INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

University Microfilms International A Bell & Howell Information Company 300 North Zeeb Road. Ann Arbor. Ml 48106-1346 USA 313/761-4700 800/521-0600 Order Number 9316207

Structural correlates of rural poverty in the Republic of Korea

Park, Dae Shik, Ph.D.

The Ohio State University, 1993

Copyright ©1993 by Park, Dae Shik. All rights reserved.

UMI 300 N. ZeebRd. Ann Arbor, MI 48106 Structural Correlates of Rural Poverty in the Republic of Korea

DISSERTATION

Presented in Partial Fulfillment of the Requirement for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

By

Dae Shik Park, B.A., M.A.

*****

The Ohio State University 1993

Dissertation Committee: Approved by Dr. Joseph F. Donnermeyer Dr. Wen L. Li /j {/ Advisor Q C/ Department of Agricultural Dr. Linda M. Lobao Economics and Rural Sociology Copyright by Dae Shik Park 1993 To My Parents

ii ACKNOWLEDGMENTS

I express sincere appreciation to Dr. Joseph F. Donnermeyer, my major advisor, for his encouragement, guidance, and support throughout the research. His advise will be kept in my heart as an everlasting guide for my future life. Sincere gratitude is also expressed to Drs. Wen Li and Linda Lobao who served on my dissertation committee for their valuable suggestions and comments. Their critiques helped to clarify thoughts and improve the research. Thanks also goes to Dr. Don Thomas who served as a member of my graduate program committee. I also wish to express gratitude to various professors in the Department of Sociology, Pusan National University, who guided me into the world of sociology. Special thanks are expressed to the late Dr. Dong Shik Hong, who taught me the duty and way of a rural sociologist. Thanks are also expressed to the many faculty, friends, and graduate students who showed interest and provided encouragement to continue the research. Finally, I express my thanks to my parents and family and to my lovely wife, Hwa Soon. Without their love and consideration, this study would not have been possible. Also, I would like to share delights with my baby who will be born soon. VITA

1959 ...... Born - Sangju, Republic of Korea 1983 ...... B.A., Department of Sociology, Pusan National University, Republic of Korea 1983-1985 ...... Research Assistant, Social Survey Research Center, Pusan National University, Republic of Korea 1985 ...... M.A., Department of Sociology, Pusan National University, Republic of Korea 1985-1987 ...... Military Service, Radar Operator 1988-1989 ...... Research Assistant, Department of Sociology, Pusan National University, Republic of Korea 1990-1992 ...... Graduate Research Associate, Department of Agricultural Economics and Rural Sociology, Ohio State University, Columbus, Ohio

PUBLICATIONS

Kang, Dae Ki, Dong Shik Hong, and Dae Shik Park (1983), Rural- to-Urban Migration Intention and Their Social Factors amoncr Korean Rural Youths. Research Monograph 1, Social Survey Research Center, Pusan National University, Republic of Korea. Hong, Dong Shik, Dae Shik Park, and others (1985), The Use of Telecommunication Information and Long Outlook for the Embodiment of information Society. Telecommunication Research Monograph, Social Survey Research Center, Pusan National University, Republic of Korea. Hong, Dong Shik and Dae Shik Park (ed. and trans.) (1985), The v Sociology of Community. Seoul, Republic of Korea: Kyung Mun Sa Publishing Co. Hong, Dong Shik and Dae Shik Park (1988), "A study on the socio-economic ties between farmers and urban laborers: based on farmers' perspectives." Journal of Social Survey Research. 7 (1):93-107. Park, Dae Shik (1985) , A study on the Effectiveness of the Rural Organizations in Korean Villages. Department of Sociology, Pusan National University, Republic of Korea. Park, Dae Shik (1985), "Social conflicts and counter-measures in the Republic of Korea." Nak-Gang Chun-Chu. 2 (1):112-121.

FIELDS OF STUDY

Major Field: Agricultural Economics and Rural Sociology

Rural Sociological Theory Dr. Joseph F. Donnermyer Dr. Donald W. Thomas Sociology of Agriculture and Dr. Linda M. Lobao Natural Resources

Sociological Research Dr. Wen L. Li Methodology TABLE OF CONTENTS

DEDICATION ...... ii ACKNOWLEDGMENTS ...... iii VITA ...... v LIST OF TABLES ...... ix LIST OF FIGURES ...... xii CHAPTER PAGE I. INTRODUCTION ...... 1 Statement of the Problem ...... 1 Purposes of Study ...... 6 Significance of Study ...... 8 Overview ...... 10 II. AGRICULTURE AND RURAL POVERTY RELATED INDICATORS IN THE REPUBLIC OF KOREA: HISTORICAL TRENDS ... 11 Introduction ...... 11 Korean Agriculture ...... 11 Historical Trends in Rural Poverty Related Indicators ...... 23 III. THEORETICAL EXPLANATION OF RURAL POVERTY ...... 3 0 Introduction ...... 3 0 Individual Level-based Perspectives ...... 31 Human Capital Theory ...... 31 Functionalist Perspectives ...... 32 Structural Level-based Perspectives ...... 3 6 Marxian Perspectives ...... 3 6 Marxian Class Theory ...... 36 Dependency/World System Theory ...... 38 Non-Marxian Perspectives ...... 44 Dual (Segmented) Economy Theory ...... 44 Goldschmidt Hypothesis ...... 4 6

vii IV. RESEARCH MODEL AND HYPOTHESES ...... 51 Introduction ...... 51 Research Model ...... 51 Hypotheses ...... 56 Farm Structure/Characteristics and Rural Poverty ...... 56 Non-Farm Characteristics andRural Poverty .... 60 Spatial/Political Economical Characteristics and Rural Poverty...... 62 V. RESEARCH METHODOLOGY ...... 69 Introduction ...... 69 Data Units and Collection ...... 69 The Measurement of Variables ...... 71 The Measurement of Independent Variables ..... 72 The Measurement of Dependent Variables ...... 77 Methods of Analysis ...... 80 VI. FINDINGS ...... 83 Introduction ...... 83 Descriptive Statistics of the Independent and Dependent Variables ...... 83 Multiple Regression Analysis ...... 104 National Patterns ...... 110 Blocked Regression ...... 118 Regional Comparisons ...... 126 VII. SUMMARY AND CONCLUSIONS ...... 13 5 Summary of Findings ...... 135 Descriptive Statistics of the Variables ..... 135 Multiple Regression Analysis ...... 13 8 Policy Implications ...... 147 Limitations of the Research ...... 153 Recommendations for Future Research ...... 154 NOTES ...... 157 BIBLIOGRAPHY ...... 158

viii LIST OF TABLES

TABLE PAGE 2.1 Major Korean Agriculture Related Indicators ... 13 2.2 Comparison of Households' Real Income Between Farmers and Urban Laborers...... 16 2.3 Numbers of Farm Households by Size of Cropland ...... 17 2 .4 Trends of Land Ownership ...... 19 2.5 Composition of Farm Household Income ...... 21 2.6 The Proportion of Farm Income to Living Expenses by Farm Size ...... 22 2.7 Historical Trends of Official Poverty Population...... 25 2.8 Official Rural and Urban Poverty Population ...... 27 2.9 Trends of Inequality in Income Distribution ... 28 5.1 The Level of International Competition Power of Main Agricultural Commodities ...... 76 6.1 Distribution of Percent Smaller Farming ...... 84 6.2 Distribution of Percent Middle Farming ...... 86 6.3 Distribution of Percent Larger Farming ...... 87 6.4 Distribution of Percent Part-time Farming ..... 88 6.5 Distribution of Tenancy Rate ...... 90 6.6 Distribution of Percent Non-farm Population ... 91 6.7 Distribution of Percent Female ...... 92 6.8 Distribution of Percent High School Graduates ... 94 ix 6.9 Distribution of Region 1 (YOUNGNAM) ...... 95 6.10 Distribution of Region 2 (HONAM)...... 96 6.11 Distribution of Proximity to Metropolitan Cities ...... 97 6.12 Distribution of International Competition Power of Commodities ...... 99 6.13 Distribution of Total Official Poverty Population ...... 100 6.14 Distribution of Protected at HomePopulation .... 102 6.15 Distribution of Self-supporting Low Income Population ...... 103 6.16 Correlation Coefficients ...... 105 6.17 Regression of Total Official Poverty Population ...... Ill 6.18 Regression of Protected at Home Population .... 114 6.19 Regression of Self-supporting Low Income Population ...... 116 6.20 Blocked Regression of Total Official Poverty Population ...... 119 6.21 Blocked Regression of Protected at Home Population...... 122 6.22 Blocked Regression of Self-supporting Low Income Population ...... 124 6.23 Regression of Total Official Poverty Population by Region ...... 127 6.24 Regression of Protected at Home Population by Region ...... 130 6.25 Regression of Self-supporting Low Income Population by Region ...... 132 7.1 Summary of Multiple Regression (National Pattern) ...... 139 7.2 Regression of Total Official Poverty Population by Region (Summary) ...... 144

x 7.3 Regression of Protected at Home Population by Region (Summary) ...... 145 7.4 Regression of Self-supporting Low Income Population by Region (Summary) ...... 146

xi LIST OF FIGURES

FIGURE PAGE 4.1. The Research Model of Rural Poverty in the Republic of Korea...... 52 CHAPTER I INTRODUCTION

Statement of the Problem

Since the 1960s, the Republic of Korea has recorded high rates of economic growth. Today, the Republic of Korea is considered to be a success story in economic development. But this "success" is misleading. Korean development has been extremely unbalanced among different sectors and re­ gions of the country. For example, during 1965-88, the average annual economic growth rate of the agricultural sector was 3.2 percent, while that of the industrial sector was 15.1 percent (World Bank, 1990:181). In 1989, average real income of farm households was 74.5 percent of urban laborers' income (Chang, 1991:57). Even though poverty has been alleviated substantially, the current poverty level in Korea remains high. About 2.2 million persons (5.2 % of the total population) had incomes below officially defined minimum standard of living costs in 19901) (Ministry of Health and Social Affairs,1990).

1 The purpose of this dissertation is to examine the structural correlates of rural poverty in the Republic of Korea. The problem of poverty in Korea has been studied by many social scientists and governmental research institutes (Chung, 1985; Hong, 1986; Ju et al., 1989; Kim, 1985; Lee and Ho, 1988; Lim, 1979; Lim et al., 1989; Suh et al., 1981). But these studies focused almost exclusively on urban areas. Rural Korea long has had a disproportionate share of the nation's poverty. Currently rural areas com­ pose 25.6 percent of the nation's population but 74.5 per­ cent of all people living below the official poverty line (Korean Statistical Association, 1990; Ministry of Health and Social Affairs, 1990). Unfortunately, rural poverty has been the direct focus of only a small number of studies by social scientists (Chung et al., 1989; Chung and Oh, 1992; Kim, 1980; Wang, 1989). Most previous studies of Korean rural poverty (includ­ ing general poverty) have various limitations. First, the vast majority of research has focused on the characteristics of individuals and has been driven largely by a social psychological research framework. That is, most previous studies were individual-level based approaches emphasizing the relationship between poverty, modernity and human capi­ tal. As a result, these studies did not adequately address structural aspects of poverty. Second, most previous studies tended to be descriptive. Enumeration of frequencies and percentages were used as the analytical methods. Multi-variate analysis such as multiple regression was rarely applied. In addition, there were very few exploratory studies to test research hypotheses for the development of poverty theories. Third, the data collection method of most previous studies was the survey approach, such as personal interview. As a result, qualitative research, such as participant observation, was very rare. Also, most previous studies did not adequately investigate the historical trends of poverty. Finally, most previous research did not consider the importance of regional effects in understanding poverty. One study (Lim et al., 1989) considered the regional dimen­ sion of poverty. However, the analytical units of the study were cities, not rural areas. Viewed from an international political economy perspective, such as dependency/world system theory, the pattern of rural poverty in Korea has been in large part the result of its relationship to first world nations (mainly Japan and United States). Korea's integration into the modern world began with the Kanghwa treaty with Japan in 1876. In 1910, Korea was annexed to Japan. During the colonial administration period (1910-45), Japan exploited raw materials and food stuffs from Korea. Japan used Korea as a market to sell Japanese manufactured goods and a military base for continental invasion. Most industrial facilities were established in the north part of Korea (mainly cities) for convenient continental invasion. As a result, rural areas were ignored in the process of national development (Eckert et al., 1990; Ku, 1985; Hoare and Pares, 1988). In 1945, Korea was liberated from Japanese colonial occupation. The end of the Second World War saw the country divided into the South and North. Most heavy industrial facilities and natural resources were located in the North. As a result, the South was reduced to a poor agrarian coun­ try with some light industries. The division of Korea into two competing ideological camps precipitated the Korean war in 1950 (Eckert et al., 1990). Since 1945, the Republic of Korea (South Korea) has been dependent on the United States. Korea has been cited as a good example of dependent development (Lim, 1987). However, the relationship with United states has been main­ tained by political and military interests as well as the logic of an international division of labor. Since 1960s Korea has pursued an economic development policy which emphasizes the increased export of industrial goods. However, Korea's economic growth has been uneven among its regions. Also Korea has become more dependent on the first world. For example, currently the United States has been pressing Korea to enlarge the export of agricultural commodities. This pressure will have a serious negative effect on the socio-economic conditions of rural Korea because Korean agricultural commodities have little potential to compete in international markets (Ministry of Agriculture, Forestry and Fisheries, 1991a). Unlike many Latin American countries, the Korean Government has adopted more stringent interventionist policies in the process of national development (Burmeister, 1990; Midgley, 1988). Thus, viewing the situation from a national context, rural poverty in Korea has been created and maintained due to the uneven national development poli­ cies of the Korean Government. Historically Korean rural development policies have been mainly top-down approaches grounded in the modernization perspective. Until now, Korean rural development policies have not adequately con­ sidered regional characteristics, the importance of partici­ patory democracy, and the basic needs of the rural popula­ tion (Hong, 1988). Also, due to large-scale farming orient­ ed agricultural policies, middle and smaller sized farming have been ignored. As a result, the current level of rural poverty in Korea remains high. Dependency/world system theory is useful for explaining differences between the first world and the third world in terms of economic well-being. However, dependency/world system theory has limitations in its applicability to ex­ plain uneven development within one national context.2) Dependency/world system theory offers better a explanation of Korea’s relationships with Japan and United States. The theory also deals with why rural areas of Korea are more impoverished than urban areas, but does not really deal with poverty variations within rural areas. That is, one of the most serious problems in the dependency/world system ap­ proach is the level of analysis. Therefore, what is needed is a perspective that looks at poverty variations based on smaller spatial scales such as within rural regions. A number of perspectives have addressed this issue, but almost always in the first world context— that is, the United States context (Lobao and Schulman, 1991; Romo et al., 1988) .

Purposes of Study

Although poverty is a very broad concept, the conceptu­ alization and measurement of poverty has focused on the economic conditions of individuals and aggregate levels (e.g. community, county, region, and state). Many previous studies of poverty confined their attention to identifica­ tion of those individuals, families, or households with income below the poverty line (Chung and Oh, 1992; Beeghley, 1984; 1981; Suh et al., 1981). Howes and Markusen (1981:438) defined poverty as "the household's inability to secure a livelihood, that is, to feed, clothe, and shelter itself at a basic standard of living." In this study, rural poverty is defined as a social position in which household income is insufficient to meet some basic needs in rural communities (Tomaskovic-Devey, 1987 and 1988a). That is, rural poverty refers to the proportion of the population with household income below the official poverty line in a county. The major purposes of this study are as follows: (1) to describe the historical trends of agriculture and rural poverty related indicators in Korea. This historical analysis will be used to identify the changing process of socio-economic conditions in rural Korea; (2) to investigate structural correlates of rural poverty in Korea. Structural level-based perspectives will be emphasized in this research because most previous studies of Korean rural poverty did not adequately investigate various structural aspects of rural poverty; (3) to test the relevance of the research model of rural poverty across different regional settings. The regional comparisons may uncover different effects of specific regional characteristics on rural poverty; and (4) to suggest some policy alternatives to help reduce and alleviate rural poverty in Korea. 8 Significance of Study

Increasing unevenness among regions and economic sectors may bring about political instability and have serious negative effects on the economic development of Korea. Recent changes in the economic structure of rural Korea have served to bring attention to rural poverty. The relatively higher level of poverty in the agricultural sector and rural areas is likely to become a heavy burden for the Korean Government. As mentioned earlier, most previous studies of Korean rural poverty did not adequately address structural aspects of poverty. Therefore, the structural focus of this study may have some significance for research, policy, and theory. First, the structural focus is significant in poverty research because structural factors may be more fundamental causes of rural poverty in developing countries such as Korea. This study is a revised application of the Goldschmidt hypothesis and related research hypotheses to rural Korea. Therefore, this study may contribute to agri­ cultural sociological research because it tests the appli­ cability of the Goldschmidt hypothesis in the situation of developing countries. Second, the structural focus of this study may provide policy alternatives to help reduce and alleviate rural poverty in Korea. Also, the structural focus may provide 9 some insights for long-term plans to reduce rural poverty. Finally, the structural focus of this study contributes to sociological development theory. This research takes the study of rural poverty in a developing country (Korea) to a lower level of analysis than dependency/world system theory, presenting a different theory of poverty. The regional focus also may have some significance for research, policy, and theory. First, the regional focus of this research seems to be very timely because recently region has "re-appeared as a potentially important consider­ ation for sociologists " (Rankin and Falk, 1991:224). That is, regional differences have been emphasized in recent studies of poverty and well-being in rural communities (Lobao, 1990; Lobao and Schulman,1991; Markusen, 1987; Rankin and Falk, 1991; Romo et al., 1988). Also, a regional focus may contribute to testing the relevance of the Goldschmidt hypothesis across different regional settings in Korea. Second, the regional focus of this study may be signif­ icant for poverty policy. The regional focus will provide hypothesis testing on structural correlates of poverty within a regional context. Therefore, this study can sug­ gest policy alternatives to reduce rural poverty which are relevant to each specific region. Finally, the regional focus of this study may have significance for the development of sociological theories 10 such as theories of development and under-development be­ cause "region is an useful theoretical concept which needs to be more adequately theorized and incorporated into socio­ logical analyses" (Rankin and Falk, 1991:224). In contrast to U.S. studies of rural poverty, such as Lobao (1990), this study will add a more regional political economy perspective to the understanding of regional/spatial variations.

Overview

This dissertation is divided into seven chapters. In the present chapter, an overview and rationale for the study have been discussed. Chapter II describes the historical trends of agriculture and poverty related indicators in Korea. Chapter III reviews theoretical perspectives of rural poverty. Chapter IV develops the research model and hypotheses emphasizing structural aspects of rural poverty, which are based on the literature review of rural poverty. Chapter V describes the methodology, including data collec­ tion procedures, measurement of variables, and analytical methods. Chapter VI presents the findings of descriptive and multiple regression analysis, and provides interpre­ tations related to the research hypotheses. Finally, Chapter VII presents the summary and conclusions of the re­ search and recommendations for future research and suggests policy alternatives to help alleviate rural poverty. CHAPTER II AGRICULTURE AND RURAL POVERTY RELATED INDICATORS IN THE REPUBLIC OF KOREA: HISTORICAL TRENDS

Introduction

This chapter presents historical trends of agriculture and rural poverty related indicators in the Republic of Korea. The first part of this chapter is a historical analysis of Korean agriculture. The second part of this chapter is an explanation of historical trends of rural poverty related indicators for Korea.

Korean Agriculture

Basic characteristics of Korean agriculture include small-scale farming, family farming, and rice farming (Hong, 1988). The cultivated land is about a quarter of the total national area (99,173 km2). Average farm size per farm has been very small: .90 ha in 1965 and 1.21 ha in 1990 (Ministry of Agriculture, Forestry and Fisheries, 1986 and 1991d; Wang, 1989). The agricultural production in Korea

1 1 12 has depended largely on family labor. That is, the proportion of family labor in the agricultural production has been over 70 percent (72.3 % in 1965, 75.4 % in 1970, and 79.8 % in 1980). Currently the proportion of family labor in Korean agricultural production has increased to 80.8 percent in 1990. This seems to be attributable to a large rural-to-urban migration of farm laborers and the spread of agricultural machinery, such as the power tiller. Rice farming has been a major part of Korean agricultural production. The proportion of rice paddy fields to total cultivated land was 57 percent in 1965, 59.5 percent in 1980, and 66.2 percent in 1990 (Ministry of Agriculture, Forestry and Fisheries, 1991b; Hong, 1988). The current increase in the proportion of rice paddy fields to the total cultivated land seems to be attributable to many land recla­ mation projects and the decrease in upland fields due to non-farming demands. Korean rice farming, which is located at Asian monsoon climate zone, needs concentrated labor at the rice-planting season (May-July) and harvest time (October-November). Remaining time needs relatively low levels of labor demand. Table 2.1 summarizes major Korean agriculture related indicators. The proportion of the rural population to the national population has declined continuously, from 66.4 percent in 1966 to 25.6 percent in 1990. Since the 1960s, the Korean Government has focused on export-oriented 13 Table 2.1 Major Korean Agriculture Related Indicators (unit= %) Indicators 1965 1970 1975 1980 1985 1990 Proportion of Rural Population 66.4* 58.8 51.6 42.7 34.6 25. 6 Proportion of Farm Population 55.1* 44.7 37.5 28.4 20.8 15.6 Share of Agriculture, Forestry and Fisheries in GNP 37.6 26.6 24.9 14.9 12.8 9.1 Proportion of Old Farmers# 15.3 20.2 26.3 32.1 39.7 53.1 Rate of Food Self-supply • 80.4 73.0 43.2 48.4 43.0 Proportion of Debts to Farm Income 6.3 3.8 11.8 35.3 42.9

* The Statistics of 1966 # Old Farmers: 50 Years Old and More & The Statistics of 1989 Source: Ministry of Agricultureforestry and Fisheries, Annual Statistics of Agriculture. Forestry and Fisheries. 199Id and various years; Ministry of Agriculture, Forestry and Fisheries, Agricultural Census. 1991b and various years; Economic Planning Board, Annual Statistics in Korea. 1991 various years. 14 industrialization. As a result, Korean agriculture and rural areas have been neglected. The Korean Government has treated the agricultural sector and rural areas as fields to provide cheap labor and food for the industrial labor force. One major consequence of this was a rapid rural-to-urban migration. During last 25 years, the settlement pattern in Korea has reversed completely, from a rural dominated soci­ ety to an urban dominated society. Today, three-fourths of the national population lives in cities. The proportion of the farm population to the national population was 55.1 percent in 1965. The proportion de­ clined to 44.7 percent in 1970, 28.4 percent in 1980, and 15.6 percent in 1990. Similarly, the share of agriculture, forestry and fisheries in GNP declined from 37.6 percent in 1965 to 9.1 percent in 1990 (Ministry of Agriculture, Forestry and Fisheries, 1991c; Hong, 1988). The proportion of old farmers (50 years old and more) among the working farm population (14 years old and more) has increased from 15.3 percent in 1965 to 53.1 percent in 1989. This seems to be attributable to the heavy rural-to-urban migration of younger persons. It was estimated that in 1979, Korea was 80.4 percent self-sufficient with respect to food. By 1990, this had declined to 43 percent. In 1990, the import liberation rate (not subject to import taxes and tariffs) of farm products in Korea was 82.8 percent (Ministry of Agriculture, Forestry 15 and Fisheries, 199Id). That is, Korean agriculture has become more dependent on other countries, such as the U.S. The proportion of debts to farm income has increased from 6.3 percent to 42.9 percent in 1990. The economic condition of farm households has deteriorated continuously. Table 2.2 shows the comparison of households' real income between farmers and urban laborers. Historically, the real income of farmers has been lower than that of urban laborers. The gap of real income between farmers and urban laborers has increased from 14.2 percent in 1970 to 27.6 percent in 1985. Table 2.3 represents the numbers of farm households by size of cropland. The proportion of smaller farm households (less than 0.5 ha in the size of cropland) decreased from 42.9 percent in 1960 to 28.8 percent in 1980. In contrast, the proportion of middle farm households (0.5-2.0 ha in the size of cropland) increased from 50.8 percent in 1960 to 64.8 percent in 1980. The proportion of larger farm house­ holds (more than 2.0 ha in the size of cropland) has stayed at about 6 percent of total farm households. The trend in the composition of farm households since 1980 represents a bi-polar pattern. That is, the proportion of smaller farms and larger farms are increasing, while the percentage of middle-sized farms are decreasing. This pattern is similar with U. S. agriculture (Rogers et al., 1988). This bi-polar pattern seems to be attributable 16 Table 2.2 Comparision of Households' Real Income Between Farmers and Urban Laborers (unit=l,000 Won) Year Farmer (A) Urban Labors (B) A/B (%)

