3* O -m- ABSTRACT FARM DEPOPULATION AND ITS ASSOCIATIONS WITH CHANGES IN THE COMPONENTS OF AGRICULTURAL SYSTEMS AND RESOURCES; A GEOGRAPHICAL STUDY OF THE NORTHERN APPALACHIANS IN THE TWENTIETH CENTURY By Warren Douglas Slocum Farm depopulation, an old but ever present phenomenon in the United States, has not received the research attention that it deserves, given the serious implications its continuance will have on the welfare of the United States. Farm depopulation, the dependent variable in this study, is equal to the county farm population per- centage remaining from a prior census count at the end of each one of the study's time periods, 1910-1930, 1940-1950, 1950-1960, and 1960-1970. In this dissertation, the areal patterns of depopulation in the northern part of a well-known region of the United States, the Appalachians, are analyzed as to their temporal and areal relation- ships with the agricultural geography and agricultural systems of the Region. Agricultural subsystems and their attributes, i.e., agri- cultural variables, are perceived as the recipients of stimuli from exogenous technological forces and consequently serve in varying degrees as indicators of the importance of outside inputs (resources) Warren Douglas Slocum to agricultural systems and subsystems. Twenty-one agricultural variables reflect those characteristics of agriculture hypothesized as most likely to be changed by technological forces. The natural and geographical setting of an agricultural system influences the ultimate effect of technological forces on parts of an agricultural system. To gain independence among the variables and a general descrip- tion of farming, the agricultural variables are grouped by principal component analyses for each time period. Through the interpretation of the variable loadings, the principal components-~the major "dimensions" of agriculture for the study area--are obtained and these are mapped for areal analysis. These components represent the subsystems, e.g., mechanization, capital investment, off farm inputs, labor, land, or type of farming most important in an area. The next stage of analysis applies to a series of stepwise multiple regressions with the components serving as "truly" independent variables and the farm population retention percentages as the dependent variable. Farm depopulation and retention rates are associ- ated in varying degrees with each component. In the first two time periods farm depopulation is found to be only slightly related to the agricultural components. In the time periods following World War II, the farm depopulation is associated significantly and collectively with several agricultural components. The agricultural components of 1949-50 account for 82.6 percent of the farm depopulation variance in the 19505. The agricultural components of 1969-70 account for 73.3 percent of the farm depopulation variance in the 19605. Off farm employment is most closely associated with ftpalachi tie: the setting. Warren Douglas Slocum farm population losses; whereas dairying is most nearly related to farm population retention. Farm population decreases of the northern Appalachians are now associated with agricultural characteristics under the forces of both the technological system and the geographical setting. FARM DEPOPULATION AND ITS ASSOCIATIONS WITH CHANGES IN THE COMPONENTS OF AGRICULTURAL SYSTEMS AND RESOURCES; A GEOGRAPHICAL STUDY OF THE NORTHERN APPALACHIANS IN THE TWENTIETH CENTURY By Warren Douglas Slocum A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Geography 1976 <7?) C1P'rTig! mm; W“ I In- © Copyright by WARREN DOUGLAS SLOCUM 1976 In Memory of my father, Bernard E. Slocum, who respected a college education, but never had the opportunity to obtain a high school education. ii ACKNOWLEDGMENTS Advancement through graduate school probably requires most of all steadfastness from the candidate and patience from those individuals closely associated with him. Frequently, the following is stated: "One obtains from college what he or she is willing to put into it." There is considerable truth in this statement, but in reality no human being is a world to himself in motivation, knowledge, assurance, financial resources, skills, faith, or whatever. This candidate is indebted to many persons: academicians, relatives, friends, neighbors, and public servants for their advice, guidance, shared ideas, and encouragement. I am grateful for the Opportunities granted to me by several academic institutions, not all of which can be named. The excellent foundational education in geography, I obtained from the Pennsylvania State University in the late 19605 has proven extremely valuable in my doctoral program and subsequent work. Study at Michigan State University in the early 19705 has widened my perspectives and per- ceptions. Dr. Georg Borgstrom, my major professor throughout my [doctoral studies, has kindly and patiently given of his valuable time and ideas, and has provided encouragement at the most apprOpriate ‘times. The Department of Geography provided financial assistance, ‘research and teaching experience during a transitional time in my life. iii Valuable ex vanes, nee Michigan p1 ”an Ecol! i‘scussion briter‘s e Tc him, In: 11181! C011 Valuable experience in field study and insight into rural pe0ple's values, needs, and perceptions were acquired in the Ontonagon County, Michigan project while employed as a research aid by the College of Human Ecology. Dr. Margaret Bubolz and Dr. Joanne Eicher through discussions on the Ontonagon study unwittingly contributed to the writer's evolving dissertation research. To my relatives and family, my wife Ann, daughters Maria and Andrea, my mother, and "grandparents," I express my appreciation for their continued forbearance, work, and support. Friends and students of the University community provided at certain critical times invaluable assistance and advice in computer science, statistics, and cartography. Sherman Hollander, Mike Lipsey, and Mark Walters provided much appreciated cartographic and drafting skills. To all the typists, including my wife and mother-in-law, I give my sincere thanks. Finally, recognition and words of gratitude are expressed to the Michigan Department of Natural Resources (DNR), Bureau of Environ- mental Protection and staff for employment that has given to me both the necessary desire and the financial resources to complete this degree's final requirements. My writing skills since the completion of most of this dissertation have been hopefully improved through the writing of a final report that the DNR will submit to the United States Environmental Protection Agency. iv LIST OF LIST 0? 21mm TABLE OF CONTENTS LIST OF TABLES . . . . LIST OF FIGURES . CHAPTER I. INTRODUCTION: A THEMATIC AND CONCEPTUAL ORIENTATION . . . . . . . . . . The Setting in Population Issues . General Statement of the Problem . . Thesis and Significance of Technology in the Explanation Some Definitions and Operational Measures of Depopulation . . . Historical Depopulation A Tentative and Theoretical Account of World-wide Depopulation . Major Occurrences of Depopulation Throughout the World . Selection of the Subject Population; An Agricultural Population . . . . Agricultural Depopulation in the Northern Appalachians 1910-1930 Agricultural Depopulation in the Region During the Depression Decade . . . . . . . High Rates of Farm Depopulation in the Region, 1940 to 1970 . Disregard of the Region's Farm Depopulation; Some Psycho- political Reasons . . Selection of the Study Area; The Northern Appalachians . Rationale and Importance of the Study Page . xiii 12 15 22 26 30 30 32 33 4S CENTER CHAPTER Agricultural DeveIOpment with Social Planning . . . . . . . . . Human Living Conditions and Public Policy . . . . . . . General Guides to the Research . 2. A REVIEW OF DEPOPULATION AND RELATED RESEARCH: A BASIS OF THEORETICAL IDEAS . An Interdisciplinary and Chronological Overview . . . . . . . . . . The Nature of the Depopulation Phenomena . Definitions of Depopulation Measurements of Depopulation . Classifications: Types of Depopulation . General Locational and Spatial Patterns of Depopulation . The Depopulation Syndrome. Signs, Symptoms, and Characteristics The Basic Causes of Depopulation . The Major Demographic Aspects Contributing Forces and a Theory . Geographical Level of Analysis . Summation of the Temporal Importance of Pull and Push Forces . . . 3. PROCEDURE OF THE INVESTIGATION: AN APPROACH TO REPRESENTATION AND EXPLANATION OF FARM DEPOPULATION IN THE NORTHERN APPALACHIANS A Specific Statement of the Problem Farm Population; Choice of the Dependent Variable . . . . . . . . . . . . . Definitions and Criteria of Population Censuses . . . . . . . . . . . . . . The Quality and Source of Decennial Farm Population Data . . . . Retention of the Farm Population . Subregions of Prolonged Farm Depopulation . . . . . . . . Subregions of Least Farm Depopulation vi Page 51 52 53 @ 57 62 62 65 68 71 81 89 89 90 93 94 96 96 97 98 102 107 112 116 CHAPTER An Overview of Farm Depopulation . . . . The 1910 to 1930 Period The 1940 to 1950 Period The 1950 to 1960 Period The 1960 to 1970 Period Methodology Relationship of the Problem to the Method of Analysis . A Systems Perspective The System Concept . Agricultural Systems . The General Model . General Procedure of the Analysis 4. ANALYSIS OF STABILITY AND CHANGE OF AGRICULTURAL VARIABLES AND ATTRIBUTES IN THE NORTHERN APPALACHIANS, 1910-1970 . . . . . . . Important Temporal Associations of Agricultural Attributes . The Traditional Attributes and Their Associations . . . . . . . . . . . Temporal Patterns of the Important Agricultural Change Variables and Their Associations . 5. ANALYSIS OF THE COMPONENTS OF AGRICULTURE: SOME DOMINANT CHARACTERISTICS AND SYSTEMS OF FARMING IN THE NORTHERN APPALACHIANS, 1909-10 AND 1929-30 . . . . . . . . . . . Identification of the Region's Agricultural Components . . . . . . . . The Components of Agricultural Systems in 1909-1910 . . . . . The Components of Agricultural Systems in 1929-1930 . . . . . . . . . . 6. ANALYSIS OF THE COMPONENTS OF AGRICULTURE: SOME DOMINANT CHARACTERISTICS AND SYSTEMS OF FARMING IN THE NORTHERN APPALACHIANS, 1949-1950 . . . . . . . . . vii Page 119 119 120 121 122 123 123 125 127 128 128 129 132 132 132 138 154 154 156 163 183 CHAPTER Identification of the Region's Agricultural Components . . . . . . . . . . . . . . The Components of Agricultural Systems in 1949-1950 . . . . . . . . . . 7. ANALYSIS OF THE COMPONENTS OF AGRICULTURE: SOME DOMINANT CHARACTERISTICS AND SYSTEMS OF FARMING IN THE NORTHERN APPALACHIANS, 1969-1970 . . . . . . . . . . . Identification of the Region's Agricultural Components . . . . . . . The Components of Agricultural Systems in 1969-1970 . . . . . . . . . . 8. EXPLANATIONS FOR PERIODIC AGRICULTURAL POPULATION CHANGES IN THE NORTHERN APPALACHIANS, 1909- 1970 . . . . . . . . . . . . . . . . . Form of the Explanation The Nature and Application of the Stepwise Regressions . . . . . . . . . Temporal Associations of Agricultural Com- ponents with Farm Population Retention Ratios O O I O O O O O O O O O I I O O O C Major Agricultural Attributes Correlated with Ensuing Farm Population Changes . Net Status of Agricultural Attributes Correlated with Preceding Farm Population Changes . . Geographical Analysis with Residuals . The Relative Areas Associations of Farm Population Changes to the Selected Agricultural Components . . Areal Relationships of Subsequent Farm Population Changes to Agricultural Components . . . . . . . . . . . . . Areal Relationships of Preceding Farm Population Changes to Agricultural Components . . . . . . The Importance of Time Lag Effects of Agriculture . . . . . . . viii Page 183 183 215 215 215 276 276 277 278 279 286 292 294 295 312 319 CHAPTER Page 9. SUMMARY, CONCLUSIONS, AND THE IMPLICATIONS FOR FUTURE RESEARCH . . . . . . . . . . . . . . . . . . . . 323 Summary . . . . . . . . . . . . . . . . . . . . . . . 323 The Problem . . . . . . . . . . . . . . . 323 The Procedure of Analysis . . . . . . . . . . . . 324 Findings and Results . . . . . . . . . 326 Conclusions and Implications for Research . . . . 327 EIIII3ILIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . 328 APPENDIX .IK. VARIABLES INCLUDED IN THE RESEARCH . . . . . . . . . . . . 348 I3. SOURCES OF DATA . . . . . . . . . . . . . . . . . . . . . 350 (:2. CORRELATION COEFFICIENTS MATRIX (1909-1910) . . . . . . . 352 II). CORRELATION COEFFICIENTS MATRIX (1929-1930) . . . . . . . 353 IEEA CORRELATION COEFFICIENTS MATRIX (1949-1950) . . . . . . . 355 'I=3 CORRELATION COEFFICIENTS MATRIX (1969-1970) . . . . . . . 357 ix Table 10. 11. 12. LIST OF TABLES Total Experienced Civilian Labor Force in Agriculture by Region within the United States, 1950, 1960, and 1970 Farm Population Change for Appalachian and Non-Appalachian Areas of New York and Pennsylvania 1910-1970 Change in Farm Population, United States, Regions and Divisions by Decades, 1920- 1970 . . . . . . . . . . . Change in Rural Farm Population, United States, Regions and Divisions by Decades, 1910-1970 . . . Rural Nonfarm and Rural Farm P0pu1ation by States in the Eastern United States, 1960 and 1970 . . . . . . . . . . . . . . Populations of the Major Pennsylvanian Cities in the Northern Appalachians, 1930 to 1970 Agricultural Variables Interrelated for Two or More Time Periods; the Traditional and Change Attributes of Agricultural Systems . Highest Linear Correlations Among Agricultural Variables, 1909-1910 . . . . . . . . . . Highest Linear Correlations Among Agricultural Variables, 1929-1930 . . . . . . liighest Linear Correlations Among Agricultural Variables, 1949-1950 liighest Linear Correlations Among Agricultural variables, 1969-1970 0 o o o o o o 0 Major Loadings from the Principal Components Analysis of Agriculture, 1909-1910 Page 25 27 28 29 44 46 133 140 141 144 151 157 Table Page 13. Major Loadings from the Principal Components Analysis of Agriculture, 1929-1930 . . . . . . . . . . . 165 14. Major Loadings from the Principal Components Analysis of Agriculture, 1949-1950 . . . . . . . . . . . 185 15. Number of Commercial Dairy Herds in Pennsylvania, 1950-67 . . . . . . . . . . . . . . . . . . 190 16. Percent Farm Operators Working Off the Farm 100 or More Days, 1950-1960 . . . . . . . . . . . . . . . 193 17. Average Age of Farm Operators, 1940-1969 . . . . . . . . . 