas. In the future, what we will probably see in rural areas are (1) primarily old people who are not willing to migrate and who depend for their livelihood on remittances from their children working in urban cities, and (2) a 8‘ See also “Taiwan: A Partner for Peace and Progress” by Michael Bociurkiw at Taiwan He@dlines (http://www.taiwanheadlines. gov.tw/doubletent112000/ 10_10_pS.htm). 85 High-tech factories require a stable power supply and a good environmental protection plan and system, which are extremely reliant on government assistance and support. With limited resources, the government is not able to create this type of environment island-wide. .285 smaller group of young villagers who operate either small-size family factories establised by their parents in the old days or factories they themselves established more recently. In sum, 1 strongly believe that rural emigration will continue. Viewed fiom the “push” side, first, agricultural production is unprofitable. Second, traditional rural industries are declining. Without upgrading, these local factories are not able to produce competitive industrial products, and/or absorb well-educated young villagers. To survive, factories may hire more and more low-cost immigrant workers or move overseas to seek cheap labor. The emigration of highly educated young villagers is inevitable. Viewed fi'om the “pull” side, newly- developed factories will attract highly educated young villagers. These factories are usually located in specific areas such as a few science-based industrial parks. There, companies pay more than local factories to attract well-educated employees. Inevitably, highly educated young villagers are likely to emigrate to work for these companies than to stay in the village to work in the traditional industry. Given the likelihood of continuing rural out-migration in Taiwan. I would like to offer some suggestions for future research on internal migration there. First, qualitative data must be included in the analysis and qualitative research is necessary. Simply utilizing quantitative data is not enough to discover the underlying meanings of and patterns in relationships between migration and its influential factors. Qualitative analysis provides meaningful contexts to explain social phenomena. Second, key variables such as the social class/socioeconomic status and social networks should be collected, examined, and discussed. Literature demonstrates that social class influences the decision of migration, 286 and power dynamics within the family during the process of migration. Including the social class/socioeconomic status into the internal migration research is necessary for understanding internal migration. In addition, within a Taiwanese society, social networks influencepeople's daily life. Certainly, social networks influence the people's decision of migration. Without including social networks, an internal migration research on Taiwan is incomplete. Fourth, the dynamics of power within the domestic unit during the process of migration needs to be 30phisticatedly measured. Instead of using a proxy the combination of gender and age, the concept of power dynamics during the process of migration needs to be paid more attention. Understanding the dynamics of power within the domestic unit can push the internal migration research forward, because it provides a new perspective for understanding and explaining migration. 287 APPENDIX A The Chronicle of the Migration Policies in the Ching Dynasty (between the mainland China . and Taiwan) Year Regulations 1684-1695 1684 1 696 1718 1732 1 740 1 746 l 747 1760 1761 During this period, labor migration was encouraged by the Ching officers on Taiwan, because of the labor shortage for cultivating. First, migrants should have applied a permission. Second, no wives, children, or another families were allowed to migrate. Third, Hakkas were prohibitated to migrate. Therefore, the only category of legal immigrants was male-Hokkienese. Hakka males were allowed to migrate between the mainland and Taiwan. Migrating back and forth between the mainland China and Taiwan was