1970 1,439 1,678 85.8

1980 2,319 3,118 74.4

1985 3,040 4,199 72.4

* Standard Year= 1980 Source: Economic Planning Board, Annual Report on Urban Laborers * Familv Budget. 1991 and various years; Ministry of Agruculture, Forestry & Fisheries, Annual Report on the Farm Household Economy Survey. 1991c and various years. 17 Table 2.3 Numbers of Farm Households by Size of Cropland (unit=l,000) Size of Year Cropland 1960(%) 1970 (%) 1980(%) 1990(%)

Less than 1,009(42.9) 860(34.6) 612(28.8) 557(31.5) 0.5 ha

0.5-2.0 ha 1,193(50.8) 1,463(58.9) 1,377(64.8) 1,087(61.5)

More than 148( 6.3) 160( 6.5) 139( 6.5) 173( 9.8) 2.0 ha

Source: Ministry of Agriculture, Forestry & Fisheries, Arqicultural Census. 1980 and 1990; Ministry of Agriculture, Forestry & Fisheries, Annual Report on the Farm Household Economy Survey. 1991c and various years. 18 to a large rural-to-urban migration of smaller farmers relative to middle-sized farmers, and a relaxation of the regulation limiting the size of landholding. Recently, the Korean Government set free the regulation of an upper limit in landholding (former upper limit was 3.0 ha). Table 2.4 shows trends of land ownership. In 1945, the proportion of owner (full-owner), part-owner, and tenant were 14.1 percent, 40.1 percent, and 45.8 percent, respec­ tively. But these patterns were changed by two land re­ forms. The first land reform was implemented in 1948 under the U.S. Military Government. Nearly 200,000 ha of farm­ land, which was seized from the colonial Japanese, was sold to 505,000 tenant farmers (sharecroppers) and small farmers. The land price was three times the value of a year's product payable in installments over a 15 years period (Chung and Oh, 1992). The second land reform was implemented by the Korean Government in 1950. The government bought 601,000 ha of farmland from landlords by issuing land securities guaran­ teeing 5 years installment at the price of 150 percent of a year's product from the land. The land was sold to 1,166,000 farmers under the same condition (Chang, 1988; Chung and Oh, 1992; Hong, 1988). As a result, the land dis­ tribution pattern was changed greatly. However, since 1950 the proportion of owners has decreased continuously, while the proportion of part-owners has increased. That is, the 19 Table 2.4 Trends of Land Ownership

Year No. of Farm Owner Part-owner Tenant % (1,000) % % %

1945 100.0(2,019) 14.1 40.1 45.8

1948 (U.S. Military Government)

1950 REFORM...... (April) 1950 100.0(1,970) 92.0 8.0 0 (December) 1959 100.0(2,267) 80.1 18.0 1.9

1974 100.0(2,269) 68.4 27.8 3 . 8

1983 100.0(1,948) 40.2 56.9 2.9

1985 100.0(1,880) 35.3 62.6 2.1

Source: Ministry of Agriculture, Forestry & Fisheries, Agricultural Census. 1991b; Hong, Understanding of Rural Sociology. 1988, P 253. 20 proportion of part-owners has increased from 8 percent in 1950 to 62.6 percent in 1985. The proportion of borrowed land to total cultivated land has increased from 16.4 percent in 1965 to 36.5 percent in 1990 (Ministry of Agriculture, Forestry and Fisheries, 1991d; Hong, 1988). Table 2.5 represents the composition of farm household income. In 1965, farm income composed 78.9 percent of total farm household income. The proportion of farm income to total farm household income has decreased continuously. As a result, in 1990, the proportion of farm income to total farm households' income was 56.8 percent. That is, non-farm income became a very important source of farm household income. Table 2.6 shows the proportion of farm income to living expenses by farm size. In 1965, farm households with 1.5 and more ha in farm size had farm incomes above living ex­ penses. In 1970s, the proportion of farm income to living expenses was improved greatly. This seems to be attribut­ able to the introduction of double grain price system and continuously good harvests during the 1970s. However, since 1980 the situation has grown from bad to worse. As a result, by 1990, only farm households with "more than 2.0 ha" in farm size had farm incomes above living expenses. This trends seems to be related to the process of deteriora­ tion in farm household economy and the enlargement of par­ ticipation in non-farm jobs. 21 Table 2.5 Composition of Farm Household Income (unit=%) Year Farm Income Non-Farm Income

1965 78.9 21.1

1970 75.9 24.1

1980 65.2 34.8

1990 56.8 43.2

Source: Ministray of Agriculture, Frestry & Fisheries, Annual Report on the Farm Household Economy Survey. 1991c and various years. 22 Table 2.6 The Proportion of Farm Income to Living Expenses by Farm Size (unit=%) Farm Size 1965 1970 1975 1980 1985 1990

Less Than 0.5 ha 58.6 54.3 71.3 39.6 35.6 33.5

0.5 - 1.0 ha 83.8 93.9 107.8 75.0 59.8 56.3

1.0 - 1.5 ha 96.8 103.6 124.0 89.9 82.9 84.3

1.5 - 2.0 ha 103.3 119.3 138.4 101.5 103.1 96.5

More Than 2.0 ha 112 .5 112.4 149.3 124.4 115.2 105.5

Source: Ministry of Agriculture, Forestry & Fisheries, Annual Report on the Farm Household Economy Survey. 1991c and various years. 23 Historical Trends in Rural Poverty Related Indicators

In general, rural communities have been variously defined by population size and density, occupation, and socio-cultural characteristics (Flora et al., 1992; Hong, 1988). In Korea, cities refer to communities of 50,000 or more inhabitants, while rural areas refer to all counties (137) which have no cities. Therefore, the concept of rural areas in Korea is similar with the concept of non-metropoli­ tan counties in U.S. (Flora et al., 1992; Hong, 1988). One way to identify historical trends of rural poverty in Korea is to review previous research. For example, Suh et al. (1981) conducted research on the state of poverty and some alleviation measures. They used a rather complicated method in estimating the poverty incidence. The so-called "Hypothetical Market Basket Approach" has been applied to the problem of estimating minimum productive survival costs. But Suh et al. (1981) modified the hypothetical market basket approach by estimating the minimum level of food demand which may be relatively feasible for measuring objec­ tively. Other basic demands were estimated indirectly by means of the Engel Coefficient (proportion of food cost to total income). Employing this approach, they estimated the absolute rural poverty rate as follows: 35.8 % in 1965, 11.7 % in 1976, and 9.0 % in 1980. Chung and Oh (1992) estimated that in 1988 the absolute poverty line for a farm household of five members was 3,779,568 won ($ 5,525) a year. Rural poverty in terms of absolute deprivation was defined as "a condition under which the income of rural farm household does not meet the expen­ diture of the household necessary to subsist with a minimum level of health and decency" (Chung and Oh, 1992:95). The time series poverty line was calculated by applying the purchasing price index and income deflator of farm house­ holds. Chung and Oh (1992) argued that the estimated "Poverty Head-Count Ratio" of the farm sector in 1988 was 6.5 percent, compared with 33.7 percent in 1967. Relative rural poverty was defined as a condition "that the rural farm household income is lower than one half of the average household income" (Chung and Oh, 1992:95). They estimated that the poverty line of relative deprivation for a farm household of five members was 5,351,500 won ($ 7,824) a year. They estimated that the "Relative Poverty Head-Count Ratio" in the farm sector was 17.9 percent in 1988, compared with 31.6 percent in 1967. Table 2.7 summarizes historical trends of official poverty population in Korea. During the 1970s, the percent official poverty population decreased greatly, but since 1980 the proportion of official poverty population has increased again. Two sub-dimensions of official poverty population are identified: "Protected at Home Population" 25 Table 2.7 Historical Trends of Official Poverty Population

Total Official Protected at Self-supporting Year Poverty Pop. Home Low Income

Freq. (%) Freq. (%) Freq. (%)

1970 2,422,748 (7.51) 306,250 (0.95) 2,116,498 (6.56)

1975 1,279,222 (3.62) 375,350 (1.06) 903,872 (2.56)

1980 1,782,056 (4.67) 282,000 (0.74) 1,500,056 (3.93)

1985 2,210,000 (5.42) 282,000 (0.69) 1,928,000 (4.73)

1990 2,174,000 (5.08) 339,000 (0.79) 1,835,000 (4.29)

* Percent is the proportion of official poverty population to total national population. Source: Ministry of Health and Social Affairs, Annual Statistics Report of Health and Social Affairs. 1991d and various years. 26 and "Self-supporting Low Income Population" (refer to Chapter V for the definition and measurement of these). Table 2.8 represents official urban and rural poverty population in 1991. There is a large difference between the rural and urban areas. About two-thirds of the poverty population live in rural areas. In rural areas, the proportion of "Protected at Home Persons" is much higher (68.11 %). Over-all, rural Korea has a very disproportion­ ate share of the nation's poverty population. Table 2.9 represents trends of inequality in income distribution. The Gini Coefficient was used to measure the inequality of income distribution. Gini Coefficient values range from 0 to 1 (the higher the value, the greater the degree of inequality). From 1965 to 1980, the inequality of income distribution in rural Korea worsened. However, it improved in the 1980s. This tendency was very similar with those of national and urban income distribution. In gener­ al, the income distribution of rural areas has been more equitable than that of urban areas. This seems to be partly attributable to the heavy out-migration of small farmers and two land reforms in 1948 and 1950 (Chang,1988; Hong, 1988). In sum, basic characteristics of Korean agriculture include small-scale farming, family farming, and rice farm­ ing. Korean agriculture and rural areas have been neglected in the national development process. Also, Korean agricul­ ture has become more dependent on other countries. 27 Table 2.8 Official Rural and Urban Poverty Population (1991)

Total Protected at Home Self-supporting Region Population low Income Po. Freq (%) Freq (%) Freq (%)

Urban 719,835 105,543 614,292 (35.28) (31.89) (35.93) Rural 1,320,752 225,366 1,095,386 (64.72) (68.11) (64.07) National 2,040,587 330,909 1,709,678 (100.00) (100.00) (100.00)

Source: Ministry of Health and Social Affairs, Current Conditions of 1991*s Livelihood Protection Population. 1991. 28 Table 2.9 Trends of Inequality in Income Distribution (Gini Coefficient)

Region 1965 1970 1980 1985 1988

Urban .3428 .3550 .3891 .3631 .3355

Rural .2985 .2995 .3555 .3199 .2895

National .3428 .3550 .3891 .3631 .3355

Source: Korean Statistical Association, Social Indicators. 1991 and various years. 29 Socio-economic conditions of farm households have deteri­ orated continuously. Non-farm income has now become a very important source of farm household income. Rural Korea has a very disproportionate share of the nation's poverty popu­ lation. CHAPTER III THEORETICAL EXPLANATION OF RURAL POVERTY

Introduction

Analysts have classified perspectives on poverty into two general schools: individual level-based perspec­ tives and structural level-based perspectives (Schiller, 1989; Tomaskovic-Devey, 1988b). Individual level-based perspectives locate the causes of poverty in the personal and socio-cultural characteristics of the poor, while struc­ tural level-based perspectives locate the causes of poverty in social structure. Individual level-based perspectives emphasize the improvement of schooling, training, and skills as the solution to poverty. Structural level-based perspec­ tives focus on changing the structural conditions of in­ equality within a society or between societies by advocating policies that change the poor's access to resources and job opportunities. In the first part of this chapter, individual level- based perspectives will be reviewed rather briefly because the focus of this research is on the structural correlates

30 31 of rural poverty in Korea. In the second part, structural level-based perspectives will be reviewed in detail.

Individual Level-based Perspectives

While there have been a number of individual level- based perspectives of poverty, these perspectives can be divided into two general types: human capital theory (as a part of neo-classical economic perspective) and various functionalist perspectives (i.e. sociological counterparts to the economic perspective).

Human Capital Theory

Human capital theory is a part of neo-classical econom­ ic theory. Human capital theory argues that poverty is due to lack of human capital. The major concern of human capi­ tal theory is "activities that influence future monetary and psychic income through investment in human capital" (Becker, 1964:1). Forms of investments in human capital include schooling, on-the-job training, medical care, migration, and seeking information about prices and incomes. Human capital theory argues that more highly educated and skilled persons almost always tend to earn more than others (Beker, 1964). The theory argues that poverty is due to lack of human capital. Therefore, the solution to poverty would be in the 32 direction of greater investment in human capital (Becker, 1964). The main assumptions of human capital theory, such as free and informed choice about job informa­ tion, and workers as rational economic actors, have been criticized as unrealistic by many social scientists (Rosenbaum et al., 1990; Tomaskovic-Devey, 1988). Human capital theory "depended too much on the idea of a personal decision being taken by the (potential) employee" and "made no distinction between the various groups in the labor market" (Valkenburg and Vissers, 1980:156).

Functionalist Perspectives

Functionalism has been developed by Parsons, Merton, and other sociologists as a comprehensive theory of society. In this sense, a functionalist perspective may be considered as a structural level-based perspective. However, most empirical research based on a functionalist perspective including derivative theories of poverty, are predominantly at the individual level. Hence, in this study, the func­ tionalist perspective is discussed as an individual level- based perspective. A functionalist perspective argues that inequalities in power, wealth (income), or prestige are related to the functional needs of a society (Davis and Moore, 1945). Two functionalist based theories focused on poverty and 33 inequality in society are the culture of poverty and status attainment theories. According to culture of poverty theory, poor people may have deviant lifestyles or value orientations to work and society (Lewis, 1968). This theory tries to understand poverty as a subculture with its own structure and ratio­ nale. This theory assumes that people are poor because of their cultural traits (Lewis, 1968). According to this theory, the poor are not prepared to take advantage of education, jobs, or other opportunities because they do not share middle-class work and family-oriented values of soci­ ety. The solution to poverty includes attempts to change the values of the poor. Culture of poverty theory emphasiz­ es the correction of inadequate ways of bringing up children through psychological treatment, education, and counseling for parents, as the means to wipe out the vicious circle of poverty (Banfield, 1970; Lewis, 1968). The culture of poverty has been criticized as follows: (1) individualistic bias for legitimating poverty; (2) ethnocentrism; (3) the exaggeration of social pathology; (4) sexism; (5) inability to offer policy alter­ natives and (6) failure to deal with the conditions of the culture of poverty such as unemployment (Goodwin, 1972; Holman, 1978; Marklund, 1990; Rabow et al., 1983; Tomaskovic-Devey, 1988b). 34 Another functionalist perspective-based theory on inequality is status attainment theory. Status attainment theory "rests on a functionalist conception of social struc­ ture in which social positions are conceived of as levels of performance" (Horan, 1978: 534). Like human capital theory, status attainment theory assumes fully open and competitive allocation of individuals to jobs (Goldman and Tickamyer, 1984). That is, status attainment theory is the sociolog­ ical counterpart of human capital theory (Knottnerus, 1987). The basic question of status attainment theory is how the status individuals achieve in their careers is affected by the statuses ascribed to them earlier in life (Blau and Duncan, 1967). According to Blau and Duncan (1967), status attainment theory investigates the process of occupational mobility by tracing the inter-dependence among four determi­ nants of occupational achievement, two of which are social background (father's education and father's occupation), and two of which are training and early experience that prepare the individual for a subsequent career (education and first job). By analyzing patterns of occupational movements, the conditions that affect them, and some of their consequences, status attainment theorists attempt to explain the dynamics of the stratification system. Status attainment theory has been frequently used in studying occupational mobility and stratification (Goldman and Tickmyer, 1984; Krymkowski, 1991; Wegener, 1991). However, status attainment theory fails "to address class- imposed limits to mobility and achievement or uneven development of segmented labor markets. It offers neither an account of the historically specific, class-mediated origins of job prerequisites nor a theory of production relations, social reproduction, or mediating political relations" (Goldman and Tickamyer, 1984:197). In a study on status attainment of Costa Rican rural high school boys, Hansen and Haller (1973:279) suggest that "there may be a number of differences in attainment processes among cultures and between developed and developing countries." Thus Hansen and Haller (1973) argue that it is important to incorporate other variables, such as physical access to educational institutions and access to nonfamilial sources of financial support, into the status attainment model. Knottnerus (1987:119) argues that the status attainment theory's image of society, "which emphasizes a growing middle mass of workers enjoying a high standard of living" is different from the reality (shrinking middle class). In sum, human capital theory and functionalist perspec­ tives of poverty have many similarities. Both perspectives argue that individual attributes and qualifications are the cause of poverty (Schiller,1989). Also, both perspectives assume that human beings are rational actors within the marketplace and that workers have free and informed choice about job information (Rosenbaum et al., 1990). Both 36 perspectives assume that inequality is regarded not only as necessary but also fundamentally just (Lobao, 1990). Like human capital theory, functionalist perspectives emphasize personal investment in education and training as the solu­ tion to poverty. The ideological stance of human capital theory and functionalist perspectives is to justify and keep the privileged positions of affluent. That is, both per­ spectives operate in support of the status quo and dominant elites of society (Huaco, 1986; Lobao,1990).

Structural Level-based Perspectives

Structural level-based perspectives focus "on the context in which inequality is created, that is, on the social order that constrains individuals' choices and generates the positions individuals occupy" (Lobao, 1990:7). Structural level-based perspectives of poverty can be divid­ ed into Marxian perspectives and Non-Marxian perspectives.

Marxian Perspectives

Considering the political/economical context of Korea, Marxian class theory and dependency/world system theory may be two good examples of Marxian perspectives of poverty. Marxian Class Theory: Marx believed that societies follow historically determined economic forces. He argues 37 that history is a series of inevitable conflicts between economic classes. Private property and capitalism are considered as the causes of poverty and alienation in soci­ ety. Marx argues that successive systems for the production and distribution of wealth primarily condition accompanying social and cultural institutions (Marx, 1867/1967; Marx and Engles, 1848/1948; Mayer, 1991). According to Wood (1987:7), the struggle between capi­ tal and labor "over the rate of surplus value is not only the basis of class conflict in capitalist society but also the major determinant of the amount of profit, the rate of capital accumulation, and general living standards." The Marxists argue that if capitalism was eliminated, the pover­ ty problem would go as well. Marx argued that poverty and socio-economic conditions were neither necessary nor justi­ fiable (Marx, 1867/1967). In Marx's view, class is deter­ mined by one's relationship to the means of production. Class exists basically in two major forms: the bourgeoisie who own the means of production and the proletariat who must support themselves by selling their labor to those who own the means of production. The bourgeoisie seek to maximize their profit by minimizing the amount of return to the proletariat. According to Marx, the struggle between the proletariat and capitalists lead to the overthrow of the capitalists. Capitalists will continue to be weakened by the increased 38 concentration of their power while the proletariat will become stronger through economic solidarity and party organization. Marx argued that the stratification of capi­ talists and workers was neither necessary nor justifiable (Levine, 1985; Marx and Engels, 1848/1948 and 1867/1967; Mayer, 1991; Ritzer, 1988; Turner, 1982). According to Marxists, "the irrationalities of capitalism (insecurity, poverty, waste, etc.) will be eliminated through rational planning of the use of productive resources. A new form of personal freedom will be created based on the elimination of competitive individualism and the traditional division of labor" (Mayer, 1991:98). Marxian class theory provides many insights into the nature of class conflict and the persistence of poverty in capitalist societies. Marx's view of social classes and the importance of economic factors is a major contribution to social sciences. However, Marxian class theory has been criticized on various grounds: (1) economic determinism (Ritzer, 1988); (2) ethnocentrism, that is a model based on Western societies (Wiarda, 1988); (3) the neglect of inter­ mediate class groupings (Lobao, 1990); (4) zero-sum game thinking (Arndt, 1989); and (5) the neglect of race, gender, and age relationships (Lobao, 1990). Dependency/world system theory: Dependency/world system theory has presented many variations depending on the emphasis of different theorists (Chirot and Hall, 1982; Long, 1980; Staniland, 1985). For example, Frank (1969) argues that the under-development of the third world was caused by the globalization of capitalism. Frank explains a global capitalist system as a whole chain of metropolis- satellite relationships. Frank argues that "all parts of the economies of underdeveloped nations should be seen to be within the web of capitalism" (Ruccio and Simon, 1988:125). According to Frank, "the metropolis expropriates economic surplus from its satellites and appropriates it for its own economic development. The satellites remain under-developed for lack of access to their own surplus and as a consequence of the same polarization and exploitative contradictions which the metropolis introduces and maintains in the satell­ ite's domestic economic structure" (Arndt, 1989:127). Also, he argues that satellites can develop when their ties with the metropolis are weakest and that today the most under­ developed regions are those which had the closest ties to the metropolis in the past (Frank, 1969). Dependent development theory has some distinctive characteristics within the dependency school. According to Staniland (1985:132-140), Cardoso argued that the forms of dependency can be changed and that scholars should inves­ tigate specific characteristics of dependent societies. Cardoso believed that earlier dependency theory was sim­ plistic and over-mechanical. Cardoso tried to "discover how internal and external processes of political domination 40 relate to one another and external factors are interwoven with internal ones, as well as to determine the links between social groups that in their behavior actually tie together the economic and political spheres" (Staniland, 1985:133). Cardoso argues that the development of the satellite (the periphery) is possible in the dependent context although there are structural limitations on such a style of development (Staniland, 1985). Evans (1979) suggested that the expansion of multina­ tional corporations brought about a new form of dependency: dependent development. Evans argued that dependent develop­ ment is characterized by a triple alliance of foreign capi­ tal, the domestic state, and local capital. Although this triple alliance may enhance the development of a modern sector in some under-developed areas, the growth will be uneven. As a result, the level of inequality will be in­ creased throughout society (Evans, 1979; Staniland, 1985). World system theory is a variant of dependency theory (Staniland, 1985). World system theory argues that "societal changes can be understood as a global process" (Peacock et al., 1988:839). Wallerstein argues that "it is necessary to think of world history as the history of the transformation of political empires into a world economy" (Janowitz, 1977:1091-1092). He argues that "a world empire is based on political and military domina­ tion, whereas a capital world-economy relies on economic 41 domination" (Ritzer, 1988:281). He tried to ascertain the conditions under which a political empire was transformed into a world economy. He argued that a capitalist world- economy will be more stable than a world empire. Wallerstein divided the world economy based upon a geographically differentiated division of labor: core, pe­ riphery, and semi-periphery. The core dominates the world economy and exploits the rest of the world system. The periphery refers to "those that provide raw materials to the core and are heavily exploited by it" (Ritzer, 1988:281). The semi-periphery refers to those that stand between the core and the periphery. Wallerstein's approach was both holistic and economistic. He treated "domestic political processes as determined externally or as reflections of domestic interests" (Staniland, 1985:145). World system theory predicts that the economic divergence between the core, semi-periphery, and periphery will be increased while economic convergence will be increased within each zone (Peacock et al., 1988). Although there are some variations within dependency/ world system school as presented in the earlier part, its members share general assumptions. First of all, they argue that under-deveiopment and poverty of the third world results from past and present exploitation by the first world's dominant economic system. Second, they argue that the world capitalist system is a direct cause of 42 under-development and persistent poverty in the third world. Finally, dependency/world system theorists argue that a radical political movement toward socialism can interrupt the vicious cycle of under-development in the third world (Arndt, 1989; Bergesen, 1990; Chirot and Hall, 1982; Howe and Sica, 1980; Long, 1980; Staniland, 1985). The major criticisms of dependency/world system are as follows: (1) the lack of empirical grounding and insuffi­ cient definitions of major concepts such as capitalism and dependency (Brenner, 1977; Staniland, 1985); (2) the lack of consideration of cultural diversities (Chirot and Hall, 1982; Worsley, 1990); (3) its failure to address the pecura- rities of local contexts (Staniland, 1985); (4) the exagger­ ation of the power of international economic systems and an unwillingness to consider dynamics within a country, such as the role of the state, social class system and internal conditions of inequality, and geographic/regional character­ istics (Foster-Cater, 1985; London and Smith, 1988; Staniland, 1985); (5) over-mechanical and economistic logic (Brenner, 1977; Staniland, 1985); and (6) little attention given to the position of the Eastern bloc socialism states in the world economy (Chirot and Hall, 1982). Marxian class theory and dependency/world system theory have some similarities. First, both theories focus on the social justice of income distribution. Second, both theo­ ries emphasize liberation from oppressive and exploitative 43 relationships. Third, both theories argue that control of the economic surplus of a society is the most important factor in the process of development (Wilber and Jameson, 1988; Staniland, 1985). Finally, both theories concentrate on under-development. As a result, they have not focused on the process of development. Neither theory worried "very much about precise forms that post-revolutionary development might or should take" (Arndt, 1989:131). However, Marxian class theory and dependency/world system theory have some differences. First, the major difference between Marxian class theory and dependency/world system theory is the locus of power and control of develop­ ment which these theories identify (Wilber and Jameson, 1988). That is, Marxian class theorists emphasize the internal relationship to means of production, while depen­ dency/world system theorists emphasize the relationship of uneven exchange among nations. Second, Marxian class theo­ rists "call for social revolution to replace the middle class with control by workers and peasants" (Wilber and Jameson, 1988:21). In contrast, dependency/world system theorists emphasize the elimination of dependency through the establishment of New International Economic Order (Arndt, 1989). 44 Non-Marxian Perspectives