200 18. Hay and Corn Silage and Grain Production Trends in the Northeast, 1949-1969 . . . . . . . . . . . 206 19. Major Loadings from the Principal Components Analysis of Agriculture, 1969-1970 . . . . . . . . . . . 216 20. Land in Farms for the United States, Censuses of 1850 to 1969 . . . . . . . . . . . . . . . . . . . . . 220 21. Assets and Farm Debt in the United States, l950-1972 . . . . . . . . . . . . . . . . . . . . . . . . 221 22- Average Expenditures for Gasoline-Petroleum Products and Fertilizer Per Farm in Major Dairy Counties, 1954 and 1969 . . . . . . . . . . . . . . 223 23- Status of Dairying in the United States, 1958 to 1972 . . . . . . . . . . . . . . . . . . . . . . 229 24~ Milk Supply, Use, and Carryover, 19505 to 19705 . . . . . . . . . . . . . . . . . . . . . . . . . . 230 25- 'The County Pastureland/Cropland Acreage Ratios, 1929, 1949, and 1969 . . . . . . . . . . . . . . 238 26- Sales of Crops and Crop Sales as Percent of Total Sales in the Southeastern Counties of the Northern Appalachians, 1964 . . . . . . . . . . . 249 27° Value and Sales of Crops in Southwestern Pennsylvania, 1964 and 1965 . . . . . . . . . . . . . . . 251 28. Linear Correlations between Capital and Labor, 1949-50 and 1969-70 . . . . . . . . . . . . . . . 256 xi Table 29. 30. 31. .32. 335. 34l. 355. 36. 37. Average Yields of Corn Harvested for Grain per Farm in Selected Counties, 1949 and 1969 . . . . Trends in Corn Production in Pennsylvania, 1944-1949 and 1964-1969 . Temporal Correlations between the Degree of Tractor Acceptance or Mechanization and the Pasture-Cropland Ratios . Correlation Matrices for Selected Agricultural Components and Farm Population Retention Rates, 1909-1930 Correlation Matrix for Selected Agricultural Components, 1949-50 and Farm Population Retention Rates, 1950-1960 . . Correlation Matrix for Selected Agricultural Components, 1929-30 and Farm Population Retention Rates, 1920-1930 . . . Correlation Matrix for Selected Agricultural Components, 1949-50 and Farm Population Retention Rates, 1940-1950 Correlation Matrix for Selected Agricultural Components, 1969-70 and Farm Population Retention Rates, 1950-1960 . . . Correlation Matrix for Selected Agricultural Components, 1969-70 and Farm Population Retention Rates, 1960-1970 xii Page 264 265 275 280 283 287 288 289 291 Figure 1. 2. 3. 4. 10. 11. 12. 13. LIST OF FIGURES Main Areas of Population Losses, 1960 to 1970 . Main Areas of Population Losses, 1940 to 1970 . The Appalachian Region The Northern Appalachians; Counties of the Study Area . . . . . . Subregions of the Appalachian Region Physiographical Regions of the Northern Appalachians . . . . . . . . . . Percentage Retention of the Farm Population by County in the Northern Appalachians, 1910 to 1930 . . . . . . . . . . . . Percentage Retention of the Farm Population by County in the Northern Appalachians, 1940 to 1950 . . . . . . . . . . . Percentage Retention of the Farm Population by County in the Northern Appalachians, 1950 to 1960 . . . . . . . . . . . . . Percentage Retention of the Farm P0pulation by County in the Northern Appalachians, 1960 to 1970 . . . . . . . . . . . . . Basic Model of Technological Interaction with a Population System through Natural and Agricultural Systems . . . . National Trends in the Comparative Use of Agricultural Resources . . . . . Expenditures for Commercial Fertilizer per Acre of Total Cropland, 1954 xiii Page 23 35 37 42 43 108 109 110 111 130 146 147 Figure 14. Replacement of Horses with Tractors . 15. Areal Distribution of Component One Scores, Intensive Agriculture, 1909-1910 16. Areal Distribution of Component Two Scores, Cattle and/or Dairy Farming, 1909-1910 17. Areal Distribution of Component Three Scores, Cropland Availability, 1909-1910 18. Areal Distribution of Component One Scores, Intensive Agriculture, 1929-1930 . IS). Areal Distribution of Component Two Scores, Dairying, 1929-1930 . 2C). Areal Distribution of Component Three Scores, Self-Sufficing Farms, 1929-1930 . . 21.. Areal Distribution of Component One Scores, Dairying and Off Farm Work Ratio, 1949-1950 . 22. Locational Change in Bituminous Coal Strip Mining in Pennsylvania, 1944 to 1959 23.. Areal Distribution of Component Two Scores, Capital Intensive Agriculture, 1949-1950 24. Areal Distribution of Component Three Scores, Labor Oriented Agriculture, 1949-1950 . 25- Areal Distribution of Component One Scores, Dairying and Off Farm Work Ratio, 1969-1970 . 26. Areal Distribution of Component Two Scores, Capital Intensive Agriculture, 1969-1970 27- Areal Distribution of Component Three Scores, Labor Oriented Agriculture, 1969-1970 . 28- Price Trends of Feed Grains . 29- Milk-Feed Price Ratios 30- 'Trends in Feed Grains Consumed 31. Trends in Sources of Protein Consumed . xiv Page 148 158 160 162 167 171 177 188 191 209 212 226 241 253 270 271 272 273 Figure 32. Residuals from Regression of Retention Rates of Farm Population, 1910-1930, on Scores of Associated Agricultural Components, 1909-1910 . 33. Residuals from Regression of Retention Rates of Farm POpulation, 1950-1960, on Scores of Associated Agricultural Components, 1949-1950 . :54. Percent of All Farm Operators Working 100 or More Days Off Their Farms, 1954 . 355. Residuals from Regression of Retention Rates of Farm Population, 1950-60, on Scores of Dairying-Off Farm Work Ratio, 1949-50 . . . . . . . . . . . . 365. Residuals from Regression of Retention Rates of Farm Population, 1950-60, on Scores of Labor Oriented Agriculture, 1949-1950 37'. Residuals from Regression of Retention Rates of Farm Population, 1940-1950, on Scores of Associated Agricultural Components, 1949-1950 . . . . . 33.. Residuals from Regression of Retention Rates of Farm Population, 1960-1970, on Scores of Associated Agricultural Components, 1969-1970 . 39- Residuals from Regression of Retention Rates of Farm Population, 1960-1970, on Scores of Associated Agricultural Components, 1949-1950 . . XV Page 297 305 307 308 309 314 315 321 5.1 . 5‘ Jim” ‘5 ”:3. new 1 CHAPTER 1 INTRODUCTION: A THEMATIC AND CONCEPTUAL ORIENTATION The Setting in Population Issues Among the major issues and concerns of the times are overpopu- 1a1Li<>n and the interrelated and concomitant problems: famine, pollution Of"t11e environment, energy shortages, and depletion of the earth's finite resources. No similar level of awareness exists for the serimous problem of depopulation. In the United States many rural areas have: sustained large population losses and have sent, and continue to send, people to urban and metropolitan areas. The investigation of the caluses of rural population decreases have been inadequate. In a market economy where jobs are frequently lost in the countryside due to the adoption of innovations and acceptance of technological changes, the net nurvement of human resources out of rural areas often sentences the remaiJring rural p0pulation to difficult social and economic readjust- ments and subjects such areas of depopulation to many years of exPloitation, economic dependence, and social stagnation. The depopulation of rural areas in the United States (Figure 1) has come about primarily through outmigration and secondarily through growing natural decrease (Beale, 1964; 1969). Natural decrease occurs ‘whe“ deaths exceed births--traditionally, a rare condition reached 1 ordinarily after much outmigration. As people of reproductive ages generally compose the larger proportion of the outmigrants, the median age and dependency ratio of the source population significantly increase. A knowledge of the reasons for population changes is basic to the planning of a healthy economy and society. To successfully treat an illness, one needs to define well the primary causes. The treat- ment of causes is preferred to the treatment of symptoms in a long run strategy for problem solving. The complex depOpulation phenomenon appears to have numerous causes, and is thus relatively poorly researched. To reach more meaningful results, a population, because of its heterogeneity, needs to be disaggregated. This writer has chosen the farm population as the subject of study because of the major declines in its numbers over several decades. Particularly the associations of the areal patterns of agricultural depopulation with the evolving characteristics and spatial patterns of agricultural systems are explored and analyzed over time in this dissertation. General Statement of the Problem In this effort of explaining agricultural depopulation, associations will be made of farm pOpulation losses with changes in selected characteristics of agricultural systems, i.e., mechanization, resource inputs, land use, labor, and production, and with notable differences in the environmental and socioeconomic conditions. Empirically, it is quite evident in many rural areas of the United States that agriculture and country life have deteriorated catas- trophically. To be answered in this research is to what extent changes in the agricultural geography account for the historical and spatial agricultural depopulation in the study area. The areal patterns of the farm population declines are anticipated to give clues to the causes of and suggest more specific hypotheses for the major farm population losses. Due to the adOption of new technology and increasing amounts of new and substituted inputs, e.g., capital, feed grains, and fertilizer, obtained primarily off the farm, significant changes have occurred in the functioning of the agricultural systems of some regions. Some practices of the "new agriculture" are in general ecologically and sociologically unsound, and developments and the evidence support this view (Milk, 1972, pp. 233-234). Little emphasis and attention have been placed on the negative social and economic con- sequences of the "new agriculture" upon the conditions and welfare of the agricultural population. Needed is more awareness that new inno- vations may be used sometimes ineffectively and destructively in both an environmental and sociological sense in areas not possessing the optimal site and situational characteristics, and socioeconomic con- ditions for their successful implementation. Ecological assessments of agricultural systems for under- developed areas are crucially needed both overseas and at home but such are easily neglected where there exists the immediate need to increase food production. Elements of some agricultural systems have clearly debilitating effects on physical and human resources. By studying how the agricultural systems have evolved and functioned in different time periods and selected areas within a region of a accep' appli. aigra' gsiie. asses. mi: 9556111 developed nation, the writer anticipates that the findings of a case study may provide some evidence of the adverse consequences of complete acceptance of new agricultural technology. Some findings may be applicable to the deve10ping world, now faced with massive rural-urban migration, unemployment, urban hunger and general malnutrition. A few guidelines to successful rural development may evolve. Realistic assessments of agricultural technologies and systems' limitations in varying physical and cultural areas and regions of the world are essential to human welfare. Thesis and Significance of Technology in the Explanation Does the degree of farm depOpulation in an area or region differ according to the nature of the prevailing agricultural systems and the changing agricultural geography? Agricultural systems are of course subsystems of a culture, and therefore, are affected by the technological changes within that culture. Examples of rural depopu- lation come from deve10ping and industrialized countries and capital- istic and socialistic nations (Council of Europe, Vol. 1, 1968, p. 1) thus, there appears to be a variety of causes, some that are culture- specific. Some caution needs to be taken in generalizing from one culture. Each population possesses and belongs to a culture; there- fore, it is really not presumptuous to assume that within this "elaborate mechanism" (White, 1949, p. 166) or large system may be the major factors underlying the depopulation process. One of the most pervasive aspects of the United States culture has been techno- logical change. Technological developments have vastly altered the V .. ClltatT new! Cufiter. '1 mung}. “If“ .1 Ian's s at hand techno: all hum 5?" med .ientua: 135%: SEE fgx 39 m 44'- . i«1:13; character of Western agriculture, the agrarian society, and the self concepts of many unemployed and underemployed persons who may feel superfluous. Culture is a system composed of technological, sociological, and ideological subsystems. Naturally each is related to the other, but the strengths of these relationships is not equal. The major influence is effected by the technological system. This is so because man's survival, and ultimately culture itself, depends upon the means at hand to obtain the necessities of life from the environment. "The technological system is therefore both primary and basic in importance; all human life and culture rest and depend upon it" (White, 1949, pp. 364-365). With the application of much new technology, i.e., new inputs and mechanization in agriculture, a surplus of agricultural labor eventually developed in the United States. Too many farmers and farm laborers were generally thought to be the basic cause of the trouble- some food surpluses and of the low farm incomes in the United States. Large numbers of emigrants in the late 18005 and early 19005 added additional labor; but farmers after World War 11 experienced diffi- culty finding farm help. In the United States the earliest demand for more machines and their invention stemmed from the labor shortages resulting from the Mexican War, the California Gold Rush, and the Civil War (Barger and Landsberg, 1942, p. 198); similar events since, e.g., World War I and World War II have generated demands for further deveIOpment of new agricultural technology and the adoption of labor- saving methods and mechanization. The new technology generally substituted for much farm labor which was particularly scarce during and after wars; and, the strong dependence at these critical times upon the succession of new and "more modern" technical creations and developments to increase production, probably led to an over- acceptance of these "new scientific ways and means" and to the develop- ment of "technological