allowed, but it must have been permitted by the Ching officers. However, this new policy did not differ from the previous. Also, only males were allowed to migrate. Liang-min (good guys) who were willing to registrate in Taiwan’s registration data could apply to migrate to Taiwan with their families. This new policy allowed other family members move with male migrants. All migration was forbidden. Family reunification was allowed. First, grand-parents and parents were allowed to migrate to Taiwan to reunite. Second, young generations were allowed to migrate to Taiwan to take care of old grand-parents and parents. Third, wives and children were not allowed to migrate, but those who migrated with those permitted by the first and second rules were allowed. However, this policy did not change the policy set up in 1740 that all migration was forbidden. Family reunification of Hakkas was allOwed for only one year. Family reunification of all Taiwan immigrants for only one year. The family reunification program in 1760 did not increase many immigrants in Taiwan, so the Ching government decided to prohibit migration between the mainland China and Taiwan again. 288 continued ...... Year Regulations 1776 Families of the Ching officers on Taiwan were allowed to migrate to Taiwan. 1788 All emigrants must have registered in Taiwan registration data. 1875 All emigration was allowed. Sources: Chi, Chia-Lin. 1989. Taiwan Shih (The History of Taiwan). Taipei, Taiwan: Tzu Li Wan Pao Hse. p. 150-3. 289 BIBLIOGRAPHY Agresti, Alan, and Barbara Finlay. 1986. Statistical Methods for the Social Sciences. San Francisco, California: Dellen Publishing Company. Amin, Samir. 1974. “Modern Migrations in Western Africa.” Pp. 64-124 in Samir Amin (ed.), Modern Migrations in Western Afiica. London: Oxford University Press. Asher, Herbert B. 1983. Causal Modeling. Newbury Park, California: SAGE Publications, Inc. Ayeni, B. 1981. “Lagos”. Pp. 17- 55 rn Michael Pacione (ed), Problems and Planning in Third World Cities. London: Croom Helm Ltd. Bociurkiw, Michael. 2000. Taiwan: A Partner for Peace and Progress. [Online] Available http://www.taiwanheadlines. gov.tw/doubletenth2000/ 10_10 _pS.htm, October 9, 2000. Borts, George H., and Jerome L. Stein. 1964. Economic Growth in a Free Market. New York: Columbia University Press. Bowles, S. 1970. “Migration as Investment: Empirical Tests of the Human Investment Approach to Geographic Mobility,” Review of Economics and Statistics. 52(??): 356-362. Boyd, M. 1989. “Family and Personal Networks: Recent Developments and New Agendas,” International Migration Review. 23(3): 638-670. Browning, Harley L. and Waltraut F eindt 1969. “Selectivity of Migrants to A Metropolis 1n A Developing Country: A Mexican Case Study,” Demography. 6(4): 347- 357. Campbell, Karen E., and Peter V. Marsden. 1990. “Recruitment and Selection Processes: The Organizational Side of Job Searches.” Pp. 59-79 in Ronald L. Breiger (ed.), Social Mobility and Social Structure. New York: Cambridge University Press. Chang, Lung-Sheng. 1984. “Regional Disparity and Regional Development Policies in Taiwan.” Pp. 545-572 in Proceedings of the Conference on Urban Growth and Economic Development in the Pacific Region. Taipei, Taiwan: Institute of Economics, Academia Sinica. 290 Chang, Ming-Cheng. 1979. “Migration Selectivity in Taiwan,” Journal of Population Studies. 3: 43-68. Chant, S. 1992. “A Framework for the Analysis of Gender-Selective Migration.” Pp. 197-206 in S. Chant (ed.), Gender and Migration in Developing Countries. New York, NY: Belhaven Press. Chattopadhyay, Arpita. 1998. “Gender, Migration, and Career Trajectories in Malaysia.” Demography 35:335- 344 Chi, Chia-Lin. 1989. Taiwan Shih (The History of Taiwan). Taipei, Taiwan: Tzu Li Wan Pao Hse. Chi, Peter S. K. 1991. “Economic Aspects of Changing Family Structure in Taiwan,” Human Ecology Forum. 19(4): 24-27. Chiang, Lan-Hung. 1978. “Male and Female Out-migration in Taiwan: An Examination of Characteristics and Propensities to Migrate,” Journal of Population Studies. 