Non-Marxian perspectives focus on the structure of industry and farming. Segmented (dual) Economy Theory and Goldschmidt Hypothesis are two representative Non-Marxian perspectives of rural poverty and well-being. Dual (Segmented) Economy Theory: Dual economy theorists "use the concept of economic sector to emphasize the impact of economic organization on socio-economic processes" (Tolbert et al., 1980:1101). Dual economy theorists view "economic sectors as structural entities which derive from the nature of modern industrial capitalism" (Beck et al., 1978:706). Dual economy theory views the capitalist dynam­ ics of concentration and centralization as the origin of a dual economy (Hodson and Kaufman, 1982). Dual economy theorists argue that there are two distinct privatized economic sectors and their specific firms in the U.S economic structure: core (monopoly and oligopoly) sector and peripheral (competitive) sector (Tolbert et al., 1980). Also, many dual economy theorists recognize the third sector, which is organized by the state (Bloomquist and Summers, 1982). According to dual economy theory, these sectors are different "not only with respect to market structure, but also in their work arrangements, employment patterns, and opportunity structures" (Baron and Bielby, 1984:454-455). In the core sector, firms "are noted for high produc­ tivity, high profits, intensive utilization of capital, high incidence of monopoly elements, and a high degree of union­ ization" (Tolbert et al., 1980:1098). The firms in the core sector "tend to be large and have a major share of produc­ tion in their home industries, elaborate divisions of labor, internal job ladders, and stable employment" (Hodson and Kaufman, 1982:729). Core firms expand through a combination of buyouts, vertical integration of the total production and distribution of a product and expansion/diversification into other product lines (i.e., horizontal integration). Positions in core firms are highly paid and require highly skilled workers (Hodson and Kaufman, 1982; Lobao, 1990). In contrast, firms in the periphery sector "are noted for their small firm size, labor intensity, low profit, low productivity, intensive product market competition, lack of unionization, and low wages" (Tolbert et al., 1980:1098). The periphery sector provides less skilled, lower paid jobs. The state sector provides the function of mediating relationships in the privatized economy. The state sector has "a higher proportion of professionals, technicians, and managers than other two sectors" (Bloomquist and Summers, 1982:336). In general, earnings in the state sector fall between the core and periphery sectors (Lobao, 1990). Dual (segmented) economy theory contributes to explain­ ing the existence of the working poor that persists despite 46 various job training programs. The major contribution of dual (segmented) economy theory "has been to shift the awareness of researchers to the characteristics of jobs and the concomitant realization that some jobs are poverty-level jobs" (Tomaskovic-Devey, 1987:57). That is, dual economy theory can account for why some workers remain poor. Also, dual economy theory can explain the fact that some unions and workers have benefitted from the market power of monop­ oly or oligopoly firms. Major criticisms of dual economy theory are as follows: (1) the neglect of non-monetary aspects of segmentation (Lobao, 1990); (2) the neglect of social processes that lead to the creation of good and bad jobs (Hodson and Kaufman, 1982; Tomaskovic-Devey, 1987); (3) the lack of model speci­ fication and inadequate operationalization of major concepts (Hodson and Kaufman, 1982; Kaufman et al., 1981); and (4) the neglect of interaction among political, cultural, social, and spatial factors (Hirsch, 1980; Hodson and Kaufman, 1982). Goldschmidt Hypothesis: The Goldschmidt hypothesis parallels Non-Marxian dual economy theory, but this time applied to farming. The essence of the Goldschmidt hypoth­ esis is that increased farm scale negatively affects the quality of life in rural communities. In a study of two California communities (Arvin and Dinuba) in 1944, Goldschm­ idt (1978a:281) tried to test the "hypothesis that the 47 institution of small independent farmers is indeed the agent which creates the homogenous community, both socially and economically democratic." Goldschmidt (1978a) argued that the two communities were very similar on many demographic, economic, and ecological factors, except for the farm struc­ ture of their respective hinterlands. Arvin (Kern county) was dominated by large, industrial type farms while Dinuba (Tulare county) primarily consisted of small, family farms. Goldschmidt (1978a) found that Dinuba (small-farm community) had more economic and social vitality than Arvin (large-farm community). Compared with Arvin, Dinuba had a better quali­ ty of life in such areas as per capita business establish­ ments, retail trade, average standard of living, physical facilities, schools, parks, recreation centers, churches, and so on (Goldschmidt, 1978a:282-284). Goldschmidt (1978a) argued that farm size was the most important cause of these differences, therefore as farms become more concentrated, the quality of life in rural communites will decline. In a more recent study, using state level data derived from the 1959 census of agriculture and census of population and the 1964 census of agriculture, Goldschmidt (1978b) found a strong positive correlation (r= .76) between the proportion of large-scale farming and the proportion of farm population in the "lower class." Goldschmidt (1978b) used the concept of large scale-farming developed by Nikolitch. That is, he defined large-scale farming as the farm "which had produced $ 100,000 gross values of food and fibers" (Goldschmidt, 1978b:364). Also, Goldschmidt used a social class typology of rural population from Smith (1969). Smith's measurement of the lower class in the farm sector included "all hired farm laborers, all sharecroppers (except for the 15 percent with the highest incomes) and their proportinate share of unpaid family workers, 20 percent of the personnel of abnormal farms (e.g. institutional), and 20 percent of the personnel (not classified elsewhere) on farms with annual sales under $ 2,500 and their proportionate share of unpaid family workers, all as reported in the 1959 Census of agriculture" (Harris and Gibert, 1982:450). He put forth this finding as evidence in support of his hypothesis that the large-scale farming has a negative effect on the quality of life in rural communities. Goldschmidt hypothesis related studies have greatly contributed to the recent development of the sociology of agriculture. These studies have offered a good opportunity to bring increased attention to the socio-economic condi­ tions of rural communities. However, studies based on the Goldschmidt hypothesis have many weaknesses. Major criticisms of the Goldschmidt hypothesis related studies are as follows: (1) unclear conceptualization and inadequate measurement of major concepts, such as farm structure (Green, 1985; Lobao, 1990); (2) the lack of strict control in choosing research areas (Hayes and Olmsted, 49 1984) ; (3) the need for more longitudinal study (Gilles and Dalecki, 1988; Lobao, 1990); (4) the neglect of the symbiotic relationship between farming and the rural commu­ nity (Swanson, 1990); (5) generalizability of the results across regions (Green, 1985); (6) the lack of broader theo­ retical considerations, such as dependency theory, Marxian class theory (Lobao, 1990); (7) the need for an interdisci­ plinary approach, integrating especially economic and socio­ logical constructs (Flinn and Buttel, 1980; Nuckton et al., 1982); and (8) the lack of accounting for non-farm factors, such as industry and service sector employment (Swanson, 1990). Dual economy theory and the Goldschmidt hypothesis have some similarities. First, the major research areas of both theories are local communities. Second, both theories focus on the economic structure of communities. That is, dual economy theory pays attention to the economic structure of industry, while the Goldschmidt hypothesis centers on the economic structure of farming. Finally, both theories fail to consider the effects of international relationships on rural well-being. That is, the major concern of both theo­ ries is on social structure within a domestic setting. The following are major differences between dual econo­ my theory and Goldschmidt hypothesis. First, dual economy theory tends to focus on the economic aspects of well-being (e.g. income, earning, and unemployment), while the 50 Goldschmidt hypothesis focuses on these as well as the social aspects of well-being (e.g. social participation, quality of community services). Second, dual economy theory focuses on the quality of jobs in a local industry structure and the relationships between employers and laborers, while the Goldschmidt hypothesis focuses on the effects of large- scale farming on the quality of life in rural communities (Lobao, 1990; Goldschmidt 1978a and 1978b). Finally, dual economy theorists have adequately identified the roles of state, while the Goldschmidt hypothesis did not. For exam­ ple, according to the dual economy theory, "the state inter­ venes in the privatized economy by providing money and in kind transfers, by formulating monetary policy, and by providing subsidies, regulations, and labor legislation" (Lobao, 1990:25). The Goldschmidt hypothesis does not consider any of these factors. CHAPTER IV RESEARCH MODEL AND HYPOTHESES

Introduction

This chapter presents the research model and hypotheses of this dissertation. The first part of this chapter pres­ ents the research model based on rural poverty theories and research. The second part of this chapter presents the hypotheses based on the results of previous studies on rural poverty and well-being.

Research Model

A structural perspective-oriented research model of rural poverty in the Republic of Korea is presented in Figure 4.1. This research model was constructed by combin­ ing various structural perspectives on rural poverty. The research model consists of three groups of independent variables: farm structure/characteristics, non-farm charac­ teristics, and spatial/political economical characteristics.

51 Farm Structure/Characteristics Percent Middle Farming Percent Larger Farming Percent Part-time Farming Rate of Tenancy

Rural Poverty Non-Farm Characteristics * Total Official Poverty Percent Non-farm Population Population Percent Female * Protected at Home Percent High School Graduates Population * Self-supporting Low Income Population

Spatial/Political Economical Characteristics Region 1 (YOUNGNAM) Region 2 (HONAM) Proximity to Metropolitan Cities International Competition Power of Commodities

Figure 4.1 The Research Model of Rural Poverty in the Republic of Korea The relationships between farm structure/ characteristics and rural poverty are based on the Goldschmidt hypothesis and related studies. Two farm struc­ ture variables (percent middle farming, and percent larger farming) were included in the model because the farm struc­ ture variables have been investigated as very important factors affecting rural socio-economic conditions (Gilles and Dalecki, 1988; Goldschmidt, 1978a and 1987b; Green, 1985; Harris and Gilbert, 1982; Lobao, 1990; Lobao and Schulman, 1991; Skees and Swanson, 1988). Percent smaller farming will be excluded in the regression model because of an orthogonality problem.3) That is, percent smaller farm­ ing is the excluded reference category for interpreting the effects of farm structure on rural poverty. Percent part- time farming and rate of tenancy were included in the re­ search model because part-time farming and tenant farming are now interpreted as regular features of almost all farm­ ing societies including rural Korea (Fuller, 1976; Hong, 1988) . The relationships between non-farm characteristics and rural poverty are based on the insight of dual economy theory. Non-farm characteristics affecting rural poverty include such variables as percent non-farm population, percent female, and percent high school graduates. Tickamyer and Duncan (1990:80) argued that "the socio-eco­ nomic characteristics of communities have an impact on the 54 economic success of residents regardless of their own socio-economic background." Unemployment rates and union­ ization rates have been considered as good indicators of labor market power (Lobao, 1990; Tomaskovic-Devey, 1987). Unfortunately, the county data on unemployment and unioniza­ tion rates in Korea were not available. So these two indi­ cators were not included in the research model. The relationships between spatial/political economical characteristics and rural poverty are based on the insight of dependency/world system theory and regional political economical studies. Spatial characteristics may be very important in understanding poverty and well-being in rural communities because "spatial characteristics circumscribe the location of economic structure and worker power and reflect historical and ecological conditions" (Lobao, 1990:93). Recently, a number of social theorists have turned to the concept of space and incorporated it in their under­ standing of society (Foucault, 1979; Giddens, 1984; Gottdiener, 1991; Lobao, 1990). Also, several rural sociol­ ogists have used regional characteristics to explain differ­ ences in farm structure (Pfeffer, 1983), uneven development of agriculture (Mann and Dickinson, 1978; Mooney, 1982 and 1987), rural poverty (Lobao and Schulman, 1991), and differences in earnings in the rural population (Rankin and Falk, 1991). Spatial characteristics that affect rural poverty include region 1 (YOUNGNAM), region 2 (HONAM),and proximity to metropolitan cities. In this study, region is divided into three categories: YOUNGNAM (43 counties located in South-East districts), HONAM (34 counties located in South-West districts), and Others (58 counties located in North districts and 2 counties located in Jeju island). Others will be excluded in the regression model (National patterns) because of an orthogonality problem, as previously discussed. International competition strongly affects peripheral sector employment, such as agriculture and textiles indus­ tries (Bluestone and Harrison, 1982; Tigges & Tootle, 1990). According to dependency theory, "the periphery is forced to rely on the low cost of its labor for its comparative advan­ tage in the international market. As a result, peripheral economies are disarticulated because firms within the pe­ riphery region are not connected to each other, but depend on center firms for markets, technology, and capital" (Curran and Tomaskovic-Devey, 1991:681). Tigges and Tootle (1990) found that industrial restructuring in the form of increased foreign competition and an increased loss of core transformative jobs are especially threatening to rural men's employment adequacy. Thus, international competition power of commodities was included in the research model. 56 Hypotheses

Farm Structure/Characteristics and Rural Poverty

Since Goldschmidt's seminal research, a number of studies have investigated the relationship between farm structure and the socio-economic well-being in rural commu­ nities. Although previous studies generally have supported the Goldschmidt hypothesis, more recent studies have found mixed or conflicting results (Lobao, 1990:55-65). For example, Green (1985) used county-level data from Missouri to test the Goldschmidt hypothesis. The analysis was based on historical data for the time period 1934-1978. Green (1985:268) found "that farm size had only a limited indepen­ dent effect on the dependent variables reflecting quality of life." In a study of the relationship between the rural well­ being and agricultural change in the Corn Belt and the Great Plains, Gilles and Dalecki (1988:51) found that "scale of agricultural production and county socio-economic status (SES) were strongly associated, but contrary to the Goldschmidt hypothesis, change in the scale of agriculture was positively related to change in county SES." Skees and Swanson (1988) found that indicators of farm structure were not significantly associated with rural county well-being. In a study analyzing the effects of farm 57 structure and industry structure on socio-economic condi­ tions, Lobao (1990) found that industrialized farming had only weak negative relationships with socio-economic condi­ tions. Lobao and Schulman (1991) found that the effects of industrialized farming on rural poverty were more spatially variant and that the effects were weakened substantially when non-farm variables were controlled. Reif (1987:472-474) found that larger family farming was negatively related to poverty. Therefore, it is hypoth­ esized that the percent of farms in the middle-size category will be negatively related to the level of rural poverty. Following the Goldschmidt hypothesis (Goldschmidt, 1978a and 1978b), it is hypothesized that percent larger farming (2.0 ha and more in farm size) will be positively related to the level of rural poverty. In sum, it is hy­ pothesized that the percentage of middle farming units is inversely related to rural poverty and that the percentage of larger farming units in the county will be positively related to levels of rural poverty. Some scholars view "part-time farmers as small-farm operators struggling to continue a way of life and forced by declining farm income or changing economic conditions to obtain off-farm work" (Barlett, 1986:290). This view empha­ sizes "the commitment to continuity in farm production and sees part-time farming as a temporary condition for farm families trying to accumulate capital and skills for 58 entrance into farming on a full-time basis or as a mechanism for easing the exit for marginal producers from agriculture" (Barlett, 1986:290). Traditional Marxists interpret part- time farming as a step in the process of proletarianization (Pugliese, 1991). Also, part-time farming is viewed as an increasingly stable component of agricultural structure and a relatively permanent lifestyle (Albrecht and Murdock, 1984; Barlett, 1986; Bertrand, 1967; Heffernan et al., 1981; Pugliese, 1991). Bertrand (1967:303) argued that "part-time farming will remain as a definite cultural pattern" in the United States. Albrecht and Murdock (1984:393) argued that part- time farming can be seen as "one form of organizational adaptation available to farm operators." Barlett (1986) found that the reasons for part-time farming reveal a merg­ ing of income and life-style benefits. Gasson (1986: 372-374) found that while part-time farming may slow down the rate of farm structural change at macro-level, it is nevertheless associated with a rapid turnover of people. Gasson (1986) concluded that survival of the part-time farming sector does not necessarily mean survival of exist­ ing part-time farming families in farming. Pugliese (1991:143) argued that part-time farming "has tended to become a permanent, self-perpetuating condition, the expres­ sion of a new set of social relations within agriculture production." Buttel argued that part-time farming has played a considerable historic role in slowing the pace of centralization of capital in agriculture and reducing the polarization of the agricultural class structure (Barlett, 1986:291). Thompson et al. (1986) found that farm operators who engaged in off-farm employment were less likely to live in poverty than farm operators who did not engage in off- farm employment. Thus, it is hypothesized that percent part-time farming will be inversely related to rural poverty. A tenant farm is a farm where the operator rents rather than owns the land. Many empirical comparisons of the tenant farm and the full-owner farm indicate that the tenant farmer has numerous disadvantages (Albrecht and Thomas, 198 6). Also, researchers found that full-owner farms were larger and more productive than tenant farms (Goldschmidt, 1978a). Researchers have noted that high levels of tenancy result in problems for communities because tenants are less involved than owners in all types of community-level social and political organizations (Anderson & Ryan, 1943; Slocum, 1962). Also, tenants tend to be more concerned with short­ term profits and less concerned with the long-term produc­ tivity of the farm. Also, many adoption-diffusion studies found that tenancy was inversely related to the adoption of agricultural technologies (Rogers, 1983). However, in a study of 1983 Texas crop producers, Albrecht and Thomas (1986) found that farmers who rented most or all of their 60 farmland had the largest and most productive farms, used better farming practices, and were more involved in communi­ ty affairs. According to a previous Korean study (Chung and Oh, 1992), most tenant farmers were at the poverty level and below (Chung and Oh, 1992). Thus, it is hypothesized that rate of tenancy will be positively related to rural poverty.

Non-Farm Characteristics and Rural Poverty

Previous researchers (Bloomquist and Summers, 1982; Lobao, 1990; Lobao and Schulman, 1991; Tomaskovic-Devey, 1987) considered industry structure to be an important predictor of rural poverty. Industry structure was consid­ ered in terms of the proportion of employment generated by three sectors: core, periphery, and state sectors. Unfortunately, county data on industry structure in Korea were not available. Many researchers found that significant variation in poverty levels by the type of rural economy, with resource-based economies (e.g. agriculture) consis­ tently showing lower socio-economic well-being (Colclough, 1988; Tickamyer & Bokemeier, 1988; Tickamyer & Tickamyer, 1988; Tomaskovic-Devey, 1987). Weinberg (1987) found that the percentage of production on farms in a county was posi­ tively related to poverty rates (percentage of persons below poverty threshold). Therefore, it is hypothesized that percent non-farm population will be inversely related to 61 rural poverty. Women are especially vulnerable to job insecurity. Women have much more limited employment opportunities, much flatter earning curves, and higher poverty rates in rural areas dominated by agriculture and mining (Tickamyer and Bokemeier, 1988; Tickamyer & Tickamyer, 1988; Weinberg, 1987). Studies on the segmented economy found that women tended to be crowded into low-wage, peripheral industries (Hodson, 1984). Kolko (1988) found the wage gap between men and women in the same occupation expanded during the 1980s. However, Gimenez (1990:53) argued that "it is not sex but rather class that propels some women into poverty." That is, capitalist women and petty bourgeois women are not in risk of becoming poor. Propertyless women are always at risk of becoming poor."(Gimenez, 1990:52-53). Gorham (1992:35-36) found that rural women workers had a much higher absolute level of low earners than did rural and urban men or urban women. Kabeer (1991) focused on gender differentials in the distribution of poverty, using a con­ ceptual framework of basic needs and resource entitlement to distinguish between the state and process of poverty. In the empirical study analysis on rural Bangladesh, Kabeer (1991:245) found strong evidence that more women lived in poverty and more acutely, than men. Women also become impoverished through different processes than men. There­ fore, it is hypothesized that percent female will be posi- 62 tively related to rural poverty. Educational qualifications have been identified as a major factor accounting for rural socio-economic conditions (Chung et al., 1989; Chung and Oh, 1992; Kim, 1980; Lichter and Costanzo, 1987; Sheets et al., 1987; Skees and Swanson, 1988; Tigges and Tootle, 1990; Wang, 1989; Weinberg, 1987). For example, Skees and Swanson (1988) found that education was inversely associated with poverty. Lichter and Costanzo (1987) found that educational composition was a major factor accounting for the under-employment differential. Morrill and Wohlenberg (1971), Beeghley (1988) found inverse rela­ tionships between education and poverty. Therefore, it is hypothesized that percent high school graduates will be inversely related to rural poverty.