overpopulation” (Pinchemel, 1969, p. 114). Thus, human resources were pushed out of agriculture. The technological capabilities possessed by a people will also dictate the materials or substances that are considered valuable resources (Carol, 1967, p. 285). As the technology changes, different resources are required. Some areas or environments may not possess the resources needed by the updated technology; and thus, new tech- nology is sometimes ill-fitted for some areas and environments. Through additional technology, man changes his socioeconomic organi- zation and usually gains in the aggregate, access to a greater array and different sets of resources, with more impact and control upon the environment; but, sometimes for many areas this latter development may be an illusion when the resources favored and assembled by the new technology are not as available as the set of resources required by the older technology. When the right resources cannot easily be obtained and matched with the new technology (presumably of yet a higher order), an area may be unable to compete successfully in pro- duction and trade. Because of rising population such an area may have an abused environment. In an exchange economy, people of such an area are at a disadvantage in producing a good that is produced easily in other places by the new technology. If resources can be obtained economically from without, the new technology may be incor- porated into the production process. Because advances in technology generally require a greater range and variety of inputs, the proba- bility of gaining all the needed elements is lowered. The thesis of this dissertation is that changes in the character- istics of the agricultural systems brought about through the adoption of new technology, at times adversely change the value and usefulness of human, bio-physical, and socioeconomic resources existing in rural areas, and set in motion "push" factors, i.e., conditions, which facilitate the depOpulation process. Technology, e.g., innovations, inventions, and new methods alter the agricultural systems and existing mix of inputs through requiring new sets of resources, e.g., labor, level land, well-drained soil, and capital for investment-~to name only a few; these inputs sometimes are not present or owned and cannot always be gained or arranged in particular places. Some Definitions and Operational Measures of Depopulation The general topic of this research effort, depopulation, regardless of its various types, e.g., rural, agricultural, and central city, or other named for its particular place of occurrence, can be thought of as either a phenomenon or a process and either a condition or a set of conditions. In this study the descriptive and analytical tasks address the Spatial patterns and causes of agricul- tural depopulation, and secondarily, the depopulation process, for at the origin of the depopulation process are the causes. To solve a problem, e.g., agricultural depopulation, one theor- etically needs to know first what it is before attempting to discover how it came to be. Unfortunately, little research is available on the definition of depopulation and when definitions are given they tend to be too restrictive to give a complete concept of and a full measure of depopulation. The failure to delineate depopulation in nearly all the works on the subject indicates its meaning is not agreed upon. Although many personal definitions of depopulation exist, most individuals engaged in a study of depopulation would agree it is the process, accomplished fact, or consequences of popu- lation loss from a given area during a given time. The most complete definition of depopulation this writer has found comes from a British government publication on depopulation in Wales. It states, Depopulation is sufficiently defined for our purpose as a sub- stantial continuing decline in population. It is generally the result of net outmigration of sufficient volume to offset any natural increase (excess of births over deaths) of population, but in extreme cases it can result from a combination of natural decrease (excess of deaths over births) of population and net outward migration (Great Britain Ministry, 1964, p. 1). To this rather recent statement as to what depOpulation is, one may add the essence of implicit and explicit meanings from a few older studies. In the late 19th century, rural depopulation was receiving considerable attention in the British Isles (Saville, 1957, pp. 5-6). Influential people of the times engaged in discussions of the nature of the rural depopulation and whether it actually existed or was as serious as some persons thought (Ogle, 1889). At that time the conceptualization of depopulation began its primary evolution. 10 "Depopulation of the rural districts," then a phrase apparently very much in use, had a very fluid interpretation. Ogle wrote, . . sometimes no more is meant that the population of the towns is increasing more rapidly than that of the rural districts, and, at their expense; while at other times the phrase is used more properly, and means that the population of the rural districts is diminishing absolutely, and not merely in comparison with the towns (Ogle, 1889, p. 205). Thus, depopulation is defined operationally in a subsequent British study as "a diminution in the number of the inhabitants of a district, as compared with those enumerated at a preceding census" (Longstaff, 1893, p. 380). This writer has incorporated in his working definition of depopulation some of the above mentioned and more universally accepted concepts relating to the depopulation phenomenon. Depopulation is defined for statistical analysis as any absolute and relative decrease or net loss in a population--a negative p0pulation change between any two consecutive censuses. The degree of areal population loss is calculated as the percent a population at the end of a decade is of its population at the decade's beginning. There are further refinements possible in an operational definition, but as estimations sometimes have to be made for some of the basic data, there is a greater chance of obtaining more error than if a measure of depopulation is calculated from the census data. For example, agricultural depopulation is measured in a contemporary study as "the difference between actual 1961 census population of farms and the 'expected' farm population of 1961. The 'expected' farm population (is) computed by adding the estimated births and deducting the esti- mated deaths from the 1961 actual farm population" (Szabo, 1965, p. 39). 11 This measurement of depopulation generally results in areas of popu- lation loss values higher in depopulation than when only population change from census to census is used, as consideration is given to those persons who would have been born to couples in a given location had they not moved before the end of the intercensual period, essen- tially resulting in a measure of "apparent net migration" (Szabo, 1965, p. 25); however, because of the required estimates, the amount of error generally increases. The fullest extent of depopulation is obtained when allowances are made for the indirect demographic effects of net migration, especially the births which would have occurred within the study if youths, young adults, and middle aged adults had not moved away during the decade. Because of the need to obtain comparability in the popu- lation data among the time periods of this dissertation and the desire to have reliability in the data, this writer has sacrificed the poten- tially highest measures of depopulation obtainable for a measure that gives a "balanced" or "moderate" result, i.e., intercensual population change. As migration is in most instances the major component of the depopulation process, an analysis of the available net outmigration data could serve as a test and check respectively of: (l) the pro- portion of the population change and depopulation directly attributable to each of two demographic processes, migration and natural change, i.e., net outmigration and natural decrease; and (2) the adequacy of the negative population change measure in representing the full extent of‘depopulation. Where outmigration is high, one would expect, given 12 outmigration is mostly composed of the reproductive population, an under representation of depopulation when the population change value is used; however, counties with considerable outmigration may have a positive population change figure because natural change remains positive and partly compensates for migration losses. Therefore, sufficiently high natural increases will conceal the population losses from migration; thus, a county may incorrectly appear through popu- lation change data not to have lost population during two censuses and not to have depopulation. Population change gives the actual differ- ence of people living in a place on two consecutive dates, but this is a rather static and gross measure which fuses the individual effects of births, deaths, and migration. Migration could not be used in this study as the measure of depopulation for a number of reasons. Data on the migration of the farm population is incomplete. Net outmigration seldom represents all the depopulation; as high out- migration has occurred for many years among the farm population, natural decrease now accounts for much of the population losses. Historical Depopulation A Tentative and Theoretical Account of World-wide Depopulation The phenomenon of depopulation, encountered in many parts of the world throughout recorded time and particularly associated with the well-to-do world in recent times should be seriously considered as an important topic by social scientists interested in theory building. The study of regressive aspects of population and cultural change (Sestini, 1962, pp. 479-490) is an important undertaking for 13 gaining perspectives on the rise and fall of civilizations and on the consequences of man's utilization of resources and the environment. Depopulation occurred presumably periodically throughout man's prehistory. As man through trial and error searched for more security, power, and better and more abundant resources, discovered and explored his resource base; he at times became the victim of his enemies and natural disasters and decreased in numbers in many places. There were, however, periods of time when some populations were in balance with their resources, e.g., as with some of the American Indian tribes; nevertheless, it may be presumed that as prehistoric man gained in numbers, more conflicts arose, resources at times were depleted, and man had to find a new home. It is doubtful whether at any time, all places were experiencing either stability or increases in their human populations. With man's mental and cultural development, human decisions and actions became over generations ever more the prime forces initiating the depopulation process. Man's decisions were not always in his best interests. The natural environment changed significantly only over long geological time periods and natural disasters happening infrequently and unpredictably had a comparatively minor role in periodic reductions in the human populations. Nevertheless, the recognition of man as a change agent came relatively late in history as Buffon of France in the 18th century was the first to consider man as a powerful influence on earthly changes (James, 1972, p. 136). Although George Perkins Marsh, an American, released his Man and Nature in 1864 stressing man's actions in altering the earth, the 14 warning was to a Nation replete with seemingly unlimited resources. Few took his words seriously until the mid-20th century, a time when his book was "rediscovered" and reprinted (Marsh, 1965). Early man was almost completely dependent upon the land resource--the biota, soil, water, and minerals associated with it. He had to expend at times enormous amounts of labor to obtain the necessities of living. As man gained progressively more knowledge and management skills, important factors in increasing production, he began to have more leisure time to develop his culture, to live at a higher level than the local resources directly provided, and to obtain resources and goods from outside his community. At this point man began to live beyond "his" resources. Once man had obtained tools and techniques, either indigenously or commercially, he greatly increased his effectiveness in using the local resources. This accumulation of capital, e.g., tools, seemingly at times proved to be a detriment to mankind as it made possible the release of many more resources in the shortrun from sources and areas lacking an adequate land resource base. Inadequate time remained for the renewal of some primary and necessary natural resources, e.g., soils, forest, and animals. The speeded up resource withdrawal process made feasible the growth of human populations often at the expense of other populations upon which humans depended for survival. The accumu- lation and application of capital goods along with the use of short term vision, the latter which is understandable given the short life expectancy, led to overpopulation, to a lower standard of living and health, and necessitated attempts to re-establish a balance between 15 resources and people through outmigration to sparsely settled or unsettled areas. Depopulation thus took place in the areas of origin. Unfortunately, a group's way of living, preconceived notions, and adjustment abilities were frequently inappropriate for their new environments; therefore, the new settlement was sometimes abandoned. Many areas of destination, including central cities of today, have eventually had depopulation tendencies. This was the history of many places in the United States. ". . . Settlement, unsettlement, and resettlement have been the principal occupation and source of wealth in this country" (Lord, 1962, p. 348). Especially since the Industrial Revolution man has greatly speeded up the exploitation, wastefulness, and abandonment of both the natural and his man-made environments, e.g., farms. He has character- istically disregarded his stewardship role as caretaker of the land and perceived his world in the shortrun and in terms of the necessities of life. He has failed to foresee the eventual negative longrun consequences of his actions. Major Occurrences of Depopulation Throughout