2:60-81. China Times 1999. Farmland Must Not Be the Victim of Conglomerates. [Online] Available http ://www.taiwanheadlines. gov.tw/ 1 999 1 202/1 999 12020 1 .htrnl, December 2, 1999. China Times 2000. Taiwan’s Future Economic Challenge. [Online] Available http://www.taiwanheadlines.gov.tw/20000103/2000010301.htrnl, January 3, 2000. Chuang, Ying-Chang 1972. “Tai-wan Nung-tsun Chia-tsu Tsui Hsin-tai-hua Te Shih-ying: I-ge Tien-yieh- tiao-cha-shih-li Te F en-hsi,” Chung-yiang-yien-chiou-yuan Min-tsu-hsueh Y ian- chiou-suo Chi-kan. Vol. 34. Clark, Cal. 1989. Taiwan ’3 Development: Implications for Contending Political Economy Paradigms. New York: Greenwood Press. Cohen, Myron L. 1976. House United, Housed Divided. The Chinese Family m Taiwan. New York: Columbia University Press. Connell, John, B. Dasgupta, Roy Laishley, and Michael Liption. 1976. Migration from Rural Areas: The Evidence from Village Studies. Oxford: Oxford University Press. 291 Connelly, R. 1992. “The Effect of Child Care Costs on Married Women’s Labor Force Participation.” The Review of Economics and Statistics 74:83-90. I Council for Economic Planning and Development, Republic of China. 1994. Taiwan Statistical Data Book. Taipei. Craig, J .E. 1981. “The Expansion of Education.” Review of Research in Education. 9: 151-210 Dahrendorf, Ralf. - 1959. Class and Class Conflict in Industrial Society. Stanford, CA: Stanford University Press. DaVanzo, Julie. 1981. “Microeconomic Approaches to Studying Migration Decisions.” Pp. 90-129 in . Gordon F. De J ong and Robert W. Gardner (eds), Migration Decision Making: Multidisciplinary Approaches to Microlevel Studies in Developed and Developing Countries. New York, NY: Pergamon Press. Deshpande, C. D., and B. Arunachalam. 1981. “Bombay”. Pp. 187-217 in Michael Pacione (ed), Problems and Planning in Third World Cities. London: Croom Helm Ltd. DeVoe, Pamela A. 1987. “Structural Considerations in the Contemporary Taiwanese Farm Family: Survival of the Large Family Ideal,” International Journal of Sociology of the Family. 17(1): 57-65. Diamond, Norma. 1969. K ’un Shen: A Taiwan Village. Chicago: Holt, Rinehart and Winston. Dinerman, Ina R. 1978. “Patterns of Adaptation Among Households of U.S.-Bound Migrants fi'om Michoacan, Mexico,” International Mgration Review. 12(4): 485-501. Fei, Hsiao-tung. 1939. Peasant Life in China. London: Routledge. Fei, John C. H., Gustav Ranis, and Shirley W. Y. Kuo. 1979. Growth with Equity: The Taiwan Case. New York: Oxford University Press. Ferdinand, Peter. 1996. “The Taiwanese Economy.” Pp. 37-65 in T ake-ofl for Taiwan, Peter Ferdinand (ed.). London, UK: The Royal Institute of International Affairs. 292 Findley, Sally E. 1987. “An Interactive Contextual Model of Migration in Ilocos Norte, The Philippines,” Demography. 24 (2): 163-190. Findley, Sally E. 1987. Rural Development and Migration: A Study of Family Choices in the Philippines. Boulder, Colorado: Westview Press. Freedman, Maurice. 1979. The Study of Chinese Society: Essays by Maurice Freedman. Selected and Introduced by G. W. Skinner. Stanford: Stanford University Press. Freedman, Ronald, Ming-Cheng Chang, Te-Hsiung Sun, Maxine Weinstein. 1994. “The Fertility Transition in Taiwan.” Pp. 264-304 in Arland Thornton and Hui- Sheng Lin (eds), Social Change and the Family in Taiwan. Chicago, IL: The University of Chicago Press. Frey, William H. 1987. “Migration and Depopulation of the Metropolis: Regional Restructuring or Rural Renaissance?” American Sociological Review. 52(2): 240-257. Frey, William H. 1990. “Metropolitan America: Beyond the Transition,” Population Bulletin. 45(2): 3-49. Fried, Morton H. 1974. Fabric of Chinese Society: A Stuay of the Social Life of A Chinese County Seat. New York: Octagon Books. Friedl, E. 1967. “The Position of Women: Appearance and Reality.” Anthropology Quarterly 9: 15 1-210. Fukutake, Tadashi. 1967. Asian Rural Society: China, India, Japan. Seattle: University of Washington Press. Gallin, Bernard. 1966. Hsin Hsing, Taiwan: A Chinese Village in Change. Berkeley, CA: University of California Press. Gallin, Bernard. 1967. “Chinese Peasant Values Toward the Land.” Pp. 367-375 in Jack M. Potter, May N. Diaz, and George M. Foster (eds), Peasant Society: A Reader. Boston: Little, Brown and Company. 