Spatial/Political Economical Characteristics and Rural Poverty

Region embodies historical and contemporary social forces that determine levels of inequality among their components (Lobao, 1990). According to Markusen (1987: 16-17), "a region is an historically evolved, contiguous territorial society that possesses a physical environment, a socioeconomic, political, and cultural milieu, and a spatial structure distinct from other regions and from the other major territorial units, city and nation." She argued that regional politics may take two quite distinct forms: 1) internal struggles around the future evolution of a region; 2) a region-wide movement in opposition to some external entity (e.g. another region, the larger state or national governments, or an outside economic force (Markusen, 1987). Regionalism refers to "the rallying around one or more distinguishing characteristics of the region, levied as a territorial claim against one or several mechanisms of the state." (Markusen, 1987:17). That is, she defined regionalism as "the consciousness of a resident population about its commonalities across a geographical space and in distinction to groups in other regions." (Markusen, 1987:238). Marxists argued that uneven regional development is the functional requirement of capital accumulation and mobility in search of an abundance of cheap labor (Mandel, 1978; Walker, 1978; Harvey, 1982). That is, many Marxists argue that regional disparities are result of the laws of capital­ ist development. Based on political economy perspective (mainly revised dependency theory), some scholars emphasized the interaction among outside and local capital, the inter­ vention of state, and locally based empowerment efforts in the process of uneven regional development (Dunford, 1988; Evans, 1979; Lovejoy and Krannich, 1982; Markusen, 1987). Historically, the socio-economic condition of HONAM (South-West districts of Korea) has been poorer, while YOUNGNAM (South-East districts of Korea) has been richer. The YOUNGNAM region has been aligned with the majority party of Korea and most rulers of Korea have been natives of YOUNGNAM. In contrast,the HONAM region has been the loca­ tion for leaders of the opposition party. As a result, the political and economic characteristics of the two regions are very distinct from each other. So, it is hypothesized that Region 1 (YOUNGNAM) will be inversely related to rural poverty and that Region 2 (HONAM) will be positively related to rural poverty. In Korea, regionalism (regional antagonism) has been characterized as a form of antagonism by one region against another. This takes three forms: 1) YOUNGNAM people's antagonism against HONAM people; 2) HONAM people's antago­ nism against YOUNGNAM people; and 3) OTHERS people's antago­ nism against HONAM people. The major causes of YOUNGNAM people's antagonism against the HONAM represents an attempt to maintain political superiority and economic advantage. This antagonism and competition has been expressed cultural­ ly in terms of negative stereotypes and prejudice about the HONAM region and people. HONAM people's antagonism against the YOUNGNAM are the result of regional discrimination in economic development and policies. The major causes of OTHERS people's antagonism against HONAM are high in-migration, prejudice, and the same negative stereotypes generally reserved for HONAM people (Chung, 1991). Thus, it is hypothesized that the relationships between the indepen­ dent variables (excluding Region 1 and Region 2) and rural 65 poverty will be different across the three regions (YOUNGNAM, HONAM, and OTHERS). Proximity to metropolitan areas affects the location of economic structures as well as local inequality (Heaton, 1980; Lobao, 1990; McGranahan, 198 0a and 1980b ). McGranahan (1980b:315) argued that the effect of proximity to metropolitan areas "has been reflected in lower median incomes and lower absolute growth in incomes over time as one moves away from metropolitan center". McGranahan (1980b) argued that proximity to metropolitan centers and community commercial centrality affected income differen­ tially, depending on education, age, and proprietorship, and that these differential effects were reflected in the dis­ tribution of income across and within communities. Also, Lobao (1990) found that poverty is higher in counties dis­ tant from metropolitan centers. She, however, found that "when the quality of the county industrial base and aspects of worker power are controlled, other U.S. regional differ­ ences diminish" (Lobao, 1990:164). She argued that "differ­ ences in economic inequality historically regarded to be a consequence of regional culture or ecology were actually more a consequence of regional political economy or of the balance of power between capital and labor embodied in particular area" (Lobao, 1990:164). It is hypothesized that proximity to metropolitan cities will be inversely related to rural poverty. 66 International competition strongly affects the agricul­ tural sector (Bluestone and Harrison, 1982; Tigges and Tootle, 1990). According to (Flora et al., 1992:151), "most natural-resource-based industries, especially agricultural production, are experiencing increased competition interna­ tionally at a time when markets are already flooded. The flight of manufacturing jobs to developing nations demon­ strates that rural labor has been drawn into competition with labor in other countries." Farmer et al. (1989) argued that increased competition from foreign companies in industries that have traditionally been located in rural areas was one of major factors leading to the deterioration of conditions in many rural areas. Attention to the spatial dimension of restructuring empha­ sizes the internationalization of the division of labor, the declining significance of national boundaries, the presence of regional disparities and competition, and so on (Tickamyer and Duncan, 1990) . Bloomquist (1988:22) argued that "intense competition from foreign manufacturing raise doubts about whether rural manufacturing industries can maintain their employment levels, because rural manufacturers tend to specialize in products vulnerable to foreign competition." That is, routine (periphery) manufacturing, such as textile and apparel industries, in rural areas is very sensitive to international competition. Bloomquist (1988) argued that 67 complex (often high-tech) manufacturing industries had a greater potential for employment stability than routine (assembly-line) manufacturing industries. World system theorists conceptualized "agrarian struc­ tures as globally formed through the dynamic of states negotiating an international division of labor" (McMichael and Buttel, 1990:100). Technological changes and the evolu­ tion of modern agro-food complexes are systematically erod­ ing sectorial boundaries between agriculture and industry (Kenney et al., 1989; Kolko, 1988). McGranahan (1980b:7) argued that "the greater vulnerability of production jobs to business cycle, foreign competition, and technological displacement places rural workers at a long-run disadvan­ tage." Also, Tickamyer and Duncan (1990) argued that indus­ tries in rural areas are particularly vulnerable to foreign competition and unfavorable exchange rates. They argued that service sector growth in rural areas is disproportionately located in low-wage consumer and personal services. The predominant occupation in rural Korea is still farming (Hong, 1988). Therefore, it is hypothesized that the international competition power of commodities in the region will be inversely related to rural poverty. As reviewed in the early part of this Chapter, non-farm forces, such as non-farm characteristics and spatial/ political economical characteristics have important effects on rural poverty. Lobao and Schulman (1991:594) found that 68 "the initial relationships between farming patterns and poverty change considerably in most regions when industrial structure, social relations, and spatial variables are controlled." Therefore, the relationships between farm structure/characteristics and rural poverty will be changed when non-farm characteristics and spatial/political economi­ cal characteristics are controlled. CHAPTER V RESEARCH METHODOLOGY

Introduction

This chapter will describe the methodological proce­ dures for testing the hypotheses and investigating the model presented in Chapter IV. Data units and collection will be considered first, followed by the measurement of variables. In final section, methods of analysis will be discussed.

Data Units and Collection

The counties of Korea were used as data units in this study. According to Lobao (1990), data at the county level offer many advantages for the study of rural poverty. For example, she argued that "counties approximate local labor markets in that most people work in their county of resi­ dence" (Lobao, 1990:96). Also, she argued that "the county is considered a superior unit for statistical analyses of socioeconomic conditions." (Lobao, 1990:96). Farmer et al. (1989:504) contend that "while there will always be a need

69 for programs geared toward alleviating the poverty of indi­ viduals, the community is perhaps the more relevant level for public policy intervention." Furthermore, many studies on the Goldschmidt hypothesis have been conducted at the county level (Lobao, 1990:96). In order to empirically examine the relationship between rural poverty and related variables, the data of all counties (137) in the Republic of Korea were collected. In Korea, these 137 counties are completely rural. The county system of Korea is different from that of the U.S. That is, the counties of Korea do not include cities, while counties in the U.S include cities. Data were obtained mainly from "Agricultural Census", "Annual Statistical Report". The Agricultural Censuses have been carried out every ten years in Korea since 1960. This study used the report on the results of the Agricultural Census conducted by the Ministry of Agriculture, Forestry, and Fisheries as of December 1, 1990. Statistical Yearbooks are published by each province (Do). Altogether, nine kinds of Statistical Yearbooks were used in this study. "Yearbooks" have been published irregu­ larly by various local newspaper publishing companies. The Yearbooks have a great deal of information about counties. However, the Yearbooks have not been published in some prov­ inces. As a result, the Yearbooks were used as supplements to make up for deficiencies (e.g. missing data) in the Agricultural Census and the Annual Statistical Reports. 71 "Annual Report on the Farm Household Economy Survey", "The Current Condition of Livelihood Protection (Official Poverty) Population", and other statistical reports and documents from the relevant governmental ministries and research institutes were used as data to describe the historical trends of rural poverty. The current condition of rural poverty is a result of broad, historical developments. Therefore, it is preferred that a longitudinal design be used to analyze rural poverty (Lobao, 1990). This study was originally designed to exam­ ine two points in time: 1980 and 1990. But, the data for 1980 had many missing cases. That is, in the case of 1980 data, many rural poverty related variables were not avail­ able on the county level. The situation for both 1960 and 1970 was even worse. As a result, this study uses county data for 1990 only.

The Measurement of Variables

The following discussion about the measurement of variables for this study will progress according to the order in which they will enter the causal models developed in Chapter IV. Measurement of the independent variables will be presented first, then the dependent variables (rural poverty) will be presented. 72 The Measurement of independent Variables

The selection of independent variables for this study was based on the review of literature as presented in the previous chapters. Farm structure/characteristics, non-farm characteristics, and spatial/political economical character­ istics are three broad categories of independent variables. The farm structure/characteristics include percent middle farming, percent larger farming, percent of part-time farming, and rate of tenancy. Percent smaller farming only will be discussed as part of descriptive analysis in Chapter 6. in the non-farm characteristics, percent non­ farm population, percent female, and percent high school graduates are included. The spatial/political economical characteristics include region 1 (YOUNGNAM), region 2 (HONAM), proximity to metropolitan cities, and international competition power of commodities. Percent Smaller Farming: "Less than .5 ha in farmland size" was used as a standard to identify smaller farming because "less than .5 ha" has been generally used as the standard of smaller farming in Korea (Chung et al., 1989; Hong, 1988). Percent smaller farming refers to the percent­ age of farming having less than .5 ha in farmland size to the total number of farm households in a county. Percent Middle Farming: ".5-2.0 ha in farmland size" was used as a standard to identify the percentage of middle 73 farming to the total number of farm households in a county. Thus, percent middle farming was measured as the percentage of farming having .5-2.0 ha in farmland size to the total number of farm households in a county. Percent Larger Farming: "More than 2.0 ha" in farmland size was used as a standard to identify the percentage of larger farming to the total number of farm households in a county. That is, percent larger farming was measured as the percentage of farming having more than 2.0 ha in farmland size to the total number of farm households in a county. Percent Part-time Farming: Part-time farming has been defined and operationalized using various criteria such as size of operation, occupation, amount of time worked on or off the farm, and other combinations (Barlett, 1986; Frauendorfer, 1966; Frank, 1983; Fuguitt et al., 1977; Fuller, 1984; Gasson, 1986; Pfeffer, 1989; Skees & Swanson, 1988). Although the controversy concerning the operational definition of part-time farming has not been resolved, the percentage of family income derived from the farm may be used to differentiate between full-time and part-time farms (Heffernan et al., 1981). That is, percent part-time farming were measured as the proportion of farms which received 50 percent or less of their gross family income from the farm. Rate of Tenancy: Rate of tenancy was measured as the percentage of borrowed farmland acreage to total farmland 74 acreage in a county. Percent Non-farm Population: Percent non-farm population was measured as the proportion of non-farm popu­ lation to the county's total population. Percent Female: Percent female was measured as the proportion of females to total county population. In multi­ ple regression analysis of this research, percent female was measured as a dummy variable: more than 50 percent=l, 50 percent or less=0. Percent High School Graduates: Percent high school graduates was measured as the proportion of high school graduates to farm household population at the age of 15 and above. This farm population indicator was used as an proxy of each county's total population, because the statistics on high school graduates among each county's total population were not available. Region 1 (YOUNGNAM): Region 1 was measured as a dummy variable: Y0UNGNAM=1, 0thers=0. Region 2 (HONAM): Region 2 was measured as a dummy variable: H0NAM=1, 0thers=0. Proximity to Metropolitan Cities: In Korea, metropoli­ tan cities generally refer to cities of 500,000 or more inhabitants (Seoul, Pusan, Taegu, Kwangju, Inchon, and Taejon). The proximity to metropolitan cities was measured as a dummy variable: Adjacent county to metropolitan cit- ies=l, Not adjacent county to metropolitan cities=0. International Competition Power of Commodities: Inter­ national competition power of agricultural commodities was measured using a Ministry of Agriculture, Forestry and Fisheries' Report (1991a): "Countermeasures to Improve International Competition Power of Agriculture and Stock- breeding Products Item by Item". According to this report, 65 main agricultural commodities were evaluated by four de­ grees of international competition power: "Very Low", "Low", "Equal (average)", and "High" (refer to Table 5.1). In this study, the international competition power of main agricul­ tural commodities in a county were measured using a 4 point scale: Very Low=l, Low=2, Equal=3, and High=4. Ten main agricultural commodities in a county were used to develop a summated scale of international competition power of main agricultural commodities. Ten main agricultural commodities in a county were selected according to the proportion of farms in a county. 76 Table 5.1. The Level of International Competition Power of Main Agricultural Commodities

Level of Commodity Acreage International (1,000 ha) Competitive­ ness High Apple,Orange,Sweet 282 Persimmon,Kiwi,Mushroom, (12 %) Vegetables,Flowers,Pear, Chicken,Hog,Sericulture, Ume,Medicinal Herbs (13) Equal Grape,Peach,Citron,Melon, 101 Watermelon,Strawberry, (4 %) Foreign Vegetables,Wild Edible Greens,Green Perilia,Unripe Beans (12) Low Rice,Barley,Soybean,Corn, Rice 1,244 Potato,Sweet Potato,Hot (51 %) Pepper,Garlic,Onion, Korean Cattle,Dairy Others 783 Farming,Peanut,Sesame, (32 %) Red Bean,Mung Bean,Adlay, Foreign Melon,Tea,Deer, Ginger,Goat,Apiculture, Fodder,Lotus Root,Jujube (26) Kidneybean,Pea,Foxtail 19 Very Low Millet,Kaoliang,Oats, (1 %) Rye,Rape,Pineapple, Banana,Hop,Cherry,Grape for Brewing,Peach for Processing (14) Total 65 2,429(100 %)

Source: Ministry of Agriculture, Forestry and Fisheries, Countermeasures to Improve International Competition Power of Agriculture and Stock-breeding Products Item bv Item.1991a. 77 The Measurement of Dependent Variables

A central question in debates about measurement has been "should poverty be measured in relative or absolute terms" (Beeghley, 1984:322). Absolute standards for poverty emphasize economic insufficiency, while relative standards stress economic inequality and the problems resulting from relative deprivation. According to Ruggles (1990:19), absolute poverty measures, which "define poverty as either income or consumption below some absolute level that repre­ sents an objective minimum, constitute the earliest and broadest class of poverty measures." An absolute poverty measure emphasizes "some particular amount of goods and services as essential to an individual's or a family's welfare" (Schiller, 1989:10). Ruggles (1990:19) argued that "disadvantages of this type of measure is that in fact it is very difficult to establish an objective minimum that really is applicable over a long period (or even across very diver­ gent population groups)." The relative poverty approach argues that "a person is poor when his or her income is significantly less than the average income of the popula­ tion" (Schiller, 1989:11). Relative poverty measures in­ clude various ones (e.g. the bottom 10 or 20 percent of the income distribution, Gini Coefficient, and one-half the median income for society). Opponents of a relative poverty measure argue that "it presents too much of a moving target 78 for policy assessment and that it is in some sense not fair to judge our antipoverty efforts against such a standard." (Ruggles, 1990:19). Also, the relative poverty measure perpetuates poverty in the statistical meaning (Schiller, 1989) . A subjective poverty measure is a relatively new one in poverty research. Subjective poverty definitions "are based on surveys that use households' own assessments of the minimum or just sufficient amounts of income or consumption needed by people like them." (Ruggles, 1990:20). The basic idea of a subjective poverty measure is "that one way to discover the minimum amount of income or consumption that people need to maintain what they consider to be a decent or minimally adequate level of living is to ask them directly" (Ruggles, 1990:21). Although poverty is a very broad concept, this study follows current official poverty measurements in Korea. Thus, rural poverty was measured by three indicators: Total Official Poverty Population, Protected at Home Population, and Self-supporting Low Income Population. Total Official Poverty Population: Total official poverty population was measured as the proportion of the population with household incomes below the official poverty line in a county. In 1990, the criteria to select the offi­ cial poverty population were two: Monthly income per capita (48,000 Won— about less than $ 60) and property worth 79 per household (3,200,000 Won— about less than $ 4,000). According to the Ministry of Health and Social Affairs, Korea (1990), the official poverty population was divided into two sub-dimensions: protected at home population and self-supporting low income population. Both sub-dimensions were used as sub-scales of total official poverty popula­ tion. Protected at Home Population: Protected at home popula­ tion is the first sub-scale. Protected at home population refers to the old and weak who are lacking in ability to earn income because of their physical conditions, such as chronic illness patients. This was measured as the propor­ tion of the protected at home population to the total popu­ lation in a county. Self-supporting Low Income Population: Self-supporting low income population is a sub-scale of total official poverty population. Self-supporting low income population refers to persons who earn their own living costs but their income is below the official poverty line. Thus, self-supporting low income population was measured as the proportion of "self-supporting low income population" to the total population in a county. 80 Methods of Analysis

The data analysis was conducted using the Statistical Package for the Social Sciences (SPSSX). Descriptive sta­ tistics were used to organize and summarize the data col­ lected on the independent variables and the dependent vari­ ables. Frequencies, percentages, means, and standard devia­ tion were computed to summarize the data. Also, these descriptive statistics were used to describe historical trends on rural poverty in Korea. Multiple regression analysis was employed in order to test the hypothesized relationships on rural poverty. In multiple regression analysis, "the estimated coefficient on any independent variables estimates the effect of that variable while holding the other independent variables constant" (Schroeder, 1986:31). That is, multiple regres­ sion coefficients (b values) measure the change in the dependent variable that results from a unit changes in the independent variables. Standardized coefficients (Beta values) measure "the change in the dependent variable that results from a one-standard deviation changes in the inde­ pendent variables" (Schroeder, 1986:31). The standardized coefficients solve the problem of dealing with different units of measure. The coefficient of determination (R ) is an estimate of the proportion of the variance of the depen­ dent variable explained by the linear combination of the independent variables. Multiple regression analyses were conducted first at the national level and then by region (YOUNGNAM, HONAM, and Others). At the national level, blocked regression analysis was added. Major purpose of the blocked regression was to investigate the change of rela­ tionships between farm structure/characteristics and rural poverty when non-farm characteristics and spatial/political economical characteristics were controlled. Blocked regression is a kind of hierarchical regression approach. Hierarchical regression usually focuses on the amount of predictive power that each additional variable contributes. In the hierarchical approach to multiple regression, the researcher specifies from the outset the order or hierarchy in which the predictor variables are to be introduced into the regression equation (Hays, 1988:662). Blocked regression is a regression method of predictor selection in which groups of independent variables are entered cumulatively, according to a specified hierarchy which is determined by theory, the logic and the purpose of the research. When the first block is completed, the analy­ sis proceeds to the second block. In blocked regression, for each block of variables entered, the increase in the variance in the dependent variable explained by that block, beyond the variance explained by the previously entered block, can be calculated (R-square Change). That is, blocked regression allows the unique partitioning of the 82 R-square of each block when it is added to the equation (Hays, 1988; Nie et al, 1975). In general, statistical tests of significance are very important in a regression analysis (Blalock, 1972; Schroeder et al, 1986). However, in this study, the significance test is not relevant as an tool to infer the generalizability about an population (Mohr, 1990), because nearly all coun­ ties of rural Korea were analyzed. The focus of multiple regression analysis will be placed on the consistency of the direction of the relationships and on the magnitude of the coefficients in interpreting the findings. CHAPTER VI FINDINGS

Introduction

This chapter presents the findings on rural poverty in the Republic of Korea. The first part of this chapter pres­ ents the descriptive statistics of the independent and dependent variables. Second, multiple regression analysis on rural poverty will be presented. In the multiple regres­ sion analysis, national patterns, blocked regression, and regional comparisons will be presented in turn.