the World Direct and inferred evidence suggests the phenomena and process of depopulation has affected the human race since its arrival upon the earth. Nearly all peoples and places appear to have been affected at least one time. Additional thought on this topic could lead to rewarding theories concerning the causes, the processes, and the con- sequences of depOpulation. Concerning the obvious and important role jpopulation declines have had in the varying successes of peoples, 16 nations, and civilizations, it is rather surprising that more attention has not been given to this important phase of population cycles. The primary causes of the depopulation phase of population cycles appear to have been some different in the prehistoric societies as contrasted with most of the groups for whom we have documents; nevertheless, wider applicable generalizations and perspectives can be gained on the nature of and the basic forces contributing to population losses and their consequences by taking into account archaeological findings and the anthropological information on primi- tive peoples of different cultural levels. Generally, the higher developed the civilization or society, the more the population losses become associated with cultural factors. It is important to recog- nize, however, that in prehistoric times the changes were probably more rapid and violent than they are today. Thus a halving or doubling of a population during a century should almost be regarded as normal, and constancy regarded with suspicion (Hollingsworth, 1969, p. 171). Early man‘s population downswings were more attributable to the sudden occurrences of natural disasters than those of historical and contem- porary primitive peoples, who experienced varying adverse effects from contacts with advanced cultures. Many population declines can be traced to the contact of two cultures with the least developed culture experiencing the losses. Petersen has listed some of the factors bringing about the decline of societies and peoples when they experience the onslaught of a more advanced culture. The resulting disease, violence, and servitude became major causes for the death and lowered fertility of millions (Petersen, 1961, pp. 334-335). Infectious diseases especially took 17 their toll of the indigenous populations of South America even before the Spanish consolidated their control. The population of central Mexico is said to have declined within the same period, i.e., the 16th century, from 25 million to one million persons and the indigenous population of the Caribbean area nearly disappeared. The slave trade in western Africa is said to have carried off an estimated maximum of 20 million persons, only one third or less of whom survived to reach the New World (United Nations, 1973, p. 19). The American Indian destruction came about mostly through violence; the order of events was generally ". . . first, land removal acts, expulsion, wars, and forced migrations; then, in sequence, food shortages, starvation, and epidemics . . . ." As the depopulation cycle decreased in intensity, reoccurrences of diseases, e.g., smallpox, cholera, tuberculosis, syphilis, diphtheria, dysentery, and trachoma continued to keep the population from increasing until the late 18005 (Phelps and Henderson, 1958, pp. 181-182). Finally, acculturation and the adopting of some material culture, e.g., guns and alcohol, of the invading cultures had a disabling effect upon the lesser developed societies. Social disorganization often was the consequence and where the family was affected and security lost, the raising of children was significantly discouraged (Petersen, 1961, pp. 335-336). Seemingly, however, many cases of depOpulation may be traced to the malfunctioning of a culture or society. Generally, the period of declining numbers is started by famine, disease, or some drastic cultural deficiency; then it proceeds to a destruction of every social institution (Phelps and Henderson, 1958, p. 181). 18 That famine and pestilence ordinarily appear together suggests a direct relation between them, but apparently the usual reason is that both the supply of food and public health depend on-- and can affect--the maintenance of social order (Petersen, 1961, p. 366). Although it is generally accepted that the hunting and gather- ing peoples had losses in numbers, there is no consensus as to the general occurrence of depopulation among primitives of ancient times (Petersen, 1961, p. 333), but it would appear natural and man-made famine and disease have always been with the human race, and these probably inflicted casualties on man, especially through the many years that it took him to occupy the earth and through the migrations that brought him into contact with new environments and sometimes with other peOple. In the preindustrial civilizations, the expansion of trade and the development of agriculture--like the later industrial revolution-- brought dramatic increases in population; but, at times these agri- culturally based civilizations had population decreases as famine, disease, and social disruption were encountered (Petersen, 1961, pp. 343 and 373). One by one all the ancient civilizations suffered reversals and some collapsed. Whether all these populations experi- enced depopulation both before and after their political setbacks or downfall is not known but undoubtedly, when social order weakened, for whatever the reasons, population losses took place. Rome is perhaps the most cited example of a civilization which fell due to certain causes, particularly internal ones. Symptoms of the disintegration appeared early in the 3rd century, before the general depopulation of the Empire with the population losses of the Roman cultural core and 19 the Italian peninsula, the decline of agriculture, and the abandonment of arable lands. Countless efforts to bring land back into production were increasingly unsuccessful. The change and neglect in land use raised the chances and effects of disease, particularly malaria. Finally, the authorities had to resort to retraction of the settlers and agricultural workers' rights to leave the land; thus, imposing serfdom, a major characteristic of the medieval era (Petersen, 1961, pp. 366-368). Although following the demise of the Western Roman Empire, there were numerous emigrations and invasions into the former Roman territory by various peoples, up to 1000 A.D. the continuing general decreases in the population can be attributed to adverse economic and social factors, plus the plagues of the 6th Century (United Nations, 1973, p. 16). The next general population decrease in Europe came from several devastating strikes of the Black Death in the mid-14th Century. Much of continental western Europe suffered declining or stagnant population changes; e.g., the German states lost 40 percent of their population due to the Thirty Years War. Sweden and Finland appear to have lost considerable population because of disastrous harvests and subsequent famine in the late 17th Century (Hollingsworth, 1969, p. 173). Ireland experienced large population decreases in the late 18405 because of potato crop failures, famine, and emigration (Woodham-Smith, 1962). Spain appears to have had little growth stability in its population numbers during the Middle Ages and from the 15 to 18th centuries experienced a "drastic decline" (United Nations, 1973, p. 17). Indications of the widespread nature of the 20 depopulation in England, France, and some other parts of Europe during the last one hundred years is presented in the literature review found in Chapter II. Europe's population seems to have been particu- larly affected by warfare; whereas, outside of Europe, famine accounted for many population declines. In the realm of the Eastern Roman Empire and Byzantine Empire, population losses occurred, an ". . . it seems indisputable that demographic decline accompanied economic reverses, pestilence, and wars of its last centuries" (United Nations, 1973, pp. 17-18). As the civilizations rose and fell in the Middle East, there appears to have been extreme fluctuations in population. Egypt's population change has reeled negatively and positively several times probably due mostly to numerous invasions, wars, and exceptionally deadly epidemics. Before and during the Dark Ages, Syria appears to have lost half of its population and depopulation and abandonment of the countryside was very evident in the 18th and 19th centuries. In addition there appears to have been declines in the Asian towns of the Ottoman Empire during the same time (Hollingsworth, 1969, pp. 248- 251 and 307-310). China and India's population decreases have been particularly caused by famine, e.g., in China in 1877-1878 as many as 13 million people may have perished. Breakdowns in social order is a major theme in Chinese history. In the Taiping Rebellion (1851-1864) probably more than 30 million perished leaving some areas completely devoid of population (Petersen, 1961, pp. 363 and 369). 21 Given the continuance of very high pOpulation growth rates in much of the deve10ping world, some of the above mentioned natural and social controls on population expansion could fall into operation at any time, resulting in large scale depopulation. Throughout the underdeveloped world, cityward migration has taken generally a lower proportion of the rural population than did the urbanization in the developed countries following the Industrial Revolution. The rural population continues to grow. City growth is mostly attributable to high population growth (Davis, 1965, pp. 15-19). Net population change in the developing world is expected to add more than 400 million to its rural populations in this decade, compared to the developed world whose rural population is expected to further diminish (Borg- strom, 1973, p. 249). There presently exists however some evidence that United States nonmetropolitan and rural areas are again experi- encing population growth at rates exceeding those of the metropolitan areas (Beale, 1975; Zelinsky, 1975). Yet, in spite of rapid city growth and rural-urban migration, the world is still essentially rural. The increasing socioeconomic disparities and agricultural population densities in rural areas of developing countries, plus the massive problems of the rapidly growing cities would suggest that there is growing danger of modern techno- logical support systems breaking down under the weight of mushrooming population, Opening the way to much more famine, pestilence, social disorder, and disruption. Actually, the traditional causes of depopu- lation have continued to Operate but at reduced levels throughout modern times within the developing nations; whereas, in the 22 industrialized world, the major underlying cause of depopulation has been technological change. New and more modern machines and methods have reduced the needs for human power. Displaced workers have sought work elsewhere, and birth rates have declined. Selection of the Subject Population; An Agricultural Population Heavily industrialized nations have high rates of agricultural depopulation. Overall, the most technologically advanced country is the United States. To achieve and maintain a high level of technical and economic progress, much geographical mobility of labor is required. Economically developed areas within wealthy countries exhibiting general population growth usually display agricultural depopulation. A cartographic display of the total population changes in the last three decades in the continental United States reveals a sub- area of atypical and persistent decreases in the total population within the Northeast, a socioeconomically advanced region with general population growth. Especially striking is a contiguous pattern of the countries suffering population declines within the central area of the region, particularly in Pennsylvania. The northern and western boundaries of the State are nearly outlined by losses in stark con- trast to the adjacent states (Figure 2). The question naturally arises as to the contribution of farm population losses to total population declines in this area. If the numerical change in the entire Pennsylvania pOpulatiOn were to be considered, many explanations would be needed to account for the geographical losses in the State's population. The entire 24 pOpulation does not require analysis to achieve the objectives of the proposed study; nevertheless, two of the major reasons for the popu- lation declines are given here as background for the study. The coal industry mechanized and the demand for coal fell as oil and natural gas became favored sources of energy. Pennsylvania had much heavy industry, e.g., iron and steel manufacturing in and around Pittsburgh, much of which was among the oldest such industrial activity in the Nation. The rate of modernization failed to keep pace with the more recently developed industrial areas, such as in the Great Lakes region; thus, markets were lost and so were jobs. Technological changes in one area affect other areas. To gain an accurate assessment of the effects of new tech- nology on population, the group chosen for study needs to be rela- tively homogeneous and defined. Although agriculture has fully engaged both absolutely and relatively few people in the Northeast (Table 1), Pennsylvania's rural population--the largest in the Nation and discussed later in this chapter--could have been economically affected through the decline of the agricultural population since many earned a living by servicing the farm families. This hypothesis is not tested however in this research as the major concern is directed toward evaluating the extent of the agricultural population decline and giving reasons for its historically persistent losses. In addition, it should be noted that the Pennsylvanian sub- area under discussion plus the most southern counties of upstate New York form the northern section of the officially designated Appalachian region, a large area of widespread depopulation. Common features 25 .ooH Oanmb .ommH use mmmm Ofinmh .ooma mHon can con madame .onmu ”moezHo> xumeesm .m.: .coHumfiamoa 0:» mo mowumfiuouomumcu .cofiumfismom ecu mo momSmcou .w.: ”moopzom amo.vma mmo.vvc mvo.oma owe: Nam.aom Hom.vaw.a nom.mm~.m guzom moH.~mm ANn.Hov.~ oa~.vmm.~ amuucou nouoz mo~.ma~ ma~.