293 Gallin, Bernard. 1978. “Rural to Urban Migration in Taiwan: Its Impact on Chinese Family and Kinship.” Pp. 261-282 in David C. Buxbaum (ed.), Chinese Family Law and Social Change: in Historical and Comparative Perspectives. Seattle, WA: University of Washington Press. Gallin, Bernard, and Rita S. Gallin. 1974. “The Integration of Village Migrants in Taipei.” Pp. 331-358 in Mark Elvin and William Skinner (eds), The Chinese City between Two Worlds. Stanford, CA: Stanford University Press. Gallin, Bernard, and Rita S. Gallin. 1982a. “Socioeconomic Life in Rural Taiwan: Twenty Years of Development and Change,” Modern China. 8(2): 205-246. Gallin, Bernard, and Rita S. Gallin. 1982b. “The Chinese Joint Family in Changing Rural Taiwan.” Pp. 142-158 in Sidney L. Greenblat, Richard W. Wilson, and Amy Auerbacher Wilson (eds), Social ‘ Interaction in Chinese Society. New York, NY: Praeger Publishers. Gallin, Rita S. 1984. “Women, Family and the Political Economy of Taiwan,” Journal of Peasant Studies. 12(1): 76-92. Gallin, Rita's. 1986. “Mothers-in-law and Daughters-in-law: Integeneration Relations Within the Chinese Family in Taiwan.” Journal of Cross-Cultural Gerontology 1: 31-49 Gallin, Rita S. 1994. “The Intersection of Class and Age: Mother-in-law/Daughter-in-law Relations in Rural Taiwan.” Journal of Cross-Cultural Gerontology 9. Gallin, Rita S. 1995. “Engendered Production in Rural Taiwan: The Ideological Bonding of the Public and Private.” Pp. 113-134 in Engendering Wealth and Well-Being, edited by Rae L. Blumberg, Cathy A. Rakowski, Irene Tinker, and Michael Monteon. Bouder: Westview Press. Gallin, Rita S. 2001. “Daughters Cry at Your Funeral: The Female Family in Rural Taiwan.” Journal of Asian Women (forthcoming). Gallin, Rita S., and Bernard Gallin. 1992. “Hsin Hsing Village, Taiwan: From Farm to Factory.” Pp. 279-293 in Ronald G. Knapp (ed.), Chinese Landscapes: the Village as Space. Honolulu, HA: University of Hawaii Press. 294 Gardner, Ross Chiffelle. 1986. Family Economy and Division: An Investigation into the Family Response to Economic Change in A Rural Taiwanese Community. Unpublished Master Thesis. Department of Anthropology, Michigan State University, East Lansing, Michigan. Gates, Hill. 1987. Chinese Working Class Lives: Getting By in Taiwan. Ithaca, NY: Cornell University Press. Goldscheider, Clavin. 1987. “Migration and Social Structure: Analytic Issues and Comparative Perspectives in Developing Nations,” Sociological Forum. 2(4): 674-696. Greenhalgh, Susan. 1985. “Sexual Stratification: The Other Side of ‘Growth with Equity’ in East Asia.” Population and Development Review 1 12265-314. Greenhalgh, Susan. _ - 1990. “Land Reform and Family Entrepreneurship in East Asia.” Pp. 77-118 in Geoffrey McNicoll and Mead Cain (eds), Rural Development and Population: Institutions and Policy. New York: The Population Council. Grigg, David. 1980. Population Growth and Agrarian Change. New York, NY: Cambridge University Press. Grindstaff, Carl F., and Frank Trovato. 1990. “Junior Partners: Women’s Contribution to Family Income in Canada.” Social Indicators Research 22:229-253. Guest, Philip. 1989. Labor Allocation and Rural Development. Colorado: Westview Press. Harrell, Stevan. 1982. Ploughshare Village: Culture and Context in Taiwan. Seattle, WA: University of Washington Press. Harris, J. R., and Michael P. Todaro. 1970. “Migration, Unemployment and Development: A Two-Sector Analysis,” American Economic Review. 60(1): 139-149. Hermalin, Albert, Paul K. C. Liu, and Deborah. Freedman. 1994. “The Social and Economic Transformation of Taiwan.” Pp. 49-87 in Arland Thornton and Hui-Sheng Lin (eds), Social Change and the Family in Taiwan. Chicago, IL: The University of Chicago Press. 295 Ho, Samuel PS. 1978. Economic Development of Taiwan, 1860-1970. New Haven: Yale University Press. Ho, Samuel PS. 1979. “Decentralized Industrialization and Rural Development: Evidence from Taiwan.” Economic Development and Cultural Change 28(1): 77-103. Holzer, Harry J. 1987. “Hiring Procedures in the Firm: Their Economic Determinants and Outcomes,” National Bureau of Economic Research Working Paper No. 2185. Holzer, Harry J. 1988. “Search Method Use by Unemployed Youth,” Journal of Labor Economics. Vol. 6( 1): 1 -20. Hsu, Francis Lang-Kwang. 