Descriptive Statistics of the Independent and Dependent Variables

Percent Smaller Farming: Table 6.1 presents the per­ centage distribution of smaller farms in Korea. Of 137 counties, 24 counties (17.5 %) had fewer than 20 percent of all farms which were classified as small. Sixty-nine coun­ ties (50.4 %) had between 20 and 30 percent in the propor­ tion of smaller farms. Forty-four counties (22.1 %) had 30 and more percent in the proportion of smaller farms. Mean

83 84 Table 6.1 Distribution of Percent Smaller Farming

Percent Smaller Frequency Percent Cummulative Percent Farming

Less Than 20 % 24 17.5 17.5

20 - 24.99 % 40 29.2 46.7

25 - 29.99 % 29 21.2 67.9

30 - 34.99 % 26 19.0 86.9

35 and More % 18 13.1 100.0

Total 137 100.0

Mean (ungrouped) = 27.416 Standard Deviation (ungrouped) = 8.625 Range (ungrouped) = 15.06 - 64.67 85 (ungrouped) and Standard Deviation (ungrouped) were 27.416 and 8.625, respectively. The range of percent smaller farming was from 15.06 to 64.67. Percent Middle Farming: Table 6.2 represents the dis­ tribution of percent of all farms in the middle farming category. Seven counties (5.1 %) had less than 50 percent in the middle farming category. Thirty counties (21.9 %) had between 50 and 60 percent, and 100 counties (73.0 %) had 60 and more percent of all farmers classified as middle- sized farms. Mean (ungrouped) and Standard Deviation (ungrouped) were 62.270 and 6.198, respectively. The range of percent middle farming was from 35.07 to 72.91. Percent Larger Farming: Table 6.3 shows the distribu­ tion of percent larger farming. Seventy-five counties (54.7 %) had less than 10 percent of all farms in the larger farming category (more than 2.0 ha in farm size). Fifty counties (36.3 %) had between 10 and 20 percent, and only 12 counties (8.8 %) had "20 and more percent" in the proportion of larger farming. Mean (ungrouped) and Standard Deviation (ungrouped) were 10.128 and 6.526, respectively. The range of percent larger farming was from 0.27 to 34.64. Percent Part-time Farming: Table 6.4 shows the distri­ bution of part-time farming. Twenty-four counties (17.5 %) had "less than 10 percent" which were part-time (farms which received 50 percent or less of their gross family income from the farm). Seventy counties (53.3 %) had between 10 86 Table 6.2 Distribution of Percent Middle Farming

Percent Middle Frequency Percent Cummulative Percent Farming

Less Than 50 % 7 5.1 5.1

50 - 54.99 % 7 5.1 10.2

55 - 59.99 % 23 16.8 27.0

60 - 64.99 % 52 38.0 65.0

65 - 69.99 % 42 30.6 95.6

70 and more % 6 4.4 100.0

Total 137 100.0

Mean (ungrouped) = 62.270 Standard Deviation (ungrouped) = 6.198 Range = 35.07 - 72.91 87 Table 6.3 Distribution of Percent Larger Farming

Percent Larger Frequency Percent Cummulative Percent Farming

Less Than 5 % 35 25.5 25.5

5 - 9.99 % 40 29.2 54.7

10 - 14.99 % 30 21.7 76.6

15 - 19.99 % 20 14.6 91.2

20 and more % 12 8.8 100.0

Total 137 100.0

Mean (ungrouped) = 10.128 Standard Deviation (ungrouped) = 6.52 6 Range = .27 - 34.63 88 Table 6.4 Distribution of Percent Part-time Farming

Percent Part-time Frequency Percent Cummulative Percent

Less Than 10 % 24 17.5 17.5

10 - 14.99 % 43 31.4 48.9

15 - 19.99 % 30 21.9 70.8

20 - 24.99 % 18 13.1 83.9

25 and more % 22 16.1 100.0

Total 137 100.0

Mean (ungrouped) = 17.379 Standard Deviation (ungrouped) = 8.901 Range = 5.11 - 57. 94 89 and 20 percent, and 40 counties (19.2 %) had 20 and more percent in the proportion of part-time farming. Mean (ungrouped) and Standard Deviation (ungrouped) were 17.379 and 8.901, respectively. The Range of part-time farming was from 5.11 to 57.94. Rate of Tenancy: Table 6.5 represents the distribution of tenancy rate. Nineteen counties (13.9 %) had less than 2 0 percent of total farms that were tenanted farms (borrowed farmland acreage to total farmland acreage in a county). Seventy-seven counties (56.2 %) had between 20 and 3 0 per­ cent, and 41 counties (30 %) had 30 and more percent tenancy rate. Mean (ungrouped) and Standard Deviation (ungrouped) were 2 6.597 and 6.528. The range of tenancy rate was from 6.40 to 43.84. Percent non-farm Population: Table 6.6 shows the dis­ tribution of percent non-farm population. Fifty counties (36.5 %) had less than 50 percent of the total population that was non-farm. Fifty-two counties (38.0 %) had a non­ farm population percent of 50 to 60, and 35 counties (25.6 %) had "60 and more percent" in the proportion of non-farm population. Mean (ungrouped) and Standard Deviation (ungrouped) were 51.440 and 12.249, respectively. The range of percent non-farm population was from 9.46 to 75.43. Percent Female: Table 6.7 shows the distribution of percent female. In fifty-two counties (38.0 %), the propor­ tion of females was less than 50 percent. Eighty-five 90 Table 6.5 Distribution of Tenancy Hate

Rate of Tenancy Frequency Percent Cummulative Percent

Less Than 20 % 19 13.9 13.9

20 - 24.99 % 32 23.3 37.2

25 - 29.99 % 45 32.9 70.1

30 - 34.99 % 29 21.2 91.2

35 and more % 12 8.8 100.0

Total 137 100.0

Mean (ungrouped) = 2 6.597 Standard Deviation (ungrouped) = 6.528 Range = 6.40 - 43.84 91 Table 6.6 Distribution of Percent Non-farm Population

Percent Non-farm Frequency Percent Cummulative Percent Population

Less Than 30 % 8 5.8 5.8

30 - 39.99 % 18 13.2 19.0

40 - 49.99 % 24 17.5 36.5

50 - 59.99 % 52 38.0 74.5

60 - 69.99 % 30 21.9 96.4

70 and more % 5 3.7 100.0

Total 137 100.0

Mean (ungrouped) = 51.440 Standard Deviation (ungrouped) = 12.249 Range = 9.46 - 75.43 92 Table 6.7 Distribution of Percent Female

Percent Female Frequency Percent Cummulative Percent

Less Than 50 % 52 38.0 38.0

50 and more % 85 62.0 100.0

Total 137 100.0

Mean (ungrouped) = 50.314 Standard Deviation (ungrouped) = 1.326 Range = 43.42 - 54.07 93 counties (62 %) had 50 and more percent of the total popula­ tion that was female. Mean (ungrouped) and Standard Devia­ tion (ungrouped) were 50.314 and 1.326. The range of per­ cent female was from 43.42 to 54.07. Percent High School Graduates: Table 6.8 shows the distribution of percent high school graduates. Forty- one counties (29.9 %) had fewer than 20 percent of the population that had graduated from high school. Sixty-seven counties (48.9 %) had between 20 and 3 0 percent of high school graduates, and 29 counties (21.2 %) had 30 and more percent who had graduated from high school. Mean (ungrouped) and Standard Deviation (ungrouped) were 24.092 and 8.033. The range of percent high school graduates was from 12.00 to 66.00. Region 1 (YOUNGNAM): Table 6.9 represents the distribu­ tion of Region 1 (YOUNGNAM). Youngnam has 43 counties (31.4 percent). Others has 94 counties (68.6 %). Mean and Standard Deviation were 0.314 and 0.466 respectively. Region 2 (HONAM): Table 6.10 shows the distribution of Region 2 (HONAM). HONAM has 34 counties (24.8 %). OTHERS has 103 counties (75.2 %). Mean and Standard Deviation were 0.248 and 0.434, respectively. Proximity to Metropolitan Cities: Table 6.11 represents the distribution of proximity to metropolitan cities. The number of counties adjacent to metropolitan cities was 26 (19.0 %). The number of counties non-adjacent to 94 Table 6.8 Distribution of Percent High School Graduates

Percent High Frequency Percent Cummulative Percent School Graduates

Less Than 20 % 41 29.9 29.9

20 - 24.99 % 45 32.9 62.8

25 - 29.99 % 22 16.0 78.8

30 - 34.99 % 13 9.5 88.3

35 and more 5 16 11.7 100.0

Total 137 100.0

Mean (ungrouped) = 24.029 Standard Deviation (ungrouped) = 8.033 Range = 12.00 - 66.00 95 Table 6.9 Distribution of Region 1 (YOUNGNAM)

Region Frequency Percent Cummulative Percent

YOUNGNAM (1) 43 31.4 31.4

OTHERS (0) 94 68.6 100.0 Total 137 100.0

Mean = 0.314 Standard Deviation = 0.466 96 Table 6.10 Distribution of Region 2 (HONAM)

Region frequency Percent Cuimnulative Percent

HONAM (1) 34 24.8 24.8

OTHERS (0) 103 75.2 100.0

Total 137 100.0

Mean = 0.248 Standard Deviation = 0.434 97 Table 6.11 Distribution of Proximity to Metropolitan Cities

Proximity to Frequency Percent Cummulative Percent Metropolitan Cities

Adjacent (1) 26 19. 0 19.0

Non-Adjacent (0) 111 81.0 100.0

Mean = 0.190 Standard Diviation = 0.394 98 metropolitan cities was 111 (81 %). Mean and Standard Deviation were 0.190 and 0.394, respectively. International Competition Power of Agricultural Commodities: Table 6.12 shows the distribution for the international competition power of agricultural commodities. Twenty-three counties (16.8 %) had summated scores of 25 or less, and 14 counties (10.2 %) scored above 30. The majori­ ty of counties (73.0 %) scored within a narrow range of 26 to 29. Mean and Standard Deviation were 27.328 and 1.986, respectively. Total Official Poverty Population: Table 6.13 shows the distribution of total official poverty population. Forty-six counties (35.4 %) had less than 10 percent in the population living below the official poverty line. Seventy- one counties (54.6 %) had between 10 and 20 percent in the population in poverty. Thirteen counties (10.0 %) had 20 and more percent in the population in poverty. Seven counties have no data of total official poverty population. Mean (ungrouped) and Standard Deviation (ungrouped) were 12.415 and 6.034, respectively. The range of percent total official population was from 2.01 to 31.67. Counties having high levels of total official poverty population (more than 20 %) numbered 13. Among these coun­ ties, 9 counties were located in HONAM region, while YOUNGNAM and OTHERS regions had 1 and 3 counties, respec­ tively. About one quarter (26.1 %) of total HONAM 99 Table 6.12 Distribution of international Competition Power of Agricultural Commodities

International Competition Frequency Percent Cummulative Power of Commodities

22 2 1.5 1.5 23 4 2.9 4.4 24 5 3.6 8.0 25 12 8.8 16.8 26 20 14.6 31.4 27 26 19.0 50.4 28 29 21.2 71.5 29 25 18.2 89.8 30 8 5.8 95.6 31 3 2.2 97.8 32 3 2.2 100.0

Total 137 100.0

Mean = 27.328 Standard Deviation = 1.986 Range =22-32 100 Table 6.13 Distribution of Total Official Poverty Population

Percent Total Official Frequency Percent Cummulative Poverty Population Percent

Less Than 5 % 17 13.1 13.1

5 - 9.99 % 29 22.3 35.4

10 - 14.99 % 41 31.5 66.9

15 - 19.99 % 30 23.1 90.0

2 0 and more % 13 10.0 100.0

Total 130 100.0

Missing Cases = 7 Mean (ungrouped) = 12.415 Standard Diviation (ungrouped) = 6.034 Range = 2.01 - 31.67 101 counties (34) had high levels of total official poverty population. In YOUNGNAM and OTHERS regions# the proportions were 1 percent and 5.1 percent# respectively. That is, HONAM region had a disproportionate share of the rural poverty. Protected at Home Population: Table 6.14 shows the dis­ tribution of percent protected at home population. Thirteen counties (10.6 %) had less than 1.0 percent of the popula­ tion who were classified as protected at home persons. Forty-nine counties (39.8 %) had between 1 and 2 percent, and 47 counties (38.2 %) had between 2 and 3 percent. Fourteen counties (11.4 %) had 3.0 and more percent of their populations classified as protected at home. Fourteen counties had no data on the numbers of protected at home population. Mean (ungrouped) and Standard deviation (ungrouped) were 2.041 and 0.844 respectively. The range of percent protected at home population was from 0.43 to 4.59. Self-supporting Low income Population: Table 6.15 represents the distribution of percent self-supporting low income population. Fifty-five counties (47.1 %) had less than 10 percent of the total population classified as self- supporting low income persons. Fifty-seven counties (47.1 %) had between 10 and 20 in this poverty category. Seven counties (5.8 %) had 20 and more percent in the pro­ portion of self-supporting low income population. Sixteen counties had no data concerning the variable. Mean 102 Table 6.14 Distribution of Protected at Home Population

Percent Protected at Frequency Percent Cummulative Home Population Percent

Less Than 1.0 % 13 10.6 10.6

1.0 - 1.99 % 49 39.8 50.4

2.0 - 2.99 % 47 38.2 88.6

3.o and more % 14 11.4 100.0

Total 123 100.0

Missing Cases = 14 Mean (ungrouped) = 2.041 Standard Deviation = 0.844 Range = 0.43 - 4.59 103 Table 6.15 Distribution of Self-supporting Low Income Population

Percent Self-supporting Frequency Percent Cummulative Low Income Population Percent

Less Than 5 % 23 19.0 19.0

5 - 9.99 % 34 28.1 47.1

10 - 14.99 % 32 26.5 73.6

15 - 19.99 % 25 20.6 94.2

2 0 and more % 7 5.8 100. 0

Total 121 100.0

Missing Cases = 16 Mean (ungrouped) = 10.915 Standard Deviation (ungrouped) = 5.806 Range = 1.12 - 29.42 104 (ungrouped) and Standard Deviation (ungrouped) were 10.915 and 5.806, respectively. The range of official poverty population was from 1.12 to 29.42.

Multiple Regression Analysis

As presented in the Table 6.16, the relationships (correlation coefficients) between percent middle farming and rural poverty represented some variations depending on rural poverty indicators. That is, percent middle farming and two rural poverty indicators (total official poverty population and self-supporting low income population) repre­ sented positive relationships. This was contradictory to the hypothesis that percent middle farming would be nega­ tively related to rural poverty. However, percent middle farming and protected at home population represented a negative relationship (r=-.180), as hypothesized. Percent larger farming and three rural poverty indicators represent­ ed negative relationships. This was contradictory to the hypothesis that percent larger farming would be positively related to rural poverty. Percent part-time farming and three rural poverty indicators represented negative rela­ tionships, as hypothesized. The relationships between rate of tenancy represented some variations depending on rural poverty indicators. That is, rate of tenancy and two rural poverty indicators (total official poverty population and Table 6.16 Correlation Coefficients

VI V2 V3 V4 V5 V6 V7 V8

VI 1.000 -.077 -.740** .103 .410** .119 -.309** -.008 V2 -.077 1.000 -.281** .430** -.157 -.249** .082 -.416** V3 -.740** -.281** 1.000 -.284** -.415** -.189* .376** -.004 V4 .103 .430** -.284** 1.000 -.113 -.150 .029 -.294** V5 .410** -.157 -.415** -.113 1.000 .354** -.377** .153 V6 .215 -.231** -.256** -.172* .414** 1.000 -.263** .284** V7 -.309** .082 .376** .029 -.377** -.044 1.000 .085 V8 -.008 -.416** -.004 -.294** .153 .432** .085 1.000 V9 .071 -.047 -.102 .044 .265** .345** -.305** -.389** V10 -.021 -.213* .119 .209* -.135 -.005 .209* -.006 Vll .022 .064 .050 .106 .010 .175* .046 .094 V12 .297** -.254** -.367** .071 .619** .438** -.474** .015 V13 -.180* -.161 -.022 -.093 .263** .209* -.282** .146 V14 .333** -.186* -.428** .122 .586** .372** -.520** -.035 V15 -.642** -.698** .749** -.375** -.179* .011 .167 .314** Table 6.16 (continued)

V9 V10 Vll V12 V13 V14 V15

VI .071 -.021 .022 .297** 180* .333** -.642** V2 -.047 -.213 .064 -.254** 161 -.186* -.698** V3 -.102 .119 .050 -.367** 022 -.428** .749** V4 .044 .209 .106 .071 093 .122 -.375** V5 .265** -.135 .010 .619** 263** .586** -.179* V6 .345** -.005 .175* .438** 209* .372** .011 V7 -.305** .209 .046 -.474** 282** -.520** .167 V8 -.389 -.006 .094 .015 146 -.035 .314** V9 1.000 -.020 -.070 .548** 297** .493** -.015 V10 -.020 1.000 .033 -.081 074 -.128 .160 Vll -.070 .033 1.000 .142 068 .127 -.030 V12 .548** -.081 .142 1.000 529** .940** -.011 V13 .300** -.074 .068 .529** 000 .477** .263** V14 .493** -.128 .127 .940** 477** 1.000 -.098 V15 -.015 .160 -.030 -.011 263** -.098 1.000 Table 6.16 (continued)

* p < .05 ** p < .01 VI = Percent Middle Farming; V2 = Percent Larger Farming; V3 = Percent Part-time Farming V4 = Rate of Tenancy; V5 = Percent Non-farm Population; V6 = Percent Female; V7 = Percent High School Graduates; V8 = Region 1 (YOUNGNAM); V9= Region 2 (HONAM); V10 = Proximity to Metropolitan Cities; VII = International Competition Power of Commodities; V12 = Total Official Poverty Population; V13 = Protected at Home Population; V14 = Self-supporting Low Income Population VI5 = Percent Small Farming 108 self-supporting low income population) represented positive relationships, as hypothesized. However, rate of tenancy and protected at home population represented a negative relationship (r=-.093). This was contradictory to the hypothesis that rate of tenancy would be positively related to rural poverty. Percent non-farm population and three rural poverty indicators represented positive relationships. This was contradictory to the hypothesis that percent non-farm population would be negatively related to rural poverty. Percent female and three rural poverty indicators represent­ ed positive relationships, as hypothesized. Percent high school graduates and three rural poverty indicators repre­ sented negative relationships, as hypothesized. The relationships between region 1 (YOUNGNAM) and rural poverty represented some variations depending on rural poverty indicators. That is, region 1 and total official poverty population represented a weak positive relationship (r=.015). Region 1 and protected at home population repre­ sented a positive relationship (r=.146). This was contra­ dictory to the hypothesis that region 1 (YOUNGNAM) would be negatively related to rural poverty. Region 1 and self- supporting low income population represented a weak negative relationship (r=-.035). Region 2 (HONAM) and three rural poverty indicators represented positive relationships, as hypothesized. Proximity to metropolitan cities and three 109 rural poverty indicators represented negative relationships, as hypothesized. International competition power of commod­ ity and three rural poverty indicators represented positive relationships. This was contradictory to the hypothesis that international competition power of commodity would be negatively related to rural poverty. Inter-correlations among independent variables were checked for the possibility of multi-collinearity (Schroeder et al.,1986). As presented in the Table 6.16, all inter-correlation coefficients were at acceptable levels for the multiple regression analysis, although some inter­ correlation coefficients (e.g. correlation coefficient between percent smaller farming and percent part-time farm­ ing was .749) were rather high. According to Bohrnstedt and Knoke (1982), the upper limit to determine multicollinearity among the independent variables is 0.80. In the multiple regression analysis, national patterns will be investigated first. Following this will be the analysis of regional patterns. 110 National Patterns

(Total Official Poverty Population)

When total official poverty population was regressed on the farm structure/characteristics, non-farm characteristics variables, and spatial/political economical characteristics, 71 percent of the variance was explained (Table 6.17). Counties with greater proportions of middle farming, as opposed to the small farming (the reference category), had lower poverty rates (Beta=-.245), as hypothesized. Counties with greater proportions of larger farming, as opposed to the small farming (the reference category), had lower pover­ ty rates (Beta=-.394). This was contradictory to the re­ search hypothesis that percent larger farming would be positively related to rural poverty. That is, this result did not support the Goldschmidt hypothesis. Counties with a greater proportion of part-time farming had lower poverty rates (Beta=-.346), as hypothesized. Counties with higher rates of tenancy had higher poverty rates (Beta=.190). This supports the hypothesis that the rate of tenancy will be positively related to rural poverty. Among non-farm characteristics, percent non-farm popu­ lation had positive effects on rural poverty (Beta=.364). This was contradictory to the hypothesis that the percent non-farm population would be inversely related to rural Ill Table 6.17 Regression of Total Official Poverty Population (N=13 0)

Independent Variables b Beta Farm Structure/Characteristics Percent Middle Farming -.233 -.245** Percent Larger Farming -.364 -.394** Percent Part-time Farming -.230 -.346** Rate of Tenancy .172 .190** Non-farm Characteristics Percent Non-farm Population .180 .364** Percent Female 1.302 . 103 Percent High School Graduates -.094 -.128* Spatial/Political Economical Characteristics Region 1 (YOUNGNAM) -.928 -.073 Region 2 (HONAM) 4.331 .317** Proximity to Metropolitan Cities -.317 -.086 International Competition Power of Commodities .614 .202** Intercept 4.784 .000

F = 25.821 Signif. F = .000 * p < .05 ** p < .01 112 poverty. This seems to be partly attributable to 1) the fact that many in the non-farm population of rural Korea are the old who are lacking in ability to earn income; 2) the fact that most non-farm jobs in rural Korea have been con­ centrated in the periphery sector, which is noted for low wages and labor intensity. Percent female had a positive effect on rural poverty (Beta=.103), as hypothesized. Counties with greater proportion of high school graduates had lower poverty rates (Beta=-.128), as expected. In the spatial/political economical characteristics, region 2 (HONAM), had a positive effect on rural poverty, as hypothesized (Beta=.317). However, region 1 (YOUNGNAM) and proximity to metropolitan cities had comparatively small negative effects on rural poverty (Beta=-.016 and Beta=-.071). International competition power of commodities had a positive effect on rural poverty. This was contradic­ tory to the hypothesis that the competition power of commod­ ities would be inversely related to rural poverty. It seems to be related to the fact that in Korea, most agricultural commodities had very low international competition power. That is, 84 percent of main agricultural commodities (in acreage) had "low" or "very low" level of international competition power (Ministry of Agriculture, Forestry and Fisheries, 1991). 113 (Protected at Home Population)

When protected at home population was regressed on the farm structure and characteristics, non-farm characteris­ tics, and spatial/political economical characteristics, 37 percent of the variance was explained (Table 6.18). Coun­ ties with a greater proportion of middle-sized farms, as opposed to the small-sized farms (the reference category), had lower levels of poverty (Beta=-.644), as hypothesized. Counties with a greater proportion of large-scale farms, as opposed to small-sized farms (the reference category), had lower levels of poverty (Beta=-.203). This was con­ tradictory to the research hypothesis that percent larger farming would be positively related to rural poverty. Counties with a greater proportion of part-time farming had lower poverty levels (Beta=-.420), as expected (Beta=-.462). Rate of tenancy had a small effect on rural poverty (Beta=-.030). In terms of non-farm characteristics, percent non-farm population had a positive effect on rural poverty (Beta=.176). This was contradictory to the research hypoth­ esis that percent non-farm population would be negatively related to rural poverty. But this was consistent with the regression result of percent total official poverty popula­ tion. Percent female had a negative effect on rural poverty (Beta=-.105). This was contradictory to the research 114 Table 6.18 Regression of Protected at Home Population (N=123)

Independent Variables b Beta

Farm Structure/Characteristics Percent Middle Farming -.084 -.644** Percent Larger Farming -.027 -.203 Percent Part-time Farming -.039 -.420** Rate of Tenancy -.004 -.030 Non-farm Characteristics Percent Non-farm Population .012 .176 Percent Female -.192 -.105 Percent High School Graduates -.013 -.130 Spatial/Political Economical Characteristics Region 1 (YOUNGNAM) .395 .224 Region 2 (HONAM) .673 .358** Proximity to Metropolitan Cities -.007 -.004 International Competition Power of Commodities .064 .152* Intercept 6.025 .000

R2 = .37 F = 5.961 Signif. = .000 * p < .05 ** p < .01 115 hypothesis that percent female would be positively related to rural poverty. Counties with a greater proportion of high school graduates had lower poverty rates (Beta=-.130), as hypothesized. In the spatial/political economical characteristics, region 1 (YOUNGNAM) had a positive effect on rural poverty (Beta=.224). This was contradictory to the research hypothesis that region 1 (YOUNGNAM) would be negatively related to rural poverty. Region 2 (HONAM) had a positive effect on rural poverty (Beta=.358), as hypothesized. Proximity to metropolitan cities had only a small effect on rural poverty (Beta=-.004). International competition power of commodities had a positive effect on rural poverty (Beta=.152). This was contradictory to the research hy­ pothesis. But, this was consistent with the regression result of percent total official poverty population.