~mm mvo.v~m ommegouoz uo>o cam whee» OH mCOmuoa HH< m:Omuom munch uo>o ecu memo» VA m:0muom ~H< :onom cam“ coma omaa . .onma cam .comH .ommm .moueum Oeuflcs ozu away“: cowmom xn ousufisoHum< a“ oouom momma :mwaw>fiu voocofiuonxm meOOHT-.H wfinmh 26 associated with Appalachia undoubtedly underly many of the population changes in the study area. Agricultural Depopulation in the Northern Appalachians 1910-1930 In respect to the rates of farm pOpulation losses, how did the northern Appalachian area of both Pennsylvania and New York (Table 2) compare with the population changes elsewhere in the Nation (Tables 3-4)? The analysis of both farm population estimates and census data, the latter mostly rural-farm, revealed the Northeastern region of the United States had the largest rate of agricultural population decline within the Nation from 1910-1930 which averaged roughly one percent per year. Within this same period, the losses within the Appalachian counties of New York and Pennsylvania averaged slightly higher than the rates of the Northeast, of the non- Appalachian counties of the two states, and of the Middle Atlantic division in which the two states are located. Therefore, during 1910 to 1930, the Appalachian counties of New York and Pennsylvania had collectively higher rates of agricultural depopulation than any region or division of states in the United States except perhaps in the 1910- 1920 period when the average negative change rate appears to have been slightly higher in New England than the Appalachian area.1 1The 1909-1910 farm population was calculated by the author using an estimated number of persons per farm as no census tabulation of farm population was made until the 1920 U.S. Census (Truesdell, 1926, p. 45). 27 Table 2.--Farm Population Change for Appalachian and Non-Appalachian Areas of New York and Pennsylvania 1910-1970. Appalachian Non-Appalachian Year Percent Percent Absolute Absolute Change Change New York 1910 257,589 663,010 1920 221,647 -l4.0 578,180 -12.8 1930 188,665 -14.9 517,781 -10.4 1940 185,753 - 1.5 530,059 + 2.4 1950 166,607 -10.3 411,047 -22.5 1960 94,332 -43.4 224,556 -45.4 1970 52,457 -44.4 138,202 -38.5 Pennsylvania 1910 723,366 327,057 1920 645,737 -10.7 302,816 - 7.4 1930 566,962 ~12.2 279,278 - 7.8 1940 612,120 + 8.0 293,048 + 4.9 1950 487,385 -20.4 217,822 -25.7 1960 217,517 -55.4 133,421 -38.7 1970 126,366 -4l.9 98,872 -25.9 Source: Calculations based on county data obtained from sources listed in Appendix B. 28 .0a .a .oa-oHoH .moomEASmm cofloaasaoa spam .OHmom .4 cw>Hmu cam .n muo> .mxcmm ”oousom n.mmo N.vma e.gvn n.0vu N.~mn m.mma N.h~- h.mmu n.vvn N.nmu o.uv- N.oma m.nn- m.nmu .uum «mm- Nmm- ooh- moH.a- oma.~- men. we“- ham- co"- com- moa.m- Hmm.a- ems- m~m.m- .soga onoa o» comm owcazu m.HNu H.H~u ~.mvn v.wmu n.wmn H.mHu w.m~u ~.omn v.~v- N.~Nu m.mm- m.~ma m.nm- ~.~mu .uum wNN- Hm“- awm.a- «mm.a- mma.a- can- «mm- Mom- aha- mov- oma.v- Amm.~- ~50- mav.a- .soce oom~ cu omaa emcezu h.mHn N.mNu c.0mn «.mmn m.m~n w.o~- ~.omn v.mmn m.mmu ~.m~u m.n~n m.o~u n.mNu m.v~- .uum com- mmm- Nam.a- mm~.a- A~4.H- Nma- mmm- oov- omm- mmv- aom.v- 055.5- owe- mmv.a- . SOFH. omm~ o» ovm~ omcmcu QQOMLDVMNN QVMNNWWNO lo-I H.o .uum BAH mm- vmm and ovH Hum- nmfi on we mm on vmws VNH ma .soge ovm~ ou omm~ omcmzu 0-. O H I m.vs .uum mHH om: Hm med- Nmmu mNHn NmVn Nod- mm- mm mac- mum- ommu mav.a- .soae ommfi o» o~m~ omcmzu ofimfiomm cwmucsoz Hmuucou zusom owe: Hmuucou cusom «mam owucm~u< cusom dauucou :uwoz umez Hmuucou couoz ummm umucm~u< cave“: vcmfiwcm zoz cemmw>fio one: nusom Hmuucou :uuoz ammonuuoz cofimom meumum pupae: mon< .o~m~-o~m~ .mopeooo x9 mcofimw>wa one mconom .moueum woufica .:0wum~=mom swam a“ omcanu--.n eanmh 29 .momsmcou COMHNHDQOQ EOHM Hocu3< xn vmumHDUHmu mwcmc—U acoohom ficm mwcmzu ”monsom ¢.Nm- NVN- «.mm- NNQ- N.NN- ama- m.NN oNN N.NN mad n.5N NNH UNNNuma o.Nm- NNN- o.mm- NNN- o.NN- mvN- o.N- NN- m.N- mN- N.NN NmN cNmocsoz N.av- moo- N.Nm- emo.N- N.om- NNN.H- N.m- NNN- m.~ Na «.2 5N NmNucmu .m.z o.me- NNo.N- o.ma- Noa.a- N.mN- ONN.N- o.m «NH N.N- om- o.N- mo“- Hmpucou .m.m 0.4m- NON.H- m.mv- ANN.N- N.NN- mov.N- N.N mma ~.N- mam- m.m mON uNucmNu< .m N.mN- NNN- H.vN- mam- N.oN- cam- N.N- com- N.N- NHL- m.e- NON- waucou .z.z N.NN- omo- N.om- ovN.~- N.ma- owm- m.N oma m.w- vmv- m.o- Hem- Hauucoo .z.m N.Nm- VNN- N.Nv- owe- N.oN- mom- w.v mmN N.oN- NNN- «.ma- «VN- uNucaNu< .eNz m.Nv- om- N.mm- NNN- N.VN- NNH- 0.0 mm N.o- om- o.ma- NNN- ecaamcm zoz :onN>No m.vm- va- N.om- oHN- N.N~- mam- N.v «o5 o.v om m.ON NNN pmoz N.mv- mam.N- N.om- mNm.m- N.NN- Nv¢.v- a. MN o.m- Nam- O.” HRH Epsom H.0N- Noa.N- e.NN- ovo.N- N.m~- on.N- ¢.N- omN- m.m- Nmm- m.m- mNo- Hmaocou guuoz m.oa- oNN- H.mv- mNN- o.~N- «ma- N.m NHN N.m- aNN- N.ma- Nmm- ammocouoz seamed N.mm- mma.m- N.Ha- moo.m- N.MN- moH.N- N. mm N.m- NON.N- 5.2. new- mouaum eouac: . Hum . 39:. . pom . flock. . Hum .SONC. . Hum . 39:. . Hum . sock. . Hum . SOLE onmfi ea coma emcenu coma on ommd owcenu cmma on ovmu owcmcu ovo~ ou omm~ owcmcu omm~ 6» ommfi emcenu onH OH onH owcwnu mon< .oumauo~mu .moumooa >9 m:meH>No van m:0wmom .moumum pupae: .coMuaasmom swam Hausa :« om:e:u--.v manna Aggicultu Regian De e: e 172—) 1 of the 61 hdt0‘ sented t‘ ighest filmy since lg only agr ”3 at a and p05, ‘;-L3§: 131365t 1940.50 Clase t POPUlat Signs, siderab in the by the but thi autheri Statlls 30 Agricultural Depopglation in the Region During the Depression Decade The 1930 decade represented a rather abnormal time; because of the Great Depression, there was generally a net movement of persons back to the farms (Tables 3 and 4). The Northeast particularly repre- sented this tendency. The Pacific and New England divisions had the highest return rates. Nevertheless, the northern Appalachian counties of New York registered a negative change just as in every decade since 1910; however, Pennsylvania's Appalachian counties had their only agricultural population increase within a decade since pre-l910 and at a rate exceeding any region or division except for the Pacific and possibly New England areas. High Rates of Farm Depopulation in the Region, 1940 to 1970 From 1940 in each decade, the Northeast experienced the second largest rate of regional farm population decline; however, in the 1940-50 and 1950-60 decades the Northeast's rate was unexpectedly close to the rates for the South (Tables 3 and 4). The rates of population loss from the farms within these two regions increased significantly and to record highs during these decades. The con- siderably less attention given to the agricultural population declines in the Northeast than to the South's can perhaps be partially explained by the much larger numbers of farm population involved in the South, but this does not account for the general inattention given by the authorities to the serious widespread erosion of the socioeconomic status of many farm people within the Northeast. The Middle Atlantic par this that 31 part of the Northeast had lower rates than the New England part during this period, but declines involved more people in the former. The Pennsylvania Appalachian loss rate from 1940 to 1950 was similar to that of the Middle Atlantic division, but from 1950 to 1970 the Pennsylvania Appalachian area's loss rates were much larger than the Middle Atlantic's (Tables 2, 3, and 4). Interestingly, New York's Appalachian area had lower rates than the Middle Atlantic area in all three decades. The 1950-60 decade (the period when the writer left the farm) was the time of the most serious negative changes in the northern Appalachian farm population as the loss of the Pennsyl- vania portion at -55.4 percent was of greater magnitude than any region, division, or state except for West Virginia's -62.3 percent (Banks et al., 1973, pp. 16-17). The 1950-60 decade was the period of the greatest loss rates, but although the agricultural depopulation rate for the United States decreased, rates for most areas fell very little; and thus, the ubiquitous depopulation of the farms continued at a very high level. From 1960-70 the New York Appalachian rate actually increased slightly and surpassed the Pennsylvanian Appalachian rate. Clearly the Appalachian areas of New York and Pennsylvania have had extreme losses of farm population over most decades from 1910 at rates that either closely approached those of the areas with the most relative population losses as in the South and New England, or at times as in the periods of 1920-30 and 1950-60 sustained probably the severest farm population declines in the United States. 32 Disregard of the Region's Farm Depopulation; Some Psycho- political Reasons The northern Appalachian region is the largest contiguous area minimally affected by urbanization within the Northeast relative to those places adjacent to the seaboard megalopolis; yet, seemingly the recognition of the importance of the survival of agriculture in the area to the welfare of the Northeast's enormous urban population has been curiously disregarded. The northern Appalachians is the hinterland of the largest megalopolis in the world and its resources have been exploited similarly as those of the southern and central Appalachians, i.e., with the residents bearing most of the socio- economic and environmental costs. Farmer's land resources have been taxed until recently on the basis of potential nonagricultural uses. Much agricultural land has been permanently destroyed by strip mining. The dominance of outside political and economic power accounts for much of the low economic status of the farm population in the northern Appalachians and the consequential decrease in numbers. Why these conditions have been perpetuated and have not been more widely recognized, studied, and confronted probably is closely related to: (1) the urban dominance of politics in the Northeast and (2) the West and South's control of the development of United States agricultural policy. The Northeast's urban dominated government needs to recognize the extreme importance of saving and promoting its regional agriculture and farm population through making certain the food producer is paid fairly for his products; for, in the long run the urban dweller's welfare depends upon it. The urban cost of living within the Northeast 33 would be much lower today if recognition had been given long ago to the beneficial effects of safeguarding and promoting local agriculture. The distress within the farm population and its major causes must be recognized before any corrective action and attitudinal change can be affected. Selection of the Study Area; The Northern Appalachians As illustrated in the previous discussion there exists a large area in the United States that has experienced depopulation over the last several decades (Figure 2). It is to be noted however that these areas encompass a wide variety of geographical settings; and, there are many varying explanations that can be given for pOpulation losses depending on the characteristics of the area being considered. A study of the entire portion of the United States experiencing popu- lation decreases, especially recently, would present an unnecessarily complex study; with numerous complicated explanations possible, such a study would by necessity have to be somewhat superficial. As a com- promise to gain deeper and more complete reasons for the population loss problem, this writer has chosen a region to analyze, for in so doing a considerable homogeneity can be assumed given the usual definitions of a region (Jensen, 1951). A locality, e.g., township, was not chosen for a micro-geographic study because areas of inter- mediate size, i.e., regions composed of counties are preferred for study as published data are more plentiful and accessible than on smaller minor civil division levels. Generalizations are more easily formed and conducive to a theoretically meaningful and comparative 34 report, and the results are produced upon a geographical division ordinarily used in forming policies and instituting programs. A regional study can give a number of generalizations worthy and usable for the eventual formation of a theory of depopulation and population decline. Areas especially prone to emigration or outmigration and thus to depopulation can be identified given some of their usual character- istics: intensity of the outflow of the population, low "resource endowments, and environmental handicaps." It has also been observed in several studies "the smaller the political region, social unit, or community, the more likely it is to be losing population" (Lowenthal and Comitas, 1962, pp. 84 and 86). Although the two case studies of Lowenthal and Comitas are from islands which are mountainous and very small in area, they are pertinent to most depopulated area studies because of the isolation factor. One island is said to have become more isolated because of advances in transport and experienced economic difficulties because of the collapse of its cash crop market (Lowenthal and Comitas, 1962, pp. 88-90). Knowledge of some basic characteristics shared by areas and regions of chronic depopulation can assist in the selection process of a study area; thus the signifi- cance of a study of depOpulation can be recognized in the beginning as well as in the findings. An area in the United States possessing some of the character- istics usually associated with depopulation as previously noted is the Appalachian Region (Figure 3) brought especially to the Nation's attention during the 19605 (U.S. President, Appalachian Regional 35 19 I ?~— ’ NEE: \ ”N 5‘ p. \\\\\\ The Appalachian Region Fig. 3.--The Appalachian Region. 