1943. “The Myth of Chinese Family Size,” The American Journal of Sociology. 68: 555-562. Hsu, Francis Lang-Kwang. 1948. Under the Ancestors ' Shadow: Chinese Culture and Personality. New York: Columbia University Press. Hu, Hsien-chin. . 1948. The Common Descent Group in china and Its Functions. New York: Viking Fund. Huang, N. C. 1984. “The Migration of Rural Women to Taipei.” Pp. 247-268 in J. T. Fawcett, S. Khoo and P. C. Smith. Boulder (eds), Women in the Cities of Asia: Migration and Urban Adaptation, Colorado: Westview Press. Huang, Sophia Wu. . 1993. “Structural Change in Taiwan’s Agricultural Economy,” Economic Development and Cultural Change. 42(1): 43-65. Hwang, Dolly Y. 1991. The Rise of A New World Economic Power: Postwar Taiwan. New York: Greenwood Press. Ka, Chih-Ming. 1995. Japanese Colonialism in Taiwan: Land Tenure, Development, and Dependency, 1 895-1 945. Oxford: Westview Press, Inc. Koo, Anthony K. C. 1968. The Role of Land Reform in Economic Development: A Case Study of Taiwan. New York: Praeger. 296 Ladinsky, Jack. 1967. “Sources of Geographic Mobility among Professional Workers: A Multivariate Analysis,” Demography. 4(1): 293-309. Lang, Olga. 1946. Chinese Family and Society. New Haven: Yale University Press. Lee, Everett S. 1966. “A Theory of Migration,” Demography. 3(1): 50. Lee, Teng-Hui. 1971. Intersectoral Capital Flows in the Economic Development of Taiwan, 1895-1960. Ithaca, N. Y.: Cornell University Press. Lewis, W. Arthur. 1954. “Economic Development with Unlimited Supplies of Labor,” The Manchester School of Economic Social Studies. 22(2): 139-191. Li, Chun-Hao. 1997. “Causal Modeling of Internal Migration in Taiwan.” Paper presented to the British Society for Population Studies Annual Meetings, Exeter, England, September 1997. Li, Chun-Hao. 1998. “The Migrant Women’s Participation in Labor Markets: A Taiwanese Case.” Paper presented to the American Sociological Association Annual Meetings, San Francisco, California, August 1998. Li, Chun-Hao, Brendan P. Mullan, Rita S. Gallin, Bernard Gallin. 1996. “Household and Family Internal Migration in Taiwan.” Paper presented to the Population Association of American Annual Meetings, New Orleans, May 1996. Li, Tung-ming. 1974. “Tai-wan Ren Kou Hsing Chuang Pieh Chien Yi Lu Ca Yi Chih Yiah Chiou (in Chinese),” Tai-wan Wen Hsien. 25(2): 17-26. Liao, Cheng-hong. 1977. “Tai-wan Nung Tsun Lao Li Yi Tung Chih Yian Chiou (in Chinese),” T ai-wan Yin Hang Chi Kan. 28(4): 151-191. Lin, Hui-Sheng, Mei-Lin Lee, and Arland Thornton. . 1994. “Trends in the Timing and Prevalence.” Pp. 202-224 in Arland Thornton and Hui- Sheng Lin (eds), Social Change and the Family in Taiwan. Chicago, IL: The University of Chicago Press. Lin, Kuan-Pin. 1985. The Structure of Personal Consumption in Taiwan: 1 951-1 980. Economic Monograph Series No. 6. Taipei: Chung-Hua Institution for Economic Research. 297 Liu, Paul K. C. 1982. “Labor Mobility and Utilization in Relation to Urbanization in Taiwan (Part 1),” Industry of Free China. 57(5): 1-12. Liu, Paul K. C. 1982. “Labor Mobility and Utilization in Relation to Urbanization in Taiwan (Part 11),” Industry of Free China. 57(6): 7-17. Liu, Paul K. C. 1983. “Factors and Policies Contributing to Urbanization and Labor Mobility in Taiwan,” Industry of Free China. 59(5): 1-20. Liu, Paul K. C. 1993. “Demographic Issues and Trends as Reported in the 1990 Census of Taiwan,” Industry of Free China. 80(2): 41-51. Liu, Paul K. C., and Hsung-Hsiung Tsai. 1990. “Urban Growth and Employment Restructuring in Taiwan,” Industry of Free China. 74(2): 17-33. Long, Larry H. 1973. “Migration Differentials by Education and Occupation: Trends and Variations,” Demography. 10(2): 243-258. ' Long, Larry. 1992 “Changing Residence: Comparative Perspectives on its Relationship to Age, Sex, and Marital Status,” Population Studies. 