(Self-supporting Low Income Population)

When self-supporting low income population was regressed on the farm structure/characteristics, non-farm characteris­ tics, and spatial/political economical characteristics, 64 percent of the variance was explained (Table 6.19). Coun­ ties with greater proportion of middle-sized farms, as opposed to the small farming (the reference category), had lower poverty rates (Beta=-.204), as hypothesized. Counties 116 Table 6.19 Regression of Self-supporting Low Income Population (N=121)

Independent Variables b Beta

Farm Structure/Characteristics Percent Middle Farming -.184 -.204* Percent Larger Farming -.361 -.378** Percent Part-time Farming -.227 -.364** Rate of Tenancy .181 .209** Non-farm Characteristics Percent Non-farm Population .145 .301** Percent Female 1.233 .097 Percent High School Graduates -.101 -.146 Spatial/Political Economical Characteristics Region 1 (YOUNGNAM) -1.337 -.111 Region 2 (HONAM) 3 . 322 .258** Proximity to Metropolitan Cities -1.749 -.122 International Competition Power of Commodities .583 .203** Intercept 3.046 .000

R2 = .64 F = 17.997 Signif. F = .000 * p < .05 ** p < .01 117 with a greater proportion of larger farming, as opposed to small farming, had lower poverty rates (Beta=-.378). This was contradictory to the research hypothesis that percent larger farming would be positively related to rural poverty. But, this was consistent with the regression result of percent total official poverty population. Counties with a greater proportion of part-time farming had lower poverty rates (Beta=-.364), as hypothesized. Rate of tenancy had a positive effect on rural poverty (Beta=.209), as hypothe­ sized. In the non-farm characteristics, counties with larger non-farm populations had higher poverty rates (Beta=.301). This was contradictory to the research hypothesis. But, this was consistent with the regression result of percent total official poverty population. Percent female had a positive effect on rural poverty (Beta=.097), as hypothe­ sized. Counties with greater proportion of high school graduates had lower poverty rates (Beta=-.146), as hypothe­ sized. In the spatial/political economical characteristics, Region 1 (YOUNGNAM) had a negative effect on rural poverty (Beta=-.Ill), as hypothesized. Region 2 (HONAM) had a posi­ tive effect on rural poverty (Beta=.258), as hypothesized. Proximity to metropolitan cities had a negative effect on rural poverty (Beta=-.122), as hypothesized. That is, counties adjacent to metropolitan cities had lower rural 118 poverty levels when compared with counties non-adjacent to metropolitan cities. International competition power of commodities had a positive effect on rural poverty (Beta=.203). This was contradictory to the research hy­ pothesis that international competition power of commodities would be negatively related to rural poverty. But, this was consistent with the regression result of percent total official poverty population.

Blocked Regression

Based on the research model as outlined in Figure 4.1 in Chapter IV, farm structure/characteristics were entered first, followed by non-farm characteristics, and spatial/ political economical characteristics. Table 6.20 shows the results of blocked regression for total official poverty population. The block of farm struc­ ture/characteristics explained 30 percent of the variance. The next block entered is non-farm characteristics. The additional variance by the block of non-farm characteristics was 28 percent. The third block, spatial/political economi­ cal characteristics explained an additional 13 percent of the variance. In the block of farm structure/characteristics, percent middle farming, as opposed to percent small farming (the reference category), had a negative effect on rural poverty Table 6.20 Blocked Regression of Total Official Poverty Population (N=13 0)

Independent Variables Beta (b) Beta (b) Beta (b) Farm Structure/Characteristics Percent Middle Farming -.227 (-.216) -.147 (-.140) -.245 (-.233)1 ** Percent Larger Farming -.499 (-.462)** -.266 (-.246)** -.394 (-.364 \ ** Percent Part-time Farming -.634 (-.420)** -.150 (-.099) -.346 (-.230 1 ** Rate of Tenancy .129 ( .118) .249 ( .226)** .190 ( .172 1 ** Non-farm Characteristics Percent Non-farm Population .419 ( .207)** 364 ( .180)** Percent Female .228 (2.876)** ,103 (1.302) Percent High School Graduates -.269 (-.198)** ,128 (-.094)* Spatial/Political Economical Characteristics Region 1 (YOUNGNAM) 073 (-.928) Region 2 (HONAM) 317 (4.331)** Proximity to Metropolitan Cities ,086 (-.317) Internatxonal Competition Power of Commodities ,202 ( .614)** Intercept .000 (34.690) .000 (11.534) .000 (4.784) R2 .30 .58 ,71 R2 Change .28 ,13 p < . 05 ** p < .01

to 120 (b=-.216), as hypothesized. Percent larger farming, as opposed to percent small farming (the reference category), had a negative effect on rural poverty (b=-.462). This was contradictory to the research hypothesis. Percent part-time farming had a negative effect on rural poverty (b=-.420), as hypothesized. Rate of tenancy had a positive effect on rural poverty (b=.118), as hypothesized. When the second block (non-farm characteristics) was added in the model, the relationships between farm structure/characteristics and rural poverty (total official poverty population) presented the same patterns. However, the relationships between farm structure/characteristics and rural poverty became weaker, except for rate of tenancy. In the block of non-farm characteristics, percent non-farm population had a positive effect on rural poverty (b=.2 07). This was contradictory to the research hypothesis. Percent female had a positive effect on rural poverty (b=2.876), as hypothesized. Percent high school graduates had a negative effect on rural poverty (b=-.198), as hypothesized. When the third block (spatial/political economical characteristics) was added in the model, the relationships between farm structure/characteristics and rural poverty (total official poverty population) presented the same pat­ terns as the results of the first and second models. Also the relationships between non-farm characteristics and rural poverty presented the same patterns as the results from the 121 second model. Table 6.21 shows the results of the blocked regression for protected at home population. The block of farm struc­ ture/characteristics explained 20 percent of the variance. The second block, non-farm characteristics explained an additional 10 percent of the variance. The block of spa­ tial/political economical characteristics added another 7 percent. In the block of farm structure/characteristics, percent middle farming, as opposed percent small farming (the reference category), had a negative effect on rural poverty (b=-.090), as hypothesized. Percent larger farming, as opposed to percent small farming (the reference category), had a weak negative effect on rural poverty (b=-.047). Percent part-time farming had a negative effect on rural poverty (b=-.059), as hypothesized. Rate of tenancy had a small negative effect on rural poverty (b=-.006). When the second block (non-farm characteristics) was added in the model, the relationships between farm structure/characteristics and rural poverty (protected at home population) presented the same pattern except for rate of tenancy. The relationship between rate of tenancy and rural poverty was changed from a small negative relationship (b=-.006) to a small positive relationship (b=.011). In the non-farm characteristics, percent non-farm population had a small positive effect on rural poverty Table 6.21 Blocked Regression of Protected at Home Population (N=123)

Independent Variables Beta (b) Beta (b) Beta (b) Farm Structure/Characteristics Percent Middle Farming -.691 (-.090)** -.656 (-.085) ** -.644 (-.084' ** Percent Larger Farming -.362 (-.047)** -.241 (-.032) * -.203 (-.027 Percent Part-time Farming -.647 (-.059)** -.369 (-.033) * -.420 (-.039 ** Rate of Tenancy -.051 (-.006) .011 ( .001) -.030 (-.004 Non-farm Characteristics Percent Non-farm Population .224 ( .016) * .176 ( .0121 j Percent Female .073 ( .133) -.105 (-.192, Percent High School Graduates -.213 (-.022) * -.130 (-.013 Spatial/Political Economical Characteristics Region 1 (YOUNGNAM) .224 ( -395) Region 2 (HONAM) .358 ( .673)** Proximity to Metropolitan Cities -.004 (-.007) International Competition Power of Commodities .152 ( .064)* Intercept .000 (9.306) . 000 (7.826) .000 (6.025) R2 .20 .30 .37 R2 Change .10 .07 * p < .05 ** p < .01

ro ro 123 (b=.016). Percent female had a positive effect on rural poverty (b=.133), as hypothesized. Percent high school graduates had a small negative effect on rural poverty (b=-.022). When the third block (spatial/political economical characteristics) was added to the model, the relationships between farm structure/characteristics and rural poverty (protected at home population) presented the same patterns as the results of the first and second models, except for rate of tenancy. In the relationships between non-farm characteristics and rural poverty, the effect of percent female on rural poverty was changed from a positive rela­ tionship (b=.133) to a negative relationship (b=-.192). Percent non-farm population and percent high school gradu­ ates presented the same patterns with the results of the second model. Table 6.22 represents the results of the blocked re­ gression of self-supporting low income population. The first block, farm structure/characteristics, explained 30 percent of the variance. The block of non-farm characteris­ tics explained an additional 23 percent of the variance. Finally, the block of spatial/political economical char­ acteristics explained an additional 11 percent of the vari­ ance. In the block of farm structure/characteristics, percent middle farming, as opposed to percent small farming (the Table 6.22 Blocked Regression of Self-supporting Low Income Population (N=121)

Independent Variables Beta (b) Beta (b) Beta (b) Farm Structure/Characteristics

Percent Middle Farming -.183 [-.165) -.108 (-.098) -.204 (-.184)i * Percent Larger Farming -.407 -.389)** -.219 (-.210)** -.378 (-.361 i ** Percent Part-time Farming -.629 ; -.39l)** -.173 (-.107) -.364 (-.227 i ** Rate of Tenancy .142 .123) .246 ( .213)** .209 ( .181 1 ** Non-farm Characteristics Percent Non-farm Population ,357 ( .172)** .301 ( .145)** Percent Female ,172 (2.170)* .097 (1.233) Percent High School Graduates ,289 (-.200)** -.146 (-.10l)* Spatial/Political Economical :naracteristics Region 1 (YOUNGNAM) 111 (-1.337) Region 2 (HONAM) 258 ( 3.322)** Proximity to Metropolitan Cities 122 (-1.749) International Competition Power of Commodities 203 ( .583)** Intercept .000 (28.566) .000 (9.694) .000 ( 3.046) .30 .53 .64 R Change .23 .11 p < .05 ** p < .01 125 reference category), had a negative effect on rural poverty (b=-.165), as hypothesized. Percent larger farming, as opposed to small farming (the reference category), had a negative effect on rural poverty (b=-.389). It was contra­ dictory to the research hypothesis. Percent part-time farming had a negative effect on rural poverty (b=-.391), as hypothesized. Rate of tenancy had a positive effect on rural poverty (b=.123), as hypothesized. When the second block (non-farm characteristics) was added to the model, the relationships between farm struc­ ture/characteristics and rural poverty (self-supporting low income population) presented the same pattern. In the non-farm characteristics, percent non-farm population had a positive effect on rural poverty (b=.172). This was contradictory to the research hypothesis. Percent female had a positive effect on rural poverty (b=2.170), as hypoth­ esized. Percent high school graduates had a negative effect on rural poverty (b=-.200), as hypothesized. When the third block (spatial/political economical characteristics) was added to the model, the relationships between farm structure/characteristics presented the same patterns as the results of the first and second models. Also, the relationships between non-farm characteristics and rural poverty presented the same patterns as the results of the second model. 12 6 In sum, as presented in the tables (Table 6.20 to Table 6.22), each block added a significant amount of explained variance. This may mean that previous research based on structural perspectives is adequate to the study of rural poverty in Korea. In general, the relationships between farm structure/characteristics and rural poverty were not greatly changed when non-farm characteristics and spatial/political economical characteristics were con­ trolled. However, the magnitudes of the coefficients were changed a little bit.

Regional Comparisons

(Total Official Poverty Population)

Table 6.23 shows the regression of total official poverty population by region. In the three regions, the amount of explained variance for rural poverty was high (from 63 percent to 76 percent). In YOUNGNAM, significant variables were mainly farm structure/characteristics and proximity to metropolitan cities. In HONAM, significant variables were international competition power of commodi­ ties (b=-2.668) and percent non-farm population (b=.26l). In OTHERS, significant variables were non-farm characteris­ tics, such as percent non-farm population (b=.229), percent female (b=3.169), and percent high school graduates Table 6.23 Regression of Total official Poverty Population by Region

Independent Variables YOUNGNAM (N=43) HONAM (N=34) OTHERS (N=53) Beta (b) Beta (b) Beta (b) Farm Structure/Characteristics Percent Middle Farming -.744 (-.350)* -.239 (-.195] Percent Larger Farming -.388 (-.391)** -.169 (-.127 -.306 -.258 Percent Part-time Farming -1.068 (-.347)** -.350 (-.189 -.038 -.032 Rate of Tenancy .289 ( .198) .194 ( .118* .130 .122 Non-farm Characteristics Percent Non-farm Population .139 ( .043) .352 ( .261)** .546 ( .229)** Percent Female .121 (1.763) .021 ( .364) .257 (3.169)* Percent High School Graduates -.068 (-.029) -.237 (-.194) -.425 (-.290)** Spatial/Political Economical Characteristics Proximity to Metropolitan cities —.424(-4.041)** -.210 (-2.668) .201 (2.689) International Competition Power of Commodities .231 ( .516) .471 (1.092)** .074 ( .190) Intercept .000 (21.291) .000 (-9.280) .000 (9.447) R2 .63 .76 .64 p < .05 ** p < .01

ro 128 (b=— .290). However/ in general, the three regions (YOUNGNAM, HONAM, and Others) presented very similar patterns except for the case of proximity to metropolitan cities. In farm structure/characteristics, percent middle farming, as op­ posed to percent small farming, had negative effects on rural poverty across the three regions, as hypothesized. Percent larger farming, as opposed to percent small farming, had negative effects on rural poverty across the three regions. These results were contradictory to the research hypothesis that percent larger farming would be positively related to rural poverty. Percent part-time farming had negative effects on rural poverty across the three regions, as hypothesized. Rate of tenancy had positive effects on rural poverty across the three regions. In non-farm characteristics, percent non-farm popula­ tion had positive effects on rural poverty. This was con­ tradictory to the research hypothesis. Percent female had positive effects on rural poverty across the three regions, as hypothesized. Percent high school graduates had negative effects on rural poverty across three regions, as expected. In the spatial/political economical characteristics, proximity to metropolitan cities had negative effects on rural poverty in YOUNGNAM and HONAM (b=-4.041 and b=-2.668), while a positive effect in OTHERS (b=2.689). International competition power of commodities had positive 129 effects on rural poverty across the three regions.

(Protected at Home Population)

Table 6.24 shows the regression of protected at home population by region. In the three regions, the amount of explained variance in rural poverty ranged from 22 percent to 47 percent. In YOUNGNAM, significant variables were farm structure/characteristics and international competition power of commodities (b=.170). In HONAM, international competition power of commodities had a positive effect on rural poverty (b=.137). However, the relationship was not significant. In OTHERS, significant variables were percent middle farming (b=-.094) and percent non-farm population (b=.025). In general, the relationships between independent vari­ ables and protected at home population had variations by region. In the farm structure/characteristics, percent middle farming had negative effects on rural poverty across the three regions, as hypothesized. Percent larger farming had negative effects on rural poverty in YOUNGNAM and OTHERS (b=-.071 and b=-.017), while a positive effect in HONAM (b=.037). Part-time farming had negative effects on rural poverty in YOUNGNAM and OTHERS (b=-.033 and b=-.045), as hypothesized. In contrast, percent part-time farming had a small positive effect in HONAM (b=.008). Rate of tenancy Table 6.24 Regression of Protected at Home Population by Region

Independent Variables YOUNGNAM (N=43) HONAM (N=34) OTHERS (N=46) Beta (b) Beta (b) Beta (b) Farm Structure/Characteristics Percent Middle Farming -.792 (-.073)* -.141 (-.019) -.643 (-.094 Percent Larger Farming -.360 (-.071)* .292 ( .037) -.150 (-.017 Percent Part-time Farming -.515 (-.033) .088 ( .008) -.390 (-.045 Rate of Tenancy -.104 (-.014) -.283 (-.029) .120 ( .015 Non-farm Characteristics Percent Non-farm Population -.044 -.003 .031 .436 .025 Percent Female .152 .037 -.020 Percent High School Graduates (-.011 -.278 -.026 Spatial/Political Economical Characteristics Proximity to Metropolitan Cities -.184 (-.343) -.056 (-.118) .220 ( .387) International Competition Power of Commodities .389 ( .170)* .354 ( .137) -.035 (-.012) Intercept .000 (3.469) .000(-1.103) .000 (7.779) R2 .47 .22 .39 p < .05 ** p < .01

oCO 131 had a negative effect on rural poverty in YOUNGNAM (b=-.104) and HONAM (b=-.029), while a positive effect in OTHERS (b=.015). In the non-farm characteristics, percent non-farm population had positive effects on rural poverty in HONAM and Others (b=.031 and b=.025), while a small negative effect in YOUNGNAM (b=-.003). Percent female had positive effects on rural poverty in YOUNGNAM (b=.432) and HONAM (b=.037), while a negative effect in OTHERS (b=-.033). Percent high school graduates had very consistent negative effects on rural poverty across the three regions. In the spatial/political economical characteristics, proximity to metropolitan cities had negative effects on rural poverty in YOUNGNAM and HONAM (b=-.343 and b=-.118), while a positive effect in OTHERS (b=.387). International competition power of commodities had positive effects on rural poverty in YOUNGNAM and HONAM (b=.170 and b=.137), while a small negative effect in OTHERS (b=-.012).

(Self-supporting Low Income Population)

Table 6.25 shows the regression of self-supporting low income population by region. In the three regions, the amount of explained variance ranged from 47 percent to 79 percent. In YOUNGNAM, significant variables were percent larger farming (b=-.384) and proximity to metropolitan Table 6.25 Regression of Self-supporting Low Income Population by Region

Independent Variables YOUNGNAM (N=43) HONAM (N=34) OTHERS (N=44) Beta (b) Beta (b) Beta (b) Farm Structure/Characteristics Percent Middle Farming -.078 (-.042) -.233 (-.180) -.166 (-.190] Percent Larger Farming# -.336 (-.384)* -.242 (-.171) -.336 (-.323 Percent Part-time Farming -.376 (-.139) -.386 (-.197) -.213 (-.184 Rate of Tenancy .103 ( .080) .262 ( .151) .064 ( .060] Non-farm Characteristics Percent Non-farm Population .082 ( .029) .332 ( .233)* .368 ( .160)* Percent Female .257 (4.249) .018 ( .287) .167 (2.013) Percent High School Graduates -.178 (-.086) -.221 (-.171) -.352 (-.244)* Spatial/Political Economical Characteristics Proximity to Metropolitan Cities —.349(—3.774)* -.214(-2.569)* .141 (1.839) International Competition Power of Commodities .054 ( .137) .438 ( .959)** .195 ( .493) Intercept .000 (9.653) .000(-8.216) .000(-63.48) R2 .47 .79 .59

* p < .05

to ro 133 cities (b=-3.774). In HONAM, significant variables were international competition power of commodities (b=.959), percent non-farm population (b=.233), and proximity to metropolitan cities (b=-2.569). In OTHERS, significant variables were non-farm characteristics, such as percent non-farm population (b=.160) and percent high school graduates (b=-.244). However, in general, the three regions presented very similar patterns except for the case of proximity to metropolitan cities. In the farm structure/ characteristics, percent middle farming had positive effects on rural poverty across the three regions. Percent larger farming had negative effects on rural poverty across the three regions. These results were contradictory to the research hypothesis. Percent part-time farming had negative effects on rural poverty across three regions. Rate of tenancy had consistently positive effects on rural poverty across all three regions. In the non-farm characteristics, percent non-farm population had very consistently positive effects on rural poverty across three regions. Percent female had negative effects on rural poverty across the three regions. Percent high school graduates had very consistent negative effects on rural poverty across all three regions. In the spatial/political economical characteristics, proximity to metropolitan cities had negative effects in YOUNGNAM and HONAM (b=-3.774 and b=-2.569), while a positive 134 effect on OTHERS (b=1.839). International competition power of commodities had consistent positive effects on rural poverty across the three regions. This was contradictory to the hypothesis that international competition power of commodities would be negatively related to rural poverty. CHAPTER VII SUMMARY AND CONCLUSIONS

The first part of this chapter presents a summary of major findings from the study. It includes (1) descriptive statistics of the variables; and (2) multiple regression analysis and the results from testing the research hypothe­ ses. The second part of this chapter will suggest policy implications to help alleviate rural poverty in Korea. The third part of this chapter will describe limitations of the research. In the final part of this chapter, recommenda­ tions for future research will be discussed.

Summary of Findings

Descriptive Statistics of the Variables

In the proportion of smaller farming, 24 counties (17.5 % of total counties) had less than 20 percent, 69 counties (50.4 %) had between 20 and 30 percent, 44 coun­ ties (22.1 %) had 30 and more percent. In the proportion of middle farming, 100 counties (73.0 %) had 60 and more percent, 30 counties (21.9 %)

135 136 had between 50 and 60 percent. Seven counties (5.1 %) had less than 50 percent. In the proportion of larger farming, 75 counties (54.7 %) had less than 10 percent, 50 counties (36.3 %) had between 10 and 20 percent, 12 counties (8.8 %) had 2 0 and more percent. In the proportion of part-time farming, 24 counties (17.5 %) had less than 10 percent, 70 counties (53.3 %) had between 10 and 20 percent, 40 counties (19.2 %) had 20 and more percent. In the rate of tenancy, 19 counties (13.9 %) had less than 20 percent, 77 counties (56.2 %) had between 2 0 and 30 percent, 41 counties (30 %) had 30 and more percent. In the proportion of non-farm population, 50 counties (36.5 %) had less than 50 percent, 52 counties (38 %) had between 50 and 60 percent, 35 counties (25.6 %) had 60 and more percent. In the proportion of female, 52 counties (38 %) had less than 50 percent, 85 counties (62 %) had 50 and more percent. In the proportion of high school graduates, 41 counties (29.9 %) have less than 20 percent, 67 counties (48.9 %) had between 20 and 30 percent, 29 counties (21.2 %) had 30 and more percent. In the distribution of Region 1 (YOUNGNAM), YOUNGNAM has 43 counties (31.4 %) while OTHERS has 94 counties 137 (68.6 %). In the distribution of Region 2 (HONAM), HONAM has 34 counties (24.8 %) while OTHERS has 103 counties (75.2 %). In the proximity to metropolitan cities, the number of counties adjacent to metropolitan cities was 26 (19.0 %) , the number of counties non-adjacent to metropoli­ tan cities was 111 (81 %). In the summated scale on international competition power of commodities, 23 counties (16.8 %) scored 25 or less than 25, 14 counties (10.2 %) had a score of 30 and more, and 100 counties (73 %) scored between 26 and 29. In the proportion of total official poverty population, 46 counties (35.4 %) had less than 10 percent, 71 counties (54.6 %) had between 10 and 20, 13 counties (10 %) had 20 and more percent. In the proportion of protected at home population, 13 counties (10.6 %) had less than 1.0 percent, 49 counties (39.8 %) had between 1 and 2 percent, 47 counties (38.2 %) had between 2 and 3 percent, 14 counties (11.4 %) had 3.0 and more percent. In the proportion of self-supporting low income popula­ tion, 55 counties (47.1 %) had less than 10 percent, 57 counties (47.1 %) had between 10 and 2 0 percent, 7 counties (5.8 %) had 20 and more percent. 138 Multiple Regression Analysis

(National Patterns)

When the three rural poverty indicators (total official poverty population, protected at home population, and self- supporting low income population) were regressed on the farm structure/characteristics, non-farm characteristics, and spatial/political economical characteristics, the variances explained were 71 percent, 37 percent, and 64 percent, respectively. Table 7.1 presents the summary of multiple regression analyses (national patterns). The relationships between independent variables and dependent variables had comparatively consistent results. In the farm structure/ characteristics, percent middle farming had consistently negative effects on rural poverty, as hypothesized. Percent larger farming had consistently positive effects on rural poverty. This was contradictory to the research hypothesis that percent larger farming would be positively related to rural poverty. Percent part-time farming had consistently negative effects on rural poverty, as hypothesized. Rate of tenancy had positive effects on total official poverty population and self-supporting low income population, while only a weak effect on the protected at home population. Table 7.1. Summary of Multiple Regression (National Pattern)

Hypothesized Empirical Findings Test Independent Variables Effects with Results Poverty D1 D2 D3 Farm Structure/Characteristics Percent Middle Farming A Percent Larger Farming + R Percent Part-time Farming A Rate of Tenancy + + • + PA Non-farm Characteristics Percent Non-farm population + + + R Percent Female + + + PA Percent High School Graduates A Spatial/Political Economical Characteristics Region 1 (YOUNGNAM) + M Region 2 (HONAM) + + + + A Proximity to Metropolitan * • PA Cities International Competition + + + R Power of Commodities

(+)=Positive; (-)=Negative; (.)=Weak; (A)=Accepted; (PA)=Partly Accepted; (M)=Mixed; (Dl)=Total Official Poverty Population; (D2)=Protected at Home Population; (D3)=Self-supporting Low Income Population 140 In the non-farm characteristics, percent non-farm population had consistently positive effects on rural pover­ ty. These results were contradictory to the research hy­ pothesis. Percent female had positive effects on total official poverty population and self-supporting low income population, while a negative effect on protected at home population. Percent high school graduates had consistently negative effects on rural poverty, as hypothesized. In the spatial/political economical characteristics, Region 1 (YOUNGNAM) had a positive effect on protected at home population, while a negative effect on self-supporting low income population. Also Region 1 (YOUNGNAM) had only a weak effect on total official poverty population. Region 2 (HONAM) had consistently positive effects on rural poverty, as hypothesized. Proximity to metropolitan cities had a negative effect on self-supporting low income population, while only weak effects on the other two dependent vari­ ables. International competition power of commodities had consistently positive effects on rural poverty. These results were contradictory to the research hypothesis. In sum, the research hypotheses on four independent variables (percent middle farming, percent part-time farm­ ing, and percent high school graduates, HONAM) were accept­ ed. Also, the research hypotheses on another three indepen­ dent variables (rate of tenancy, percent female, and proxim­ ity to metropolitan cities) were partly accepted. The 141 research hypotheses on three independent (percent larger farming, percent non-farm population, and international competition power of commodities) were rejected. The re­ search hypothesis on Region 1 presented mixed results.