36 Commission, 1964). Numerous studies, dated throughout this century, exist on the Appalachians as is revealed in the Appalachian Bibliog- raphy (1967 and 1972); however, the great majority of these studies are on the southern and central Appalachians, long favored areas of study for sociologists, economists, and other social scientists. Much before the National Government's concentrated attention to Appalachia in the 19605. ". . . national concern turned toward the region with substantial action" during the Depression (Rothblatt, 1971, p. 24). The considerable early public awareness of the problems of the southern two-thirds of the Appalachians resulted in the writing of major reports on this area commonly termed Appalachia (Ford, 1962 and U.S.D.A., 1935). Thus, the southern and central Appalachians have particularly attracted both popular and scholarly attention; whereas, the northern Appalachians, defined for this study as southern upstate New York and western and northern Pennsylvania (Figure 4), have received scant attention. Appalachia refers to the region delimited by the Appalachian Regional DeveIOpment Act and its amendments; and, northern Appalachia includes all those counties of New York State and Pennsylvania found within the Appalachian region as so defined. The New York portion entered the program for Appalachia after the original act, and by way of an amendment, Schoharie County, New York was added in 1967 (Tyson, 1968, pp. 3-4), The reasons for the comparative paucity of research on the northern Appalachians are not easily given, but this scarcity of studies on the area requires some attempted explanations. The northern part of the Region is more like the American mainstream of culture 37 wow< xvsum osu mo mewucaou “maeflnoeaenm< auonuuoz ochuu.v .wflm Jamar N _N v N... a. 2. mm a B mm o m _ mm 9 on _¢ cu m N o. : a. n. we m. mm 3 a .6 mm mm N. a .. v. \ an on Nn km N» nv m t he 8 8 e. a 1. % o 9 on .n m 2. 3 B o. \ on 8 me «e B 8 an o... 3 emce>~>mcccm kox :ofiuauflwwucopa Xucaou v OHJNHL 38 mcwxmeoh «weak censoum NOszcow mcweoxz vcmaoNOEumOz chmz coumcflnmmz coupe: owcwco> sewn: «weak accusesdmsm :m>NH~=m pomuoeom unexcm Naaxxsnum .00 .m0 .m0 .Nm .Hm .om .mv .wv .nv .ov .mv .vv .mv .Nv .Hv .ov oflwmnonom owemuo ohmamaeo ecmfiuuou weapon exam xenon pcmfiponeacuuoz uzoucoz ooucoz cwammfiz Noouo: :moxoz wcweooxq Ocuoqu oo:0h3ma accmzmxowq .No .Ho .00 .mm xuo> 302 .mm .mm .mm .om .mm .vm .mm .Nm .~m .Om .mm .mm .mm aficm>fixm=com owcmconu mcssonu msmsmusmnu «Hammad :Omuommon mammpca :owwcwucsx oceouu confism umouom ouuoxmm owum me cuomzmuu «Mnezfiou coacwau xox :owuaowmwucepm xuczou a ousmua .mm .mm .om .om .mm .VN .mN .NN .HN .om .mH .mm .BH .OH .m~ .vH msmsmnmuueu esooum xcmmofia< vfiowmumo~u cofiuefiu ouucou conumu coonmu «NunEmu weapon vuomemwm uwmam upompam ue>mom mcouum8p< xconmema< .mm .vm .mm .MH .Nfi .HH OTC) o—e HNMQLDONQ 39 and undoubtedly the major differences of the culture of the southern and central sub-areas from the contemporary national culture have continued to awaken the curiosity of many researchers. The northern Appalachians have experienced large scale foreign immigration which the remainder of Appalachia has not. The central and southern Appalachian culture has been less affected by outside forces; thus, scholars desiring insight into traditional American culture find the southern and central Appalachians a rewarding "laboratory." Living conditions in the southern and central areas were perceived undesirable from a national viewpoint; however, some conditions, e.g., the density of family poverty incomes and unemployment in western Pennsylvania (Fuller and Baum, 1965, pp. 8 and 12), and the retardation of economic growth reached the most severe levels in subareas of the northern Appalachians. If negative population growth is an indicator of undesirable living conditions, then the deterioration in quality of life in the north began near the beginning of the 20th Century, much earlier than in most of the other parts of the Region. Early industrial influences in northern Appalachia however may have attracted some people and also had an effect on lowering fertility and thus, population growth. The majority of the southern and central Appalachian counties acquired their maximum populations by 1940 or afterwards (Zelinsky, 1962, p. 501; Hirsch, 1970, p. 90). The maintenance of high fertility rates in nonnorthern part of Appalachia (De Jong, 1968) may have kept that portion of Appalachia from experiencing much negative population change until after 1940; whereas, a number of northern Appalachian 40 counties experienced negative population change even before the first decades of the Twentieth Century. Much of the difference in demo- graphic history between the northern and other sections of the Appalachians can be related to the proximity of the northern Appalachians to the earliest, long dominant, and large northeastern centers of commerce and industry, e.g., New York, Philadelphia, Baltimore, and Pittsburgh, and the accompanying forces of urbanization, modernization, and the subsequent earlier demographic transition, than in the basically agrarian and rural Southeast. The northern Appalachian region, although lightly settled in many areas (Klimm, 1954), and somewhat isolated, has experienced outside influences to a much greater extent than the rest of Appalachia as northern Appalachia's situation made it the crossroads between the first recognized megalop- olis (Gottmann, 1961), the urbanized northeastern seaboard of the United States, i.e., Boston to Washington, D.C., and the second major industrial and urbanized area of the Middle West, i.e., Pittsburgh to Chicago. Northern Appalachia's location between the two largest urban regions in the United States may account for much of the outmigration from the area; however, the relative proximity of the Region to very large metropolitan areas seemingly should have held population in the Region as commuting distances to the nearest SMSA's were favorable (Hathaway, Beegle, and Bryant, 1968, p. 9). Nevertheless, over the years, with the exception of periods when mining or perhaps manu- facturing was being greatly increased, Pennsylvania has revealed its generally low potential for keeping and supporting additional people 41 through comparatively low population growth rates (Simkins, 1970, pp. 52-53), and in recent decades high rates of net migration. Over the years the counties in the Highland area (Figure 5), have particu- larly experienced negative population growth. Not as frequently as the Highlands area, the Appalachian Plateau portions in western Pennsylvania and the southern tier counties of New York have witnessed population declines (Figure 6). It is most significant that Pennsylvania, "the Keystone State," which possesses most of the area defined here as northern Appalachia, led during the 19605 all other states in the number of migrants it contributed to other states, with a total of nearly 400,000 (Taeuber, 1972, p. 8). The estimated loss of population through net migration for the State during the 19505 was 460,000 to 475,000 (Simkins, 1965, p. 183 and Simkins, 1970, p. 53). The State is listed among those states having a low proportion of persons born outside of the State (Petersen, 1961, p. 173) as relatively few people move into Pennsylvania. One explanation for the above population characteristics is the significant rurality of Pennsylvania. Interestingly, Pennsylvania leads all states in total numbers of rural population; in 1960 the State had nearly a half million more rural persons than the second most rural state, North Carolina (Hathaway et al., 1968, p. 27) and by 1970 Pennsylvania's rural popu- lation increased and exceeded North Carolina's by more than a half million (Table 5). Many of these rural persons live within the Appalachian portions of the State, and although proportionately the rural population is not as large as some other states, much of the 42 A Northern Appalachia 8 Central Appalachia 0 Southern Appalachia \\\\\\\\ Highlands area 20° Mi. 8 .6 M Fig. S.--Subregions of the Appalachian Region. 43 .maufinoaaenm< anonuuez as» we mcoflwom Hecanmmumonxnm--. 0 .NE 56E arm n 3626.“. 523.034 nee b.2023 v >o..o> use .02: n 2.25.... N 52¢ .2300 22.6.3 . IIIIIII 0-.- aaaaaaa ........ l ............ ......... ..... I I ,. .. . . .mww ...... {fat-au- - Jar-n no. a- }I {a I l I II a I a? a on. I"- ................ ..... o ...... ..... ........... . ............ 44 Table 5.--Rural Nonfarm and Rural Farm Population by States in the Eastern United States, 1960 and 1970. 1960 1970 Rural Rural Rural Rural Nonfarm Farm Nonfarm Farm United States 40,596,990 13,444,898 45,586,707 8,292,150 New England: Maine 423,999 48,152 464,716 22,840 New Hampshire 234,521 18,634 311,917 9,073 Vermont 191,115 48,845 275,031 26,427 Massachusetts 810,102 35,946 860,043 18,831 Rhode Island 112,635 3,956 121,206 2,359 Connecticut 525,153 24,514 673,183 14,948 Middle Atlantic: New York 2,125,633 324,746 2,441,877 190,659 New Jersey 641,686 51,357 762,468 32,432 Pennsylvania 2,861,417 356,273 3,141,289 225,238 East North Central: Ohio 2,063,722 519,513 2,257,727 370,946 Indiana 1,266,686 485,474 1,448,069 374,590 Illinois 1,377,982 562,845 1,458,822 428,726 Michigan 1,643,125 440,937 2,042,377 277,529 Wisconsin 875,734 553,864 1,093,074 415,206 South Atlantic: Delaware 131,683 21,821 141,583 11,360 Maryland 736,700 110,157 855,116 62,385 District of Columbia Virginia 1,352,340 397,176 1,524,556 192,784 West Virginia 1,028,382 120,938 1,007,267 57,445 North Carolina 1,945,855 808,379 2,421,846 374,692 South Carolina 1,050,054 351,154 1,247,095 111,528 Georgia 1,355,602 407,278 1,651,447 171,544 Florida 1,184,758 105,419 1,250,111 72,261 East South Central: Kentucky 1,137,118 547,823 1,151,565 381,696 Tennessee 1,115,517 586,744 1,300,163 316,817 Alabama 1,068,716 402,855 1,271,951 159,641 Mississippi 814,497 542,839 1,019,277 210,323 * Sources: U.S. Census of Population, 1960, Characteristics of the Population, U.S. Summary, Table 107; U.S. County and City Data Book, 1972, Table 1. Note: Of the western states only four in 1970 had rural nonfarm Populations exceeding one million persons: Texas (1,881,000), California (1,635,000), Louisiana (1,117,000), and Missouri (1,040,000). 4S northern Appalachian area is heavily rural. Many small communities greatly affect the settlement patterns (Carroll, 1971). With the decline of the primary economic activities such as lumbering, farming, and mining, along with the widespread decline in railroads (Warren, 1972, p. 11), a number of Pennsylvania cities lost population over several decades (Table 6) and some small urban places reverted, at least by census definition, to rural centers. Clearly, with many cities declining in population, urbanization explains only a small part of the recent farm depopulation in the northern Appalachians. The rural population, more than a fourth of the Pennsylvania popu- lation in 1970, depends heavily upon the primary activities; and, the recent and contemporary declines in these employments have placed many rural residents in severe socioeconomic straits. In addition to the special rural characteristics of the popu- lation living in the northern Appalachians, the effects of changes in agriculture on the farm population make the Region a worthy setting for studying depOpulation. As noted previously, the farming popu- lation in Pennsylvania and New York composes a very small percentage of the total populations of these states; i.e., 2.4 percent and 1.3 percent respectively for 1970 (Banks and Beale, 1973, p. 5); however, these minor numbers suggest the importance of programs to encourage the remaining farm population to stay in agriculture to prevent these two states becoming totally dependent on others for food. Rationale and Importance of the Study Most of the rationale for studies on rural or agricultural depopulation relate to either or all of the many effects that the 46 Table 6.--Populations of the Major Pennsylvanian Cities in the Northern Appalachians, 1930 to 1970. 1930 1940 1950 1960 1970 Pittsburgh 669,817 671,659 676,806 604,332 520,117 Erie 115,967 116,955 130,803 138,440 129,231 Scranton 143,433 140,404 125,536 111,443 103,564 Altoona 82,054 80,214 77,177 69,407 63,115 Wilkes-Barre 86,626 86,236 76,826 63,551 58,856 Johnstown 66,993 66,668 63,232 53,949 42,476 McKeesport 54,632 55,355 51,502 45,489 37,977 New Castle 48,674 47,638 48,834 44,790 38,559 Williamsport 45,729 44,355 45,047 41,967 37,918 Hazleton 36,765 38,009 35,491 32,056 30,426 Easton 34,468 33,589 35,632 31,955 30,256 Sharon 25,908 25,622 26,454 25,267 22,653 Washington 24,545 26,166 26,280 23,545 19,827 New Kensington 16,762 24,055 25,146 23,485 20,312 Pottsville 24,300 24,530 23,640 21,659 19,715 Butler 23,568 24,477 23,482 20,975 18,691 Kingston 21,600 20,679 21,096 20,261 18,325 Dunmore 22,627 23,086 20,305 18,917 17,300 Monessen 20,268 20,257 17,896 18,424 15,216 Clairton 15,292 16,381 19,652 18,389 15,051 Uniontown 19,544 21,819 20,471 17,942 16,282 Oil City 22,075 20,379 18,581 17,692 15,033 Greensburg 16,508 16,743 16,923 17,383 15,870 Meadville 16,698 18,919 18,972 16,671 16,573 Jeannette 15,126 16,220 16,172 16,565 15,209 Beaver Falls 17,147 17,098 17,375 16,240 14,375 Nanticoke 26,043 24,387 20,160 15,601 14,632 Bradford 19,306 17,691 17,354 15,061 12,672 Duquesne City 21,396 20,693 17,620 15,019 11,410 Source: 1975, pp. 19-20. Pennsylvania Statistical Abstract, 1967, p. 11 and 47 temporary or permanent loss of population from an area has upon the many facets of its society, economy, government, culture, psycho- social behavior, demography, resource use or ecology; and, any such findings may serve as an important input in a successful planning effort. The effects or consequences of depopulation are pg£_the major concern of this writing as much more research needs to be done on the conceptualization of depopulation before its consequences can be more fully understood and explained. The existence of the serious rami- fications of depopulation however provide the most potent raison d'etre for this dissertation. A growing divergence in settlements from highly concentrated to sparse states have significant implications for resource use and management in the United States. Perhaps this traditional trend has been aided and abetted because little recognition until recently has been given to the truth that human beings are the most valuable resource (Schultz, 1962), if for no other reason than their flexible laboring potential to fill multiple needs. This study is prompted by the author's many years of living on farms and in the rural society. For more than two decades the degra- dation of rural society and economy was observed and experienced by this writer while living in the northern Appalachians. For many rural folks especially in certain rural areas of the United States, a "developed nation," day-tO-day living is a real and ever more tiring struggle which worsens with increasing age and loss of sons and daughters to distant urbanized and industrialized places. "The population that remains in such communities exist in a chronically 48 depressed condition and constitute a major national welfare problem" (Bogue, 1969, p. 8). Most rural people have very limited means or know-how to