46(1): 141-158. Lu, Min-J en. 1981. “Taiwan’s Economic Development -- Retrospect and Prospect.” Paper presented ' in the Conference on the History of the Republic of China, Taiwan, August 1981. MacDonald, John S., and Leatrice D. MacDonald. 1974. “Chain Migration, Ethnic Neighborhood Formation, and Social Networks.” Pp. 226-235 in Charles Tilly (ed.), An Urban World. Boston, MA: Little, Brown. Markham, William T., P. O. Macken, C. M. Bonjean, and J. Corder. 1983. “A Note on Sex, Geographical Mobility and Career Advancement.” Social Forces 61 :1 138-46. Massey, Douglas S. 1988. “Economic Development and International Migration in Comparative Perspective,” Population and Development Review. 14(3): 383-413. Massey, Douglas S. 1990. “Social Structure, Household Strategies, and the Cumulative Causation of Migration,” Population Index. 56(1): 3-26. 298 Massey, Douglas S. 1990. “The Social and Economic Origins of Immigration,” The ANNALS of the American Academy of Political and Social Science. 510: 60-72. Massey, Douglas S., and Felipe Garcia-Espana. 1987. “The Social Process of International Migration,” Science. 237: 733-738. Massey, Douglas S., Joaquin Arango, Graeme Hugo, Ali Kouaouci, Adela Pellegrino, and J Edward Taylor. 1993. “Theories of International Migration: A Review and Appraisal,” Population and Development Review. 19(3): 431-466. Massey, Douglas S., Joaquin Arango, Graeme Hugo, Ali Kouaouci, Adela Pellegrino, and J Edward Taylor. 1994. “An Evaluation of International Migration Theory: The North American Case,” Population and Development Review. 20(4): 699-751. Massey, Douglas S., Rafael Alarcon, Jorge Durand, and Humberto Gonzalez. 1987. Return to Aztlan: The Social Process of International Migration from Western Mexico. Berkeley, CA: University of California Press. Mastel, Greg. 1999. Taiwan in the WTO: An Economic and Policy Analysis. [Online] Available http://www.taipei.org/un/wt00223.htrn, October 6, 1999. Maxwell, N. L. 1988. “Economic Returns to Migration: Marital Status and Gender Differences.” Social Science Quarterly 69: 108-121 . McAleavy, Henry. 1955. “Socialism and the Chinese Peasant Family,” Journal of Asian Studies. 34(???): 613-30. Mincer, Jacob. 1978. “Family Migration Decisions.” Journal of Political Economy 86:749-773. Mines, Richard. 1984. “Network Migration and Mexican Rural Development: A Case Study.” Pp. 136- 155 in Richard C. Jones (ed.), Patterns of Undocumented Migration: Mexico and the United States. Totowa, NJ: Rowman and Allanheld, Publishers. Mines, Richard, and Alain de J anvry. 1982. “Migration to the United States and Mexican Rural Development: A Case Study,” American Journal of Agricultural Economics. 64(3): 444-454. Mullan, Brendan P. 1986. “The Impact of Social Networks on the Occupational Status of Migrants,” International Migration. 27(1): 69-86. 299 Mullan, Brendan P., Chun-Hao Li, Brendan P. Mullan, Rita S. Gallin, Bernard Gallin. 1998. “Family and Internal Migration in Taiwan,” Asian and Pacific Migration Journal. 7(1 ): 43-66. Muth, Richard F. 1968. “Differential Growth Among Large US. Cities”. in James P. Quirk and Arvid M. Zarley (eds), Papers in Quantitative Economics. Lawrence: The University Press of Kansas. Oberai, A. S., Pradhan H. Prasad, and M. G. Sardana. 1989. Determinants and Consequences of Internal Migration in India: Studies in Bihar, Kerala, and Uttar Pradesh. Delhi: Oxford University Press. O’Connor, S. M. 1988. “Women’s Labor Force Participation and Preschool Enrollment: A Cross- National Perspective, 1965-80.” Sociology of Education 61:15-28. Park, Albert, and Bruce Johnston. . 1995. “Rural Development and Dynamic Extemalities in Taiwan’s Structural Transformation,” Economic Development and Cultural Change. 44(1): 181-208. Pamwell, Mike. 1993. Population Movements and the Third World. London: Routledge. Pedhazur, Elazar J. 1982. Multiple Regression in Behavior Research: Explanation and Prediction. Forth Worth, TX: Holt, Rinehart and Winston, Inc. Presser, H.B., and W. Baldwin. 1980. “Child Care as a Constraint of Employment: Prevalence, Correlates, and Bearing on the Work and Fertility Nexus.” American Journal of Education 85: 1201 -12 1 3. Quinn, N. 1977. “Anthropological Studies on Women’s Status.” Annual Anthropological Review 6: 1 81—225. Ranis, Gustav, and John C. H. Fei. 1961. “A Theory of Economic Development,” American Economic Review. 