) (Blocked Regression)

In the case of blocked regression analysis, the vari­ ances of three rural poverty scales (total official poverty population, protected at home population, and self-support­ ing low income population) explained by the block of farm structure/characteristics were 30 percent, 20 percent, and 30 percent, respectively. The variances of the three rural poverty scales added by the block of non-farm character­ istics were 28 percent, 10 percent, and 23 percent, respec­ tively. The variances of the three rural poverty scales added by spatial/political economical characteristics were 13 percent, 7 percent, and 11 percent, respectively. That is, each block added a significant amount of explained variance. However, the relationships between farm structure/characteristcs and rural poverty were not greatly changed when non-farm characteristics and spatial/ political economical characteristics were controlled, although the magnitude of the coefficients were changed a little bit. 142 (Regional Comparisons)

The amounts of explained variance for total official poverty population were comparatively high (from 63 percent to 76 percent) across the three regions. In general, the patterns of relationships between the independent variables and rural poverty presented variations across the regions. The relationships between the independent variables and rural poverty presented variations across the sub-dimensions of rural poverty. Table 7.2 presents a summary of the regression results for total official poverty population by region. The re­ search hypotheses on six independent variables (percent middle farming, percent part-time farming, rate of tenan­ cy, percent female, percent high school graduates, and proximity to metropolitan cities) were accpted across the three regions. The research hypotheses on three independent variables (percent larger farming, percent non-farm popula­ tion, and international competition power of commodities) were rejected across the three regions. Table 7.3 presents a summary of the regression results for protected at home population by region. The research hypotheses on two independent variables (percent middle farming and percent high school graduates) were accepted across the three regions. The research hypotheses on per­ cent part-time farming was partly accepted. The research 143 hypotheses on percent non-farm population was partly reject­ ed. The research hypotheses on five independent variables (percent larger farming, rate of tenancy, percent female, proximity to metropolitan cities, and international competi­ tion power of commodities) presented mixed results. That is, seven independent variables presented some variations by region. Only percent middle farming and percent high school graduates did not. Table 7.4 presents a summary of the regression results for self-supporting low income population by region. The hypotheses on five independent variables (percent middle farming, percent part-time farming, rate of tenancy, percent female, and percent high school graduates) were accepted across the three regions. The research hypothesis on three independent variables (percent larger farming, percent non­ farm population, international competition power of commodi­ ties) were rejected across the three regions. The research hypothesis on proximity to metropolitan cities presented mixed results. Table 7.2. Regression of Total Official Poverty Population by Region (Summary)

Hypothesized Empirical Findings Test Independent Variables Effects with Results Poverty R1 R2 R3 Farm Structure/Characteristics

Percent Middle Farming —— — — A Percent Larger Farming + ——— R Percent Part-time Farming ——— — A Rate of Tenancy + + + + A Non-farm Characteristics

Percent Non-farm Population — + + + R Percent Female + + + + A Percent High School Graduates ——— — A Spatial/Political Economical Characteristics Proximity to Metropolitan Cities ———— A International Competition Power of Commodities + + + R

(+)=Positive; (-)=Negative; (A)=Accepted; (R)=Rejected; (Rl)=Youngnam; (R2)=Honam; (R3)=Others Table 7.3. Regression of Protected at Home Population by Region (Summary)

Hypothesized Empirical Findings Test Independent Variables Effects with Results Poverty Rl R2 R3 Farm Structure/Characteristics

Percent Middle Farming ——_ — A Percent Larger Farming + — + — M Percent Part-time Farming — — • — PA Rate of Tenancy + — — + M Non-farm Characteristics

— Percent Non-farm Population • + + PR Percent Female + + + — M Percent High School Graduates A Spatial/Political Economical Characteristics Proximity to Metropolitan Cities ““ + M International Competition Power of Commodities + + M

(+)=Positive; (-)=negative; (.)=Weak: (A)=Accepted; (PA)=Partly Accepted; (PR)=Partly Rejected; (M)=Mixed; (Rl)=Youngnam; (R2)=Honam; (R3)=0thers Table 7.4. Regression of Self-supporting Low Income Population by region (Summary)

Hypothesized Empirical Findings Test Independent Variables Effects with Results Poverty Rl R2 R3 Farm Structure/Characteristics

Percent Middle Farming ———— A Percent Larger Farming + — — — R Percent Part-time Farming ———— A Rate of Tenancy + + + + A Non-farm Characteristics

Percent Non-farm Population — + + + R Percent Female + + + + A Percent High School Graduates A Spatial/Political Economical Characteristics Proximity to Metropolitan Cities mmm + M International Competition Power of Commodities + + + R

(+)=Positive; (-)=Negative; (A)=Accepted; (R)=Rej ected; (M)=Mixed; (Rl)=Youngnam; (R2)=Honam; (R3)=0thers 147 Policy Implications

Based on the overview of historical trends of rural poverty related indicators and the multiple regression analyses, and the review of previous poverty studies, some policy implications and alternatives to help reduce and alleviate poverty in the rural Korea are suggested. First, historically Korean rural development policies have been mainly top-down approaches. As a result, the needs and demands of local people have not been adequately reflected in the planning and implementation processes of development projects. The results of this dissertation indicate that factors associated with poverty rates are not uniform across rural Korea. Therefore, we need people- centered, bottom-up rural development policies to alleviate poverty in rural Korea. People centered development re­ quires basic reforms in political and administrative struc­ tures which move the locus of development initiative from the central government to the people (Korten, 1984; Maeda, 1981). Second, the rural sector of Korea has been treated as a production field to provide cheap food in the national development processes. This is wrong. Today, rural prob­ lems are no longer synonymous with farm problems. For exam­ ple, according to the finding of this study, the average proportion of non-farm population in rural Korean counties was 51.4 percent. Therefore, policies to alleviate rural poverty should be focused on non-farm sectors as well as the farm sector. In this study, percent part-time farming had consistently negative effects on rural poverty. Thus it is necessary to enlarge the opportunities of off-farming jobs. The rural industrial complex program which was intro­ duced by the Korean Government in 1980s, was a plan to enlarge the opportunities off-farming jobs. But the program was not successful. Most jobs in the rural industrial complex were low wage jobs for unskilled labor force (Chung and Oh, 1992). The amount of poverty in a community is conditioned by the quantity and quality of jobs available (Tomaskovic-Devey, 1988). The enlargement of low wage jobs in the rural industrial complex is not a real solution to alleviate rural poverty. It just enlarges the number of working poor. Therefore, it is important to consider the types and quality of jobs in developing rural industries. According to Lobao (1990), in the long run diverse local firms contribute to protecting communities from macro- economic instability. Local firms should be based on local characteristics and resources. Third, it is very important to provide the structural opportunities for the upward mobility of the rural poor. That is, it is necessary to strengthen institutional mecha­ nisms in favor of the rural poor. In this study, percent middle farming and percent larger farming had consistently 149 negative effects on the three rural poverty scales. That is, the scale of farming had reverse relationships with rural poverty. It was contradictory to the Goldschmidt hypothesis that increased farm scale negatively affects the quality of life in rural communities. Therefore, it is necessary to enlarge the financial support to help small­ sized farmers expand their farms. According to this study, in general, the rate of land tenancy presented positive effects on rural poverty. Thus, it will be important to lower the rate of land tenancy in alleviating rural poverty. In rural Korea, rent for land tenancy varies according to the quality and location of land. However, the rent for land tenancy has been very high (40-50 percent of products). Also, most land tenants are small-sized family farmers, not commercially oriented farmers (Chang, 1988; Chung and Oh, 1992) . Therefore, the rent for land tenancy should be lowered in a resonable level. In addition, current land tenancy regulations should be revised. The cultivation right of tenants should be protected in the form of laws. Korean Government should try to return land ownership to real farmers, who actually farming in the land. Therefore, the speculation of land by non-farmers should be prohibited strictly. Fourth, until now, the agricultural policies of the Korean Government have been large-scale farm oriented. The current agricultural policies of the Korean Government 150 emphasizes the larger farming system and mechanization in order to increase international competition power of commod­ ities. As a result, the financial support of the Korean government to help farmers expand their farms has been in favor of large-scale farmers. This may represent another policy mistake of the Korean Government. As presented in Chapter V, the general level of international competition power of most agricultural commodities are very low. There­ fore, the enlargement of farm size is not enough to improve the socio-economic conditions of rural communities. The Korean Government has considered 10 ha as a lower limit of adequate farm size to compete internationally. However, the enlargement of farm size in about 10 ha may not be useful to improve the international competition power of commodities. That is, 10 ha in farm size is still very small compared with major countries (e.g. U.S.) exporting agricultural commodities. The geographical characteristics of Korea may constrain the mechanization of farming. Therefore, the Korean Government needs to develop new agricultural items which can survive and compete internationally. Also, the Korean Government should enlarge investment to improve the level of agricultural technologies. Fifth, in this study, relationships between the independent variables and rural poverty had some variations across the three regions. In YOUNGNAM, significant vari­ ables to explain rural poverty were mainly farm structure/ 151 characteristics and proximity to metropolitan cities. In HONAM, significant variables to explain rural poverty were mainly spatial/political economical characteristics. In OTHERS, significant variables to explain rural poverty were mainly non-farm characteristics, such as percent non­ farm population. Therefore, the solution of rural poverty in Korea will require locally specific policy packages. In particular, HONAM presented the highest level of rural poverty compared to the other regions. This situation can be compared with the situation of the South in the U.S.. But, unlike the South in the U.S., persistent regional socio-economic differences in Korea have resulted mainly from the discrimination and intervention of the Korean Government. In Korea, the state has had more significant effects on the socio-economic in regions than the logic of capital to extract maximun surplus. That is, the Korean Government has used various unbalanced (uneven) development strategies as major means to keep political power. There­ fore, it is very important to enlarge the political autonomy of regional government in lowering regional differences and rural poverty. Sixth, the Korean Government should specify poverty policies depending on different types of the poor. Attempts to tackle rural poverty should be made in more sensitive manners to the needs of different categories of rural poor people. The working poor (e.g. self-supporting low income population) should be considered differently from the dis­ abled poor (e.g. protected at home population). For the working poor, it is important to set a higher minimum wage and to improve their labor market power by providing the opportunities of training and higher education. Also, the prices of agricultural commodities should be guaranteed in reasonable levels through various institutional supporting systems. For the disabled poor, a basic needs approach may be necessary. Basic needs refer to the minimum standard of living which a society should set for the poorest groups of it people. Basic needs approach seeks to relieve as quickly as is possible absolute poverty through intensive direct assistance to those in desperate circumstances (Streenten, 1981; Arndt, 1987). Finally, in Korea, the protection level of welfare program has been very low and unrealistic. According to Chung and Choi (1990:68), the protection level of welfare program for official poverty population was 27.7 percent to 58 percent of minimum living cost in 1989. Therefore, it is necessary to adjust the protection level of welfare programs to meet the minimum standard of living. According to Chung and Oh (1992:113), the public welfare budget of Korea was 1.5 percent of the national budget in 1988. That is, the level of pubulic welfare budget of Korea has been very low. Therefore, the Korean government should enlarge the public welfare budget, step by step. What is needed in order to 153 alleviate poverty in rural Korea is welfare reform.

Limitations of the Research

There were several limitations in this study. First, the use of county-level secondary data constrained the selection of variables. For example, most counties had no data on industry structure. As a result, this study could not adequately test the applicability of the dual (segment­ ed) economy theory on rural poverty. This study was origi­ nally designed to examine two points in time: 1980 and 1990. However, in the case of 1980 data, many rural poverty relat­ ed variables were not available on the county level. As a result, this study used a cross-sectional design in the multiple regression analysis. Second, another limitation of this study was the reli­ ability problem of official data. The official statistics of poverty tended to be different from the real conditions of poverty (Chung and Oh, 1992). That is, the official statistics of poverty may be underestimated values of the real poor. Third, the three rural poverty indicators had some missing cases. In particular, two sub-dimensions of total official poverty population (protected at home population and self-supporting low income population) had many missing cases (14 and 16, respectively). These missing cases may 154 affect the relationships between the independent variables and rural poverty. However, based on additional analysis of missing cases, the counties with missing data were not significantly different from the counties without missing data on indicators of quality of rural communities, such as water supply and medical care. Therefore, it is possible to infer that missing cases did not affect the relationship between the independent variables and rural poverty. Finally, this study had some theoretical limitations. This study tried to combine the ideas obtained from various theories of poverty and inequality, such as the Goldschmidt hypothesis, dependency/world system theory, and dual economy theory. As a result, this study may have lacked some coher­ ence of theoretical reasoning.

Recommendations for Future Research

First of all, future research should develop and test a finer measurement of rural poverty. One way to do this is to measure the minimum living costs within each region. Also it will be very important to collect the reliable data on income in rural communities. That is, future research is required to measure the level of rural poverty by using real data of incomes and minimum living costs depending on each region. Future research needs to conduct a complete update of consumption patterns "to help our poverty measurements 155 keep pace with changing social and economic norm" (Ruggles, 1990:xiv). Poverty measures are sensitive to the specifi­ cation of the accounting period (Ruggles, 1990). Thus, future research needs to develop spell (i.e., period) ori­ ented measures of income and poverty. Also, future research should investigate the relative and subjective aspects of poverty for a comprehensive understanding on rural poverty. Second, future research should use longitudinal designs to investigate rural poverty. A longitudinal design will permit testing of the historical dynamics of poverty. In a longitudinal design, qualitative research methods, such as participant observation, may be necessary in order to investigate the reality of rural poverty from an insider's perspective. A more localized case study will be useful to find out the political economical processes of rural poverty in a region. Third, future research should develop a more coherent theoretical framework to explain the causes of poverty. For example, aspects of various structural perspectives of rural poverty can be integrated into a more generalized political economical approach to poverty. Also, future study should develop theoretical strategies for linking micro and macro processes in explaining rural poverty (Turner, 1983). That is, future study should more systematically specify and articulate the various levels of analysis (individual, household, local, regional, national, and international). 156 Finally, future research should incorporate comparative research designs. It would be useful to compare the study results of one country with those of other countries. In particular, comparisons among third world countries are important. The comparative research of rural poverty will require the cooperation of social scientists from various countries. In comparative research of rural poverty across different countries, socio-cultural aspects, such as reli­ gions and custom, should be considered carefully. NOTES

1) In 1990, the criteria to select official poverty popu­ lation were two: monthly income per capita (48,00 won, that is, about $ 60) and property worth per household (3,200,000 won— about $ 4,000). 2) However, Cardoso and Faletto (1979) has used dependency theory to examine differences within a particular national context. 3) According to Blalock (1972:330), "the notion of orthogonality derives from a geometrical interpretation of statistical associations and refers to situations in which the relationships can be represented by means of perpendicu­ lar or orthogonal axes."

157 BIBLIOGRAPHY

Albrecht, Don E. and Steve H. Murdock (1984), "Toward a human ecological perspective on part-time farming." Rural Sociology. 49 (3):389-411. Albrecht, D. E. and J. K. Thomas (1986), "Farmtenure: a retest of conventional knowledge." Rural Sociology. 51 (1): 18-30. Anderson, C. A. and Bryce Ryan (1943), "Social participation differences among tenure classes in a commercialized farming area." Rural Sociology. 8 (3):281-290. Arndt, H. W. (1989), Economic Development: The History of an Idea. Chicago: The University of Chicago Press. Banfield, E. C. (1968), The Unheavenlv Citv: The Nature and Future of Our Urban Crisis. Boston: Little, Brown and Company. Barlett, Peggy F. (1986), "Part-time farming: saving the farm or saving the life-style." Rural Sociology. 51 (Fall): 289-313. Baron, James N. and William T. Bielby (1980), "Bringing the firms back in: stratification, segmentation, and the organization of work." American Sociological Review.45 (October):737-765. Beck, E. M., Patrick M. Horan, and Charles M. Tolbert II (1978), "Stratification in a dual economy: a sectoral model of earning determination." American Sociological Review. 43 (October):704-720. Becker, G. S. (1964), Human Capital. New York: Columbia University Press. Beeghley, Leonard (1984), "Illusion and reality in the measurement of poverty." Social Problems. 31 (3):322-333. Beeghley, Leonard (1988), Living Poorly in America. New York: Praeger.