cope with the impersonal forces that have made them the victims of progress and many blame themselves for their failure to succeed. In a very real sense an originally rewarding way of life and society, i.e., agrarian and small town in the United States, has succumbed to political and economic forces and negative effects of change. Change has commonly received widespread support in the United States as being completely good and virtuous, but obviously, these pronouncements are written and spoken by influential persons who have benefited by "progress," for those who have paid the costs and have received minimal benefits seldom gain the opportunity or influence to express effectively to the public and authorities the countervailing results produced by a "change for change" policy. This study grows out of a value base. Any researcher, specifically here a geographer, is inescapably guided by his own values when establishing objectives for an investigation (Harvey, 1969, pp. 3-4). The author's empirical assessment is that the farm society in many places within the United States has suffered extraordinary degrees of deterioration. The disappearance of farming as a way of life in many areas will certainly lower the richness of American life. The public needs to be aware of the dangers of encouraging capital use expansion in farming and of favoring agribusiness. Cheap food should not be the major objective of agricultural policy. The retention of the population in the rural areas to maintain viable communities and 49 to reduce the number of peOple completely dependent on others for all their necessities of life should have first priority. Unemployment will continue to increase if ways are not found to employ more peOple upon the land. Because of the large proportion of the American popu- lation now urbanized the achievement of the above presents a very formidable task, for it is now the urban population whose attitudes and votes will determine the direction of politics, farm policy, and farm programs. A new state of awareness to the farm population's precarious existence and an understanding, appreciation, and respect for farm peOple and the rural society is vitally needed to insure the social and economic well-being of the nation. In this age of stress on equal rights and commercialization, perhaps it is difficult to understand why not everyone should desire to be like everyone else. People are brought up in different social- cultural environments; the learnings and values obtained from these early experiences are carried throughout a lifetime, making the adjustments to a new environment and to a new way of living much more difficult than is usually realized. Of the two main types of forces motivating migration, expulsion, rather than attraction, is believed by Haddon (1912) to be the most influential. "People are psycho- logically reluctant to leave their traditional homes without strong forces forcing them to do so" (Kasdan, 1970, p. 2). The gap between rural and urban life settings is still significantly wide in most of the developing world which composes more than half of the world's population; and surprisingly this difference in natural and social environments still exists within the highly developed nations, 50 including the United States. Whether people are forced or enticed to urban settings from the rural society, there exists commonly among rural-urban migrants a sense of personal loss, a separation and dis- association from the urban society. People need a wide range of choices in life styles and environments in which to live, to most effectively use their abilities and follow their interests. Without conservation of rural ways of living, the probabilities of finding a satisfactory living style for many personalities are needlessly lowered. Could it be that part of the increasing crime of the cities is explained somewhat by the unemployment or inadequate and makeshift nature of jobs and the lack of alternative environments for many persons who would for example prefer more self reliance and a slower pace of living? Other adverse deveIOpments are foreseen with fewer members of the farm society. The control of food production is falling into the hands of fewer people, who are better able to get together, organize, and establish prices. Consumers are however numerous and more diffi- cult to organize for combating high prices. Fewer persons own agri- cultural land as farm consolidation continues. With the loss of direct access to the land, more and more people are finding themselves too interdependent, less self-reliant, and further away from the necessities of life and the resources from which they are derived. A considerable degree of security and independence is lost in one's life when he finds himself completely at the mercy of other people for all necessities of life. As our economy and society becomes more complex, ever more dependent on special interest groups, e.g., 51 truckers, railroaders, food processors, and chemical manufacturers, any failure in one portion of the highly artificial food system, could mean disaster for countless millions of people. Agricultural Development with Social Planning It is interesting to speculate about how United States urban problems would be different today if America's agricultural revolution had been accompanied by some such large-scale program for planning, mitigating, and stretching out the social effects of the depopulation of the countryside (McLin, 1969, p. 11). The preceding quotation refers to western Europe. In the Common Market area, nearly 90 percent of the farms are family operated and in an area one-third that of the United States, an agricultural population lives which is several times the percent of the United States population on farms. Why therefore, cannot the United States support more farm people? Although the farm pOpulation in this part of Europe declined between 1960-1970, there appears a continuance of European hesitancy to removing what many agro-industrialists consider a redundant population (Bracey, 1971, pp. 123-125). The decision to follow an American as Opposed to a Chinese agricultural settlement model has not been clearly decided (Walters, 1973, pp. 187-189), but it appears that in attitude toward saving the agrarian society and family farming, the Europeans are much more determined to do so than the Americans. Yet, even in Europe, the small European farmer represents a simpler society which the agro-industrialists consider not worth preserving. He is redun- dant, and should be paid to get off the land on which his family may have gained a living for centuries. But, to stem the drift of young people to the towns, surely governments should seek not only to preserve the countryside but also to preserve the popu- lations indigenous to it (Walters, 1973, p. 189). 52 It seems clear that an ecological balance in population between rural and urban areas needs to be planned, and the economy planned to make it work. Human Living Conditions and Public Policy Policies using migration as the major change mechanism have resulted in serious hardships for many rural and farm families, through the worsening of rural living conditions. Vast rural areas and rural populations have retrogressed significantly through the absence of a well thought out rural development policy with appropri- ately planned programs. Public policies have had much input from laissez faire economics. Much of the rural difficulties result from past unwork- able and unsuccessful agricultural programs stressing production, especially from increasing economics of scale or bigness of the agri- cultural production unit. Tweeten (1965) has found the large commer- cial farms to be more vulnerable to price decreases in the shortrun than the smaller production units (Milk, 1972, p. 232). The long time emphasis on production probably explains much of the serious neglect of rural human concerns, e.g., no specific United States manpower policy was developed for the displaced agricultural workers (BishOp, 1967, p. 15), who lost their jobs, usually not due to any fault of their own, but often because of the effects of national agricultural policy. With greatly increased production, it seems paradoxical that poverty should still be a major fact of life in some areas and sectors of the United States economy. Many farm people have long had low 53 incomes when compared to the nonfarm workers; and, this "poverty," the low standard of living, and the deteriorating living conditions are almost certainly significantly related to and a major cause of the depopulation of American farms. Therefore, one might ask justifiably: has the basic agricultural policy of the United States proved at all successful, given that there continues to be a large proportion of the farm pOpulation at the poverty level although a very large off- the-farm migration has continued for several decades? Gans has printed a relevant commentary. Clearly, then, poverty and the poor serve a number of functions for affluent groups--households, professions, institutions, corporations, and classes, among others--thus contributing to the persistence of these groups . . . (Gans, 1968, pp. 105-106). General Guides to the Research A major motivating force underlying this research is to determine as nearly as possible the meaning of depopulation. Closely allied with this mental puzzle are also the characteristics of the depopulated state of affairs and the changes through which this general but abnormal condition of life came to be. The word abnormal is used because certainly in viewing an adequate quality of life for an area the existence of the characteristics coming from depopulation would be considered undesirable. Some characteristics of depopulated states, areas, or places can be listed and some of the steps cul- minating in depopulation elucidated. To understand the processes of depopulation, the study of evolutionary events through history is considered indispensable. The events emphasized in the following 54 work are especially related to agricultural changes as the agricul- tural population is the focus of the research. Additional aspects of the descriptive part of this dissertation are to locate and to measure as accurately as possible for the northern Appalachians the Spatial patterns of significant decreases in the agricultural pOpulation during the time periods most affected by the losses. The areal and temporal farm depopulation statistical patterns are useful in establishing reasons for the movement of the farm population out of agriculture. Some of the questions which must be asked and answered to achieve this collective Objective are: (1) What variable, once a clear definition is obtained for depopulation, can be operationalized and can serve as the most accurate dependent variable representing the extent of depopulation--given several constraints, e.g., data avail- ability and variation in the population composition? (2) When did the Appalachian portions of New York and Pennsylvania begin to lose population and particularly farm population? (3) What are the major time periods or eras of farm depopulation in the northern Appalachians? (4) What areas, i.e., groupings of counties, experienced various levels of agricultural depopulation during these periods? (5) What are some of the common agricultural, socioeconomic, and environmental characteristics of each set of counties grouped by the magnitude of the farm population losses? The central objective of this study is to provide explanations particularly for the recently large farm population decreases via a 55 search for significant causes, especially the "push factors" con- tributing to the agricultural population losses. These factors have frequently been neglected in the movement of people out of an area and "pull factors" emphasized and analyzed instead (Sly, 1971, p. 5). As farm people have traditionally identified with and earned their living from the land, probably more than any other group, it seems reasonable to assume that the "push factors" influence their decisions to stay or go from the land more than do "pull factors." As agricul- ture is the major means of livelihood for most farm people, it is presumed that significant changes in the nature of farming, would have great influence on whether a farm person stayed or moved from the farm. The nature of agriculture is viewed primarily as embodying several agricultural systems and or types, each consisting of several criteria or factors basically characterizing the systems or types. The general outward appearance of farming, or type of farming, e.g., dairying, general farm, and cash crop, will have changed ordinarily slightly; however, the arrangement and individual importance of the criteria and components of and within each kind of agriculture or agricultural system, will have changed considerably. In analyzing the effect of agricultural changes on the farm population, one may prefer to regard agriculture as composed of various subsystems and elements; then, such pertinent questions as what are each system's environment, resources, and components can be more adequately answered (Dent and Anderson, 1971, p. 344). 