51(4): 533- 565. Ravenstein, E. G. 1885. “The Laws of Migration” (part 1), Journal of T he Statistical Society. 48:167-227. Ravenstein, E. G. 1889. “The Laws of Migration” (part 2), Journal of The Statistical Society. 52:241-301. 300 Rees, Albert. 1966. “Labor Economics-Effects of More Knowledge: Information Networks in Labor Markets,” The American Economic Journal. 56(2): 559-566. Ritzer, George. 1983. Sociological Theory. New York, NY: Alfred A. Knopf, Inc. Roos, P. A. 1983. “Marriage and Women’s Occupational Attainment in Cross-Cultural Perspective.” American Sociological Review 48:852-864. Salaff, J. 1981. “Singapore Women.” in N. Black and A. B. Cottrell (eds), Women and World Change. Beverly Hills, CA: Sage Publications. Sandell, S. H. 197 7. “Women and the Economics of Family Migration.” Review of Economics and Statistics 59:406-414. Saraceno, C. 1984. “Shifts in Public and Private Boundaries: Women as Mothers and Service Workers in Italian Daycare.” Feminist Studies 10. Schmink, M. 1984. “Household Economic Strategies: A Review and Research Agenda,” Latin American Research Review. 19(3): 87-101. Shaw, Paul. 1975. Migration Theory and Fact: A Review and Bibliography of Current Literature. Philadelphia, PA: Regional Science Research Institute. Shih, Ming. 1980. Four Hundred Years of Taiwanese History. San Jose, California: Paradise Culture Associates. Shihadeh, Edward. 1991. “The Prevalence of Husband- Centered Migration: Employment Consequences for Married Mothers.” Journal of Marriage and the Family 53: 432-444. Sinorama 1999. Foreign Labor Changes the Face of Taiwan. [Online] Available http://th.gio.gov.tw/show.cfm?news_id=486, December 1999. Sjaastad, Larry A. 1962. “The Costs and Returns of Human Migration,” Journal of Political Economy. 70S: 80-93. 301 Smith, James P., and Duncan Thomas. 1993. On the Road: Marriage and Mobility in Malaysia. Labor and Population Working Papers Series 93-11 (DRU-323-NICHD). Santa Monica: Rand Corporation. Smith, James P., and Duncan Thomas. 1998. “On the Road: Marriage and Mobility in Malaysia.” Journal of Human Resources 33:805-832. Speare, Alden, Jr. 1971. “A Cost-Benefit Model of Rural to Urban Migration in Taiwan,” Population Studies. 25(1): 117-130. Speare, Alden, Jr. _ 1974. “Urbanization and Migration in Taiwan,” Economic Development and Cultural Change. 22(2): 302-319. Speare, Alden, Jr., Sidney Goldstein, and William H. Frey. 1975. Residential Mobility, Migration, and Metropolitan Change. Cambridge, MA: Ballinger. Stark, Oded. 1983. “A Note on Modeling Labour Migration in LDCs,” The Journal of Development Studies. 19(4): 539-543. Stark, Oded. 1984. “Rural-to-Urban Migration in LDCs: A Relative Deprivation Approach,” Economic Development and Cultural Change. 32(3): 475-486. Stark, Oded. 1991. “Migration in LDCs: Risk, Remittances, and the Family,” Finance and Development. 28(4): 39-41. Stark, Oded, and David E. Bloom. 1985. “The New Economics of Labor Migration,” The American Economic Review. 75(2): 173-178. Stark, Oded, and J. Edward Taylor. 1989. “Relative Deprivation and International Migration,” Demography. 26(1): 1-14. Stark, Oded, and J. Edward Taylor. 1991. “Migration Incentives, Migration Types: The Role of Relative Deprivation,” The Economic Journal. 101(408): 1163-1178. Stark, Oded, J. Edward Taylor, and Shlomo Yitzhaki. 1986. “Remittances and Inequality,” The Economic Journal. 86(3 83): 722-740. 302 Stark, Oded, and Robert E. B. Lucas. 1988. “Migration, Remittances, and the Family,” Economic Development and Cultural Change. 36(3): 465-481. Stouffer, Samuel A. 1940. “Intervening Opportunities: A Theory Relating Mobility and Distance,” American Sociological Review. 5(6): 845-867. Stouffer, Samuel A. 1960. “Intervening Opportunities and Competing Migrants,” Journal of Regional Science. 2(1): 1-26. Sun, Chen, and Hsung-hsiung Tsai. 1981. “Urbanization, Regional Growth and Economic Development in Taiwan, the Republic of China,” Industry of Free China. 