158 159 Bergesen, Albert (1990), "Turning world-system theory on its head." Theory. Culture and Society. 7:67-81. Bertrand, Alvin L. (1967), "Research on part-time farming in the United States." Socioloqia Ruralis. 7 (3):295-306. Blalock, Hubert M. (1972), Social Statistics. New York: McGraw-Hill Book Company. Blau, Peter and Otis D. Duncan (1967), The American Occupational Structure. New York: Wiley. Bloomguist, Leonard E. (1988), "Rural manufacturing gets mixed reviews." Rural Development Perspectives. 4 (3):22-26. Bloomquist, Leonard E. and Gene F. Summers (1982), "Organization of production and community income distributions." American Sociological Review. 47 (June): 325-338. Blueston, Barry and Bennett Harrison (1982), The Deindustrialization of America. New York: Basic Books. Bohrnstedt, G. W. and David Knoke (1982), Statistics for Social Data Analysis. Peacock Pubs. Brenner, R. (1977), "The origins of capitalist development: a critique of neo-Smithian Marxism." New Left Review. 104: 25-92. Burmeister, Larry L. (1990), "State, industrialization and agricultural policy in Korea." Development and Change. 21:197-223. Buttel, Frederick H. and Pierre LaRamee (1991), "The disappearing middle: a sociological perspective." Pp. 151-169 in W. H. Friedland, L. Busch, F. H. Buttel, and A . P . Rudy (eds.), Towards a New Political Economy of Agriculture. Boulder: Westview Press. Cardoso, Fernando H. and Enzo Faletto (1979), Dependency and Development in Latin America. Berkeley: University of California Press. Cernea, Michae M. (1985), "Sociological knowledge for development projects." Pp. 3-22 in M. Cernea (ed.), Putting People First. New York: Oxford University Press. Chang, S. H. (1988), "The characteristics of current land problems and solutions." (in Korean). Pp. 109-191 in Korea Rural Studies Institute (ed.), The Study of Korean Agriculture and Farmers I. Seoul, Republic of Korea: 160 Yeongusa. Chang, S. H. (1991), "The development of Korean capitalism and social changes of rural society." (in Korean). Pp. 11-75 in Korea Rural Studies Institute (ed.), Korean Capitalism and Rural Society. Seoul, Republic of Korea: Society and Culture Studies Institute. Cheju-do (1990), Cheiu Statistical Yearbook. Cheju-do, Republic of Korea. Chirot, Daniel and Thomas D. Hall (1982), "World-system theory." Annual Review of Sociology. 8:81-106. Choi, I. S. and Others (1989), The Actual Conditions of Low Income Population and Counter-measures in Seoul Citv (in Korean). Seoul, Republic of Korea: The Korean Institute for Industrial Economies and Technology. Chonbuk Daily Newspaper Co. (1990), Chonbuk Yearbook (in Korean). Chonbuk Daily Newspaper Co., Republic of Korea. Chonlabuk-do (1990), Chonbuk Statistical Yearbook. Chonlabuk-do, Republic of Korea. Chonlanam-do (1990), Chonnam Statistical Yearbook. Chonlanam-do, Republic of Korea. Chungchongbuk-do (199 0), Chunqbuk Statistical Yearbook. Chungchongbuk-do, Republic of Korea. Chungchong Daily Newspaper Co. (1991), Chunqbuk Yearbook (in Korean). Chungchong Daily Newspaper Co., Republic of Korea. Chungchongnam-do (1990), Chunqnam Statistical Yearbook. Chungchongnam-do, Republic of Korea. Chung, D. Y. (1985), Urban Poverty Study (in Korean). Seoul, Repubulic of Korea: Achim. Chung, Gun Shik (1991), "Regional emotion and regional problem." Pp. 521-546 in Y. B. Koh (ed.), Contemporary Social Problems (in Korean). Seoul, Republic of Korea: Research Institute of society and Culture. Chung, Ki-Whan and Nae-Won Oh (1992), "Rural poverty in the Republic of Korea: trends and policy issues." Asian Development Review. 10 (1):91-124. Chung, M. C., S. K. Min, K. H. Choi and Y. D. Lee (1989), The Living and Well-being Conditions of Small-income Farmers (in Korean). Research Monograph 203. Seoul, Republic of 161 Korea: Korea Rural Economic Institute. Chung, M. C. and K. H. Choi (1990), A Study on the Direction of Rural Social Policy to Improve Agricultural Structure (in Korean). Research Monograph 227. Secoul, Republic of Korea: Korea Rural Economic Institute. Colclough, Glenna (1988), "Uneven development and racial composition in the deep South." Rural Sociology. 53:73-86. Curran, Sara and Donald Tomaskovic-Devey (1991), "Uneven development in North Carolina?: job quality differences between local and nonlocal firms." Rural Sociology,. 56 (4):680-698. Davis, Kingsley and Wilbert E. Moore (1945), "Some principles of stratification." American Sociological Review. 10 (April):242-249. Dunford, M. F. (1988), Capital, the State, and Regional Development. Brodesbury Park, London: Pion Limited. Eckert, C. J., K. B. Lee, Y. I. Lew, Michael Robinson and E. W. Wagner (1990), Korea Old and New: A History. Seoul, Republic of Korea: Ilchokak, Publishers. Economic Planning Board (1991 and various years), Annual Statistics in Korea. Economic Planning Board, Repulic of Korea. Evans, Peter (1979), Dependent Development: The Alliance of Multinational. State, and Local Development in Brazil. Princeton, M.J.: Princeton University Press. Farmer, F. L., T. W. Ilvento, and A. E. Luloff (1989), "Rural community poverty: a LISREL measurement model." Rural Sociology. 54 (4): 491-508. Flinn, William L. and Frederick H. Buttel, (1980), "Sociological aspects of farm size: ideological and social consequences of scale in agriculture." American Journal of Agricultural Economics. 62 (5):946-953. Flora, C. B., J. L. Flora, J. D. Spears, L. E. Swanson, M. B. Lapping, and M. L. Weinberg (1992), Rural Communities: Leaacv & Change. Boulder: Westview Press. Foster-Carter, Aidan (1985), The Sociology of Development. Ormskirk, Lancashire: Causewey Books. Foucault, M. (1979), Discipline and Punish. New York: Vintage. 162 Fox, John (1991) Regression Diagnostics. Newbury Park: Sage. Frank, A. G. (1969), Capitalism and Underdevelopment in Latin America. New York: Monthly Review Press. Frank, Walter (1983), "Parttime farming, underemployment and double activity of farmers in the EEC." Sociologia Ruralis. 23 (1):20-27. Frauendorfer, S. (1966), "Part-time farming: a review of world literature." World Agricultural and Economics/Rural Sociology Abstracts. 8 (l):5-37. Fuguitt, G. V., A. M. Fuller, H. Fuller, R. Gasson, and G. Jones (1977), Part-time Farming; Its Nature and Implications. Ashford, Kent, England: Wye College. Fuller, Anthony M. (1984), "Part-time farming: the enigmas and the realities." Pp. 187-219 in Harry K. Schwarzweller (ed.), Research in Rural Sociology and Development. Volume I. Greenwich, Conn.: JAI Press. Gans, Herbert (1972), "The positive functions of poverty." American Journal of Sociology. 78 (2):275-289. Gasson, Ruth (1986), "Part time farming strategy for survival?" Sociologia Ruralis. 24 (3/4):364-376. Giddens, A. (1984), The Constitution of Society. Cambridge: The Polity Press. Gilles, Jere Lee and Michael Dalecki (1988), "Rural well­ being and agricultural changes in two farming regions." Rural Sociology. 53 (Spring):40-55. Gimenez, Martha E. (1990), "The feminization of poverty: myth or reality." Social Justice, 17 (3):43-59. Goldman, Robert and Ann Tickamyer (1984), "Status attainment and the commodity form: stratification in historical perspective." American Sociological Review. 49 (April): 196-209. Goldschmidt, Walter (1978a), As You Sow: Three Studies in the Social Conseguences of Agribusiness. Montclair, N.J.: Allanheld, Osmum and Company. Goldschmidt, Walter (1978b), "Large-scale farming and the rural social structure." Rural Sociology. 43 (Fall): 362-366. Goodwin, L. (1972), Do the Poor Want to Work?. Washington, 163 D. C.: Brookings institute. Gorham, Lucy (1992), "The growing problem of low earnings in rural areas." Pp.21-39 in Cynthia M. Duncan (ed.), Rural Poverty in America. New York: Auburn House. Gottdiener, M. (1991), "Space, social theory, and the urban metaphor." Current Perspectives in Social Theory. 11: 295-311. Green, Gary P. (1985), "Large-scale farming and the quality of life in rural communities: further specification of the Goldschmidt hypothesis." Rural Sociology. 50 (Summer): 262-273. Hansen, David 0. and Archibald 0. Haller (1973), "Status attainment of Costa Rican males: a cross-cultural test of a model." Rural Sociology. 38 (3):269-282. Harris, Craig and Jess Gilbert (1982), "Large-scale farming, rural income, and Glodschmidt's agrarian thesis." Rural Sociology. 47 (Fall):449-458. Harvey, David (1982), The Limits to Capital. Chicago: University of Chicago Press. Hauser, Robert M. and David L. Featherman (1977), The Process of Stratification: Trends and Analysis. New York: Academic Press. Hayes, Michael N. and Alan L. Olmstead (1984), "Farm size and community quality: Arvin and Dinuba revisited." American Journal of Agricultural Economics. 66 (November):430-436. Heaton, Tim B. (1980), "Metropolitan influence on United States farmland use and capital intensivity." Rural Sociology. 45 (Fall):501-508. Heffernan, William D. (1972), "Sociological dimensions of agricultural structures in the United States." Sociologia Ruralis. 12 (October):481-499. Heffernan, W. D., G. Green, R. P. Lasley, and M. F. Nolan (1981), "Part-time farming and the rural community." Rural Sociology. 46 (3):245-262. Hirsch, Eric (1980), "Dual labor market theory: a sociological critique." Sociological Inguiry. 50 (2): 133-145. Hoare, James and Susan Pares (1988), Korea: An Introduction. New York: Kegan Paul International. 164 Hodson, Randy (1984), ’’Companies, industries and the measurement of economic segmentation." American Sociological Review. 49 (June):335-348. Hodson, Randy D. and Robert L. Kaufman (1982), "Economic dualism: a critical review." American Sociological Review. 47 (December):727-739. Holman, R. (1978), Poverty: Explanations of Social Deorivation. New York: St. Martin's Press. Hong, Dong Shik (1988), Understanding of Rural Sociology (in Korea). Seoul, Republic of Korea: Bubmunsa Publishing Co. Hong, G. Y. (ed.) (1986), The Actual Conditions of Urban Poverty and Policy Study. Seoul, Republic of Korea: Danguk University Press. Horan, Patrick M. (1978), "Is status attainment research a theoretical?11 American Sociological Review. 43 (August):534-541. Howe, Gary M. and Alan M. Sica (1980), "Political economy, imperialism, and the problem of world system theory." Current Perspectives in Social Theory. 1:235-286. Howes, Candace and Ann R. Markusen (1981), "Poverty: a regional political economy perspective." Pp. 437-463 in Amos H. Hawley and Sara Mills Mazie (eds.), Nonmetropolitan America in Transition. Chapel Hill: The University of North Carolina Press. Huaco, George (1986), "Ideology and general theory: the case of sociological functionalism." Comparative Studies in Society and History. 28:34-54. Jacobs, David (1982), "Unequal organizations or unequal attainments? An empirical comparison of sectoral and individualistic explanations for aggregate inequality." American Sociological Review. 50 (April):166-180. Janowitz, M. (1977), "A sociological perspective on Wallerstein." American Journal of Sociology. 82 (5): 1090-1097. Ju, H. J., H. J. Chang, I. S. Choi, and G. B. Lee (1989) The Actual Conditions of Low Income Population and Measures for the Stabilization of Livelihood fin Korean). Seoul, Republic of Korea: Korea development Institute. Kabeer, Naila (1991), "Gender dimensions of rural poverty: analysis from Bangladesh.11 The Journal of Peasant Studies. 165 18 (2):241-262. Kalleberg, Arne L., Michael Wallace, and Robert P. Althauser (1981), "Economic segmentation, worker power, and income inequality." American Journal of Sociology. 87 (June): 651-683. Kangwon Daily Newspaper Co. (1990), Kanowon Yearbook (in Korean). Kangwon Daily Newspaper Co., Republic of Korea. Kangwon-do (1990), Kanowon Statistical Yearbook. Kangwon-do, Republic of Korea. Kaufman, Robert L., Randy Hodson, and Neil D. Fligstein (1981), "Defrocking dualism: a new approach to defining industrial sectors." Social Science Research. 10 (March): 1-31. Kenney, Martin, Linda M. Lobao, James Curry, and W. Richard Goe (1989), "Midwestern agriculture in U.S. fordism: from the New Deal to economic restructuring." Sociologia Ruralis. 29 (2):130-148. Kim, Dong-11 (1980), "A profile of Korean rural villages, farmers and their changing quality of life." Journal of Rural Development. 3 (December):193-212. Kim, Y. S. (1985), The Theory of the Urban Poor (in Korean). Seoul, Republic of Korea: Achim Publication Co. Knottnerus, J. D. (1987), "Status attainment research and its image of society." American Sociological Review. 52 (February):113-121. Kolko, Joyce (1988), Restructuring the World Economy. New York: Pantheon Books. Korean Statistical Association (1991 and various years), Social Indicators in Korea. Seoul, Republic of Korea: The Statistical Association. Korten, David C. (1981), "Social development: putting people first." Pp. 201-222 in D. C. Korten and F. B. Alfonso (eds.), Bureaucracy and the Poor: Closing the Gap. New York: McGraw-Hill. Korten, David C. (1984), "People-centered development: toward a framework." Pp. 299-309 in D. C. Korten and R. Klauss (eds.), People Centered Development. Kumarian. Korten, David C. (1987), "Third generation NGO strategies: a key to people centered development." World Development. 166 15:145-159. Krymkowski, Daniel H. (1991), "The process of status attainment among men in Poland, the U. S., and West Germany." American Sociological Review 56 (February):46-59. Ku, Dae-Yeol (1985), Korea Under Colonialism: The March First Movement and Anqlo-Jaoanese Relations. Seoul, Republic of Korea: Royal Asiatic Society Korea Branch. Kyonggi-do (1990), Kvonqqi Statistical Yearbook. Kyonggi-do, Republic of Korea. Kyong-In Daily Newspaper Co. (1990), Kvonqqi Yearbook (in Korean). Kyong-In Daily Newspaper Co., Republic of Korea. Kyongnam Daily Newspaper Co. (1990), Kvonqnam Yearbook (in Korean). Kyongnam Daily Newspaper Co., Republic of Korea. Kyongsangbuk-do (1990), Kvonqbuk Statistical Yearbook. Kyongsangbuk-do, Republic of Korea. Kyongsangnam-do (1990), Kyongnam Statistical Yearbook. Kyongsangnam-do, Republic of Korea. Lee, H. J. and S. Y. Ho (1988), The Third World Urbanization and Poverty (in Korean). Seoul, Republic of Korea:Hangilsa. Lewis, Oscar (1968), La Vida. New York: Vintage. Levine, Rhonda F. (1985), "Marxism, sociology, and neo- Marxist theories of the state." Current Perspectives in Social Theoryr 6:149-167. Lichter, Daniel T. and Janice A. Costanzo (1987), "Nonmetropolitan underemployment and labor-force composition." Rural Sociology. 52 (3):329-344. Lim, C. H., M. G. Lee and J. T. Park (1989), A Study on the Countermeasures for Urban Poverty Population (in Korean). Seoul, Republic of Korea: Korea Development Institute. Lim, H. J. (1987), Contemporary Korea and Dependency Theory (in Korean). Seoul, Republic of Korea: Seoul National University Press. Lim, H. S. (1979), "The poverty problem in the Republic of Korea." (in Korean) Pp. 71-87 in S. J. Hong et al. (eds.), Korean Social Development Studies I. Seoul, Korea: Asian Research Center. Lobao, Linda M. (1990), Locality and Inequality: Farm and 167 Industry Structure and Socioeconomic Conditions. Albany: State University of New York Press. Lobao, Linda M. and Michael D. Schulman (1991), "Fanning patterns, rural restructuring, and poverty: a comparative regional analysis." Rural Sociology. 56 (4):565-602. London, Bruce and David A. Smith (1988), "Urban bias, dependence, and economic stagnation in noncore nations." American Sociological Review, 53 (June):454-463. Long, Norman (1980), An Introduction to the Sociology of Rural Development. London: Tavistock Publications. Lovejoy, S. B. and R. S. Krannich (1982), "Rural industrial development and domestic dependency relations: toward an integrated perspective." Rural sociology. 47: 475-95. Maeda, Justin H. J. (1981), "Creating national structures for people centered agrarian development." Pp. 136-162 in D. C. Korten and F. B. Alfonso (eds.), Bureaucracy and the Poor: Closing the Gap. New York: McGraw-Hill. Mandel, Ernest (1978) Late Capitalism. London: Versa. Mann, Susan A. and James M. Dickinson (1978), "Obstacles to the development of a capitalist agriculture." Journal of Peasant Studies. 5 (July):466-481. Marklund, staffan (1990), "Structures of modern poverty." Acta Sociologia. 33 (2):125-140. Markusen, Ann (1987), Regions: The Economics and Politics of Territory. Totowa, N.J.: Rowman and Littlefield. Marousek, Gerald (1979), "Farm size and rural communities: some economic relationships." Southern Journal of Agricultural Economics, 2 (December):57-61. Marx, Karl (1867/1967), Capital: A Critigue of Political Economy. Vol.l. New York: International Publishers. Mark, Karl and Friedrich Engels (1848/1948), Manifesto of the Communist Party. New York: International Publishers. Mayer, Tom (1991), "On Marxism." Pp. 98-123 in Henry Etzkowitz and Ronald M. Glassman (eds.), The Renascence of Sociological Theory. Itasca, Illinois: F.E. Peacock Publishers, Inc. McGranahan, David A. (1980a), "Changing central place activities in Northwestern Wisconsin." Rural Sociology. 45 168 (Spring):91-109. McGranahan, David A. (1980b), "The spatial structure of income distribution in rural regions." American Sociological Review. 45 (April):313-324. McMichael, Philip and F. H. Buttel (1990), "New directions in the political economy of agriculture." Sociological Perspectives. 33 (1): 89-109. Midgley, James (1988), "Inequality, the third world and development." International Journal of Contemporary Sociology. 25 (3-4):93-103. Ministry of Agriculture, Forestry and Fisheries (1991a), Counter-measures to Improve International Competition Power of Agriculture and Stockbreeding Products Item bv Item (in Korean). Seoul, Republic of Korea: The Ministry of Agriculture, Forestry and Fisheries. Ministry of Agriculture, Forestry and Fisheries (1991b and various years), 1990 Agricultural Census. Ministry of Agriculture, Forestry and Fisheries, Republic of Korea. Ministry of Agriculture, Forestry and Fisheries (1991c and various years), Annual Report on the Farm Household Economy Survey. Ministry of Agriculture, Forestry and Fisheries, Republic of Korea. Ministry of Agriculture, Forestry and Fisheries (199Id and various years), Annual Statistics of Agriculture. Forestry and Fisheires. Ministry of Agriculture, Forestry and Fisheries, Republic of Korea. Ministry of Health and Social Affairs (1990, 1991 and various years), Current Conditions of Livelihood Protection Population (in Korean). Seoul, Korea: The Ministry of Health and Social Affairs. Mohr, Lawrence B. (1990), Understanding Significance Testing. Newbury Park: Sage. Mooney, Patrick (1982), "Labor time, production time, and capitalist development in agriculture: reconsideration of the Mann-Dickinson Thesis." Sociologia Ruralis. 12 (3-4): 279-292. Mooney, Patrick (1983), "Toward a class analysis of Midwestern agriculture." Rural Sociology. 48 (4):563-584. Mooney, Patrick (1987), "Desperately seeking: one­ dimensional Mann and Dickinson." Rural Sociology. 52 (2): 169 286-295. Nie, N. H., C. H. Hull, J. G. Jenkins, K. Steinbrenna, and D. H. Bent (1975), SPSS (2nd ed.). New York: MaGraw-Hill. Nuckton, Carole Frand, Refugio I. Rochin, and Douglas Gwynn (1982), "Farm size and rural community welfare: an interdisciplinary approach." Rural Sociology. 47 (Spring): 32-46. O'Hare, William P. (1988), The Rise of Poverty in Rural America. Occasional paper Number 15, July. Washington, D.C.: The Population Reference Bureau, Inc. Peacock, Walter G., Greg A. Hoover and Charles D. Killian (1988), "Divergence and convergence in interantional development: a decomposition analysis of inequality in the world system." American Sociological Review. 53 (December): 838-852. Pfeffer, Max (1983), "Social origins of three systems of farm production in the United States." Rural Sociology. 48 (Winter):540-562. Pfeffer, Max (1989), "The feminization of production on part-time farms in the Federal Republic of Germany." Rural Sociology. 54 (l):60-73. Pugliese, Enrico (1991), "Agriculture and the new division of labor." Pp. 137-150 in W. H. Friedland, L. Busch, F. H. Buttel, and A. P. Rudy (eds.), Towards a New Political Economy of Agriculture. Boulder: Westview Press. Rabow, Jerome, S. L., Berkman and Ronald Kessler (1983), "The culture of poverty and learned helplessness: a social psychological perspective." Sociological Inguirv. 53 (4):419-434. Rankin, B. H. and W. W. Falk (1991), "Race, region, and earnings: blacks and whites in the South." Rural Sociology. 56 (2):224-237. Reif, Linda Lobao (1987), "Farm structure, industry structure, and socioeconomic conditions in the United States." Rural Sociology. 52 (Winter):462-482. Ritzer, George (1988), Sociological Theory (2nd. ed.). New York: Alfred A. Knopf. Rogers, Everett M., Rabel J. Burdge, Peter F. Korsching, and Joseph F. Donnermeyer (1988), Social Change in Rural Societies: An Introduction to Rural Sociology (3rd. ed.). 170 Englewood Cliffs, M.J.: Prentice-Hall. Romo, F. P., H. Korman, P. Brantley, and M. Schwartz (1988), "The rise and fall of regional political economies: a theory of the core." Research in Politics and Society. 3: 37-64. Rosenbaum, J. E., Takehiko Kariya, Rick Settersten and Tony Maier (1990), "Market and network theories of the transition from high school to work: their application to industrialized societies." Annual Review of Sociology. 16:263-299. Ruccio, David F. and Lawrence H. Simon (1988), "Radical theories of development: Frank, the modes of production school, and Amin." Pp. 121-173 in C. K. Wilber (ed.). The Political Economy of Development and Under-development (4th ed.). New York: McGraw-Hill Publishing Co. Ruggles, Patricia (1990), Drawing the Line: Alternative Poverty Measures and Their Implications for Public Policy. Washington, D.C.: The Urban Institute Press. Schiller, Bradley R. (1989), The Economics of Poverty and Discrimination. Englewood cliffs, N.J.: Prentice Hall. Schroder, Larry D., David L. Sjoquist, and Paula E. Stephan (1986), Understanding Regression Analysis: An Introductory Guide. Newbury Park: Sage. Schulman, Michael D. and Linda M. Lobao (1989), "Agrarian origins, industrial experience, and militancy: an analysis of Southern textile workers." Sociological Spectrum. 9:379-401. Sheets, Robert G., Stephen Nord, and John J. Phelps (1987), The Impact of Service Industries on Underemployment in Metropolitan Economics. Lexington, M.A.: Lexington Books. Singer, Edward G., Gary P. Green, and Jere L. Gilles (1983), "The Mann-Dickinson thesis: reject or revise?" Sociologia Ruralis 23 (3-4):276-287. Skees, Jerry R. and Louis E. Swanson (1988), "Farm structure and rural well-being in the South." Pp. 238-321 in Louis E. Swanson (ed.), Agriculture and Community Change in the U.S.: The Congressional Research Reports. Boulder: Westview. Slocum, Walter L. (1962), Agricultural Sociology: A Study of Sociological Aspects of American Farm Life. New York: Harper and Brothers. Smith, T. Lynn (1969), "A study of social stratification in 171 the agricultural sections of the U.S.: nature, data, procedures, and preliminary results." Rural Sociology. 34 (Winter):497-510. Staniland, Martin (1985), What is Political Economy?: A Study of Social Theory and Underdevelopment. New Haven: Yale University Press. Streeten, P. (1981), "Why basic needs?" Pp. 8-45 in P . Streeten (ed.), First Things First: Meeting Basic Human Needs in Developing Countries. Suh, Sang Mook and Others (1981), The Actual Conditions of Poverty and Remedy Measures (in Korean). Seoul, Republic of Korea: Korea Development Institute. Swanson, Louis E. (1990), "Rethinking assumptions about farm and community change." Pp. 19-33 in A.E. Luloff and Louis E. Swanson (eds.), American Rural Communities. Boulder, Colorado: Westview Press. Taejon-Chungnam Yearbook Compilation Committee (1990), Taeion-Chunqnam Yearbook (in Korean). Taejon-Chungnam Yearbook Compilation Committee, Republic of Korea. Thompson, A., A. 0. yeboan, and S. H. Evans (1986), "Determinants of poverty among farm operators in North Carolina." Pp. 177-199 in J. J. Molnar (ed.), Agricultural Change: Consequences for Southern Farms and Rural Communities. Boulder, Westview Press. Tickamyer, A. and J. Bokemeier (1988), "Sex differences in labor market experiences." Rural Sociology 53 (2):166-189. Tickamyer, Ann R. and Cynthia M. Duncan (1990), "Poverty and opportunity structure in rural America." Annual Review of Sociology. 16:67-86. Tickamyer, A. and C. Tickamyer (1988), "Gernder and poverty in central Appalachia." Social Science Quarterly. 69 (4):874-891. Tigges, Leann M. and Deborah M. Tootle (1990), "Labor supply, labor demand, and men's underemployment in rural and urban labor markets." Rural Sociology. 55 (3):328-356. Tolbert, Charles, P. M. Horan, and E. M. Beck (1980), "The structure of economic segmentation: dual economy approach." American Journal of Sociology. 85 (5):1095-1116. Tomaskovic-Devey, Donald (1987), "Labor markets, industrial structure, and poverty: a theoretical discussion and 172 empirical example." Rural Sociology. 52 (l):56-74. Tomaskovic-Devey, Donald (1988a), "Industrial structure, relative labor power, and poverty rates." Pp. 104-129 in Donald Tomaskovic-Devey (ed.), Poverty and Social Welfare in the United States. Boulder, Colorado: Westview Press. Tomaskovic-Devey, Donald (1988b), "Poverty and social welfare in the United States." Pp. 1-26 in Donald Tomaskovic-Devey (ed.), Poverty and Social Welfare in the United States. Boulder, Colorado: Westview Press. Turner, Jonathan (1982), The Structure of Sociological Theory (3rd. ed.). Hamewood, 111: Dorsey Press. Turner, Jonathan (1983), "Theoretical strategies for linking micro and macro processes: an evaluation of seven approaches." Western Sociological Review, 14 (1):4-15. Valkenburg, F. C. and A. M. C. Vissers (1980), "Segemntation of the labor market: the theory of the dual labor market." The Netherlands Journal of Sociology. 16:155-170. Walker, Richard (1978), "Two sources of uneven development under advanced capitalism: spatial differentiation and capital mobility." Review of Radical Political Economics. 19 (3): 28-37. Wang, In Keun (1989), "Agriculture and rural development and rural poverty problems in Korea." Korea Observer. 20 (1): 55-92. Wegner, Bernd (1991), "Job mobility and social ties: social resources, prior job, and status attainment." American Sociological Review. 56 (February):60-71. Weinberg, Daniel H. (1987), "Rural pockets of poverty." Rural Sociology. 52 (3):398-408. Wiarda, Howard J. (1988), "Toward a nonethnocentric theory of development: alternative conceptions from the third world." Pp. 59-82 in C. K. Wilber (ed.), The Political Economy of Development and Under-development (4th. ed.). New York: McGraw-Hill Publishing Co. Wilber, Charles K. and Kenneth P. Jameson (1988), "Paradigms of economic development and beyond." Pp. 3-27 in C. K. Wilber (ed.), The Political Economy of Development (4th. ed.). New York: McGraw-Hill Publishing Co. Wood, Phillip J. (1986), Southern Capitalism: The Political Economy of North Carolina. 1880-1980. Durham, N.C.: Duke 173 University Press. World Bank (1990), World Development Report: 1990 Poverty. New York: Oxford University Press. Worsley, Peter (1990), "Models of the modern world system." Theory. Culture and Society. 7:83-95.