56(4): 5-13. Taeuber, Irene B. 1970. “The Families of Chinese Farmers.” Pp. 63-86 in Maurice Freedman (ed.), Family and Kinship in Chinese Society. Stanford: Stanford University Press. Tang, Mei-Chun. 1985. “Equal Right and Domestic Structure.” Pp. 61-69 in The Chinese Family and Its Ritual Behavior, Hsieh J ih-Chang and Chuang Ying-Chang (eds.). Taipei: Institute of Ethnology, Academia Sinica. Taylor, J. Edward. . 1986. “Differential Migration, Networks, Information and Risk.” Pp. 147-171 in Oded Stark (ed.), Research in Human Capital and Development: Migration, Human Capital, and Development (vol. 4). Conn.: JAI Press Inc. Thomas, Brinley. , 1954. Migration and Economic Growth. London: Cambridge University Press. Thomas, Dorothy S. 1941. Social and Economic Aspects of Swedish Population Movements: 1750-1933. New York: Macmillan. Thorbecke, Erik. 1979. “Agricultural Development.” Pp. 132-205 in Walter Galenson (ed.), Economic Growth and Structural Change in Taiwan: the Postwar Experience of the Republic of China. Ithaca, NY: Cornell University Press. Tilly, Charles, and C. H. Brown. 1967. “On Uprooting, Kinship, and the Auspices of Migration,” International Journal of Comparative Sociology. 8(1): 139-164. 303 Todaro, Michael P. 1969. “A Model of Labor Migration and Urban Unemployment in Less Developed Countries,” American Economic Review. 59(1): 138-148. Todaro, Michael P. 1976. Internal Migration in Developing Countries: A Review of Theory, Evidence, Methodology and Research Priorities. Geneva: International Labour Office. Todaro, Michael P. 1980. “Internal Migration in Developing Countries: A Survey.” Pp. 361-403 in Richard A. Easterlin (ed.), Populatibn and Economic Change in Developing Countries. Chicago, Illinois: University of Chicago Press. Treiman, D. J ., and K. Terrell. 1975. “Sex and the Process of Status Attainment: A Comparison of Working Women and Men.” American Sociological Review 40:174-200. Tsai, Hong-Chin. 1978. “Development Policy and Internal Migration in Taiwan,” Journal of Population Studies. 2: 27-59. Tsai, Hong-Chin. 1979. “Urban and Rural Population Redistribution and Changes in Urbanization Process in Taiwan: Characteristics, Problems and Policy Implications,” Journal of Population Studies. 3: 1-42. Tsai, Hong-Chin. 1981. “Rural Industrialization in Taiwan: Its Structure and Impact on the Rural Economy,” Industry of Free China. 56(6): 17-32. Tsai, Hsung-Hsiung. 1979. Urban Growth and the Change of Spatial Structure in Taiwan. Taipei: Academia Sinica: Report of Land Use Planning Conference. Tsai, Hsung-hsiung. 1990. “Population Decentralization Policies: The Experience of Taiwan,” Industry of Free China. 73(3): 1-12. United Nations. 1985. World Population Trends, Population and Development Interrelations and Population Policies:1983 Monitoring Report, Vols. I and II. New York. Unknown. 1977. Land Reform Program Implemented in the Republic of China. Weber, Max. 1978. Economy and Society (vol. 1 and 2). California: University of California Press. 304 Wolf, M. , 1972. Women and the Family in Rural Taiwan. Stanford, CA: Stanford University Press. Wood, Charles H. 1981. “Structural Changes and Household Strategies: A Conceptual Framework for the Study of Rural Migration,” Human Organization. 40(4): 338-344. Veevers, Jean. 1977. “The Family in Canada.” In Profile Studies (Vol. 5, Part 3). Ottawa: Statistics Canada. Wang, Sung-Hsing, and Raymand Apthorpe. 1974. Rice Farming in Taiwan: Three Village Studies. Monograph Series B, No. 5. Taipei: Institute of Ethnology, Academia Sinica. Yang, Martin M. C. 1945. A Chinese Village: T aitou,Shantung Province. New York: Columbia University Press. Yang, Martin M. C. 1970. Socio-Economic Results of Land Reform in Taiwan. Honolulu: East-West Center Press. Yin, Chien-Chung. 1978. “Homogeneous Migration and Heterogeneous Migration: A General Typology of Anthropological Migration Studies,” Journal of Population Studies. 2: 1-26. Zachariah, K. C. 1966. “Bombay Migration Study: A Pilot Analysis of Migration to An Asian Metropolis,” Demography. 3(2):378-392. 305 IIIIIIIIIIIIIIIIIIIII WIlllllfllillllllzlllflllllll