56 The loss of farm population from an area is seen as taking place only under certain inadequate and unsatisfactory conditions. Some conditions of society, economy, and environment, especially related to agriculture, are thus explored and associated with agri- cultural depopulation. A major encompassing purpose of this presentation is to explore a portion of the interface of population and agricultural geography, two subfields or specialties within the discipline of geography that if interrelated to a much greater extent in future geographical research could supply many revealing insights and answers to many of the ills now confronting mankind. Finally, this study raises some questions regarding the wisdom and adequacy of past and present public policies and programs in rural deveIOpment, agriculture, and social welfare. This regional study on northern Appalachia is initiated in part to disclose the inconsistency of policies to assist declining areas through encouraging outmigration. These policies based primarily upon migration from depressed areas generally have proven unsuccessful in bringing rural resources in balance with the rural population because they have lowered an area's ability to use its own resources. In the future it will be more important to predict the consequences of policies and programs on each region's populations before the implementation of such and damage is done to certain groups. Tradeoffs often must be made in establishing solutions but the costs should be distributed as equally as possible. CHAPTER 2 A REVIEW OF DEPOPULATION AND RELATED RESEARCH: A BASIS OF THEORETICAL IDEAS An Interdisciplinary and Chronological Overview The literature cited in and surveyed for this review is not to be considered the total work available on depopulation; neverthe- less, the literature on depopulation that is discussed in this study has come from a number of academic disciplines including geography, rural sociology, demography, agricultural economics, economics, anthropology, history, and the general collective field of government, political science, and planning. Many of the references have been discovered by chance and it has not been possible to systematically search the literature in all the involved disciplines. It is nonethe- less apparent that the tOpic attracts the attention of scholars in many fields; but, there is little evidence of cross disciplinary research and exchange of ideas. The major exception to the straight discipline approach is to be found in the proceedings resulting from the European Conference of the Local Authorities (1968). Certainly, an obvious need exists for annotating and abstracting a bibliography on the whole general topic of rural population decrease. A rather extensive literature exists on various subtopics of rural population decreases which is only considered in this study at 57 58 relative points where it presents important ideas relevant to this thesis. Examples of these subtopics are: rural exodus, declining communities, natural decrease, rural-urban migration, and net out- migration. The French had a very early interest in depopulation (Jaubert, 1767) and French literature (Bertillon, 1897 and 1911; Dumont, 1890; Parodi, 1897) discussed the subject around the era when members of the Royal Statistical Society in Great Britain debated the seriousness and extent of rural depopulation during the late 18905 and early 19005 (Ogle, 1889; Longstaff, 1893; Graham, 1892; Eversley, 1907). An emphasis on the economics of rural depopulation appeared in an article by Roxby in 1912. Thompson (1925) explored the relationships between urbani- zation and rural depopulation in France, and in the United States that had significant rural-urban migration also, related studies were done (Carver, 1927; Goldthwait, 1927). The presence of prosperity probably led many scholars to conclude rural-urban migration con- tributed to the nation's economic growth. One may speculate how the concept of depopulation was intro- duced to the United States; nevertheless, by the 19205 and 19305 when the birth rates plummeted in many of the industrial countries and the extent of rural outmigration was recognized as people began moving back to the rural lands to survive the Depression, sociologists and economists (Gee, 1929 and 1933; Reuss, 1937; Smith, 1938; Spengler, 1938) awoke to the possible effects of rural population losses and "flight from the land." With the continued lowering of birth rates 59 in the 1930 decade, demographers became interested in the general subject of population decline, and Dorn's note on the natural decrease of population in some rural and urban places in the United States (Dorn, 1939) was expressive of concerns over an increasing tendency toward negative natural change in populations, seemingly akin to Spengler's theme. During the same decade geographers expressed apparently the first genuine interest in the phenomenon with an article on mountain depopulation, a subtopic with special pertinence to the geographical emphasis of this study (Toniolo, 1937). After World War 11, publications revealed an increasing geo- graphical interest in studying and reporting on depopulation in various places. A study of southwestern Ontario presented a chal- lenging analysis of sequential developments resulting in general rural depopulation (Watson, 1947). Watson concluded that general rural depopulation during the last half of the 18005 came from neither the occupance and then abandonment of submarginal farm land, nor the general deterioration of the physical environment because the first places to decline were not necessarily the ones with poor soils and adverse physical conditions for agriculture. Furthermore, depopula- tion first began with the migration of the nonfarm population who were influenced by the social and economic changes and ideas originating from the nearby cities. Depopulation continued while the state of agriculture continued to improve during the period. Nevertheless, an out movement of the farm population, starting in the late 19th century and continuing over to the 20th century, gained its impetus 60 both from the attraction and competition of the West and changes in farming methods (Watson, 1947, pp. 147-151). An indication of geographers' expanding concern for rural population decline is the number of their studies published during the 1960 decade (Lowenthal and Comitas, 1962; Field, 1963; Szabo, 1965; Lawton, 1967; Pinchemel, 1969). Lowenthal and Comitas decry however the slight attention given to the widespread extent of depopulation and places that people migrate from, and give examples of places which have experienced emigration and depopulation and some of the consequences that have ensued, in an attempt to ". . . stimulate interest in a neglected and fruitful field of inquiry" (Lowenthal and Comitas, 1962, pp. 83-84). A cursory look at other geography works on population decline during the decade, reveals the global extent of the depopulation phenomenon. Often there are extreme differences in the rural depopu- lation rates within a nation's boundaries. An example is the U.S.S.R. where the Slavic rural people are leaving the countryside whereas the rural minority ethnic groups are generally staying. In addition, a condition in one country which would encourage outmigration from rural areas may not be sufficient to do so in another country. In the Soviet Union there are high rural densities and population pres- sures upon the farm land in many places, but this has not motivated non-Slavic Russians to migrate from the land, a tendency which dis- courages the development and use of new technology to conserve labor (Field, 1963, p. 477). The out movement of the youth from the western parts of the U.S.S.R. apparently continues at a very high 61 level (Murarka, 1975, pp. 1 and 4). Yet, a study of farm depopulation in the south-central plains of Canada, where population densities are low and minority numbers (of non-English speaking origin) are signifi- cant, shows depopulation occurs where there are relatively important interrelationships between depopulation and the economic conditions of agriculture (Szabo, 1971, p. 36). The writing on rural depopulation in France by Pinchemel (1969) is an excellent example of the multi-perspectives geographers can use in analyzing a problem. This recent study of the rural popu- lation declines in France is a very important prologue to the theory, typology, and methodology of rural depopulation; its reading is an absolute necessity for any scholar interested in depopulation of the countryside. Stressed particularly in this article are both the simple and complex factors and types of rural depopulation. Causes of a general or primary nature are considered separately from the secondary causes. Lawton's recent presentation (1967) of the historical geography of rural depOpulation in England in the last century is a valuable sum- mation of findings from the literature and an account of the spawning of rural depopulation in the nation where the Industrial Revolution began. This study is particularly important as it gives possible ways rural areas may be delimited using population data on: (1) density, (2) persons engaged in primary occupations, and (3) certain adminis- trative units. Closely related to the determination of rural areas is consideration of the definition of rural population which is 62 critical to any assessment of the extent of rural depopulation (Lawton, 1967, pp. 230-233). The Nature of the Depopulation Phenomena Definitions of Depopulation Definitions frequently contain information closely related to measurement; this is to be expected as both definition and measure are closely associated--one is needed to make meaningful the other. Nevertheless, in this presentation separate consideration is given to the concepts and the employed parameters of depopulation in an effort to simplify the discussion. In the introduction of this study several definitions, particu- larly ones that could be operationalized, that is put in the form of a measure, are quoted from selected works on depopulation. The main substance of these definitions will be summarized in this chapter. Relative ideas that will add to the clarity and full meaning of the term depopulation will be added by a review of the depOpulation literature. An analysis of the meaning of depopulation in a pertinent work will frequently reveal an author's assumption that there has to be present ". . . a substantial continuing decline in population" (Great Britain Ministry, 1964, p. l) which is "diminishing absolutely" (Ogle, 1889, p. 205 and Lawton, 1967, p. 230) for depopulation to be present. This is to say if the total selected population of a chosen area decreases during a sufficiently long time period, depopulation is Considered occurring. The majority of studies use census time periods, 63 ordinarily of ten years duration and areas which are political sub- divisions. Long term depopulation is generally associated with negative population change. There are a number of additional meanings that authors have given to the depopulation phenomena. A large proportion of studies containing the term, depopulation, in their titles equate exodus or flight from the land and rural-urban migration with depopulation as a process (Bogue, 1969, p. 8; Council of Europe, 1968, Vol. 1, pp. 89- 90; Field, 1963, p. 465; Stockwell, 1968, pp. 269-271). Outmigration is a demographic process directly causing depopulation; however, natural decrease is another demographic process contributing some effects, especially after much outmigration, on population change and the ecology of the source area. Spengler primarily restricted the meaning of depopulation in his study of France to ". . . a persistent long-run excess of deaths over births" (Spengler, 1938, p. 3), and studied negative natural change due mostly to lowering fertility. Such an interpretation of depopulation at the national level is adequate if limited net outmigration has taken place, but studies of a similar nature involving civil divisions of nations permitting freedom of movement must give sufficient analytical treatment of migration if the depopulation phenomena is to be understood. In contrast, if the emphasis is on the role of natural population change in population decline, an analysis restricted to natural pOpulation decrease can be justified (Beale, 1969; Dorn, 1939). In particular, the term depopulation implies studies of a broad and comprehensive scope. Toniolo states: 64 . . . mountain depopulation has been defined as covering not only the larger causes for the decline in population but also "the total or partial abandonment Of territory by the inhabi- tants . . . with the continuous or intermittent shifting of centers of rural life and the deterioration of geographica- economic conditions in the regions under examination" (Toniolo, 1937, p. 473). Extensive conceptualizations of depopulation are found in a number of monographs written on the topic (Great Britain Ministry, 1964; Hutchinson, 1963; Rivers, 1922; Saville, 1957). It is worthy of note that although the term seems to imply an area will eventually be depleted of its population (Browder, 1963, p. 31), Toniolo (1937) interprets the subject as also a "partial" loss of population for an area. The "larger causes" and the remaining content of his statement on the meaning of depopulation strongly imply that he saw the term depopulation as including the process of depOpulation, the various causative processes, and the subsequent effects and resultant con- ditions. When depopulation is assessed as a serious problem, the effects and consequential conditions of depopulation are especially viewed as additional senses of the term. In some studies there is the especially clear implication that the process of depOpulation can be seen at work in some of the changing forms and patterns of settlement (Bogue, 1969, p. 8; Watson, 1947, pp. 152-153). Thus, as will be subsequently discussed, a temporal process is often also a spatial process which merges with evolving spatial forms and spatial structures. Frequently, the social and economic conditions of an area experiencing population losses are seemingly considered as integral characteristics and a part of depopulation (Bollinger, 1972, p. 567; Bracey, 1958, 65 pp. 387-388; Council of Europe, 1968, pp. 89-90; Lowenthal and Comitas, 1962, pp. 89-93; MacDougall, 1973, pp. 57-91 and 123-148; Young, 1972, pp. 289-300). Some of the meaning of depopulation is inseparable from the nature of the population being studied as their characteristics affect the type of depopulation examined, e.g., mountain, rural, nonfarm, agricultural, or central city, each of which has its own deviations from the general conceptualization of depopulation. For example, as the agricultural population is associated with large land areas, there needs to be the inclusion of land abandonment in the formulation of the general agricultural depopulation process; whereas, in the central city situations a primary concern would be the abandoned housing. A conceptualization of depopulation to result in the most meaningful and revealing studies needs to be stated from a human ecological perspective, that is with sufficient consideration given to the role of physical, economic, and social environmental factors in the depopu- lation process. Measurements of Depopulation Population change in both absolute and relative figures for a census period, or from census to census, is the traditional measure of depopulation; however, there are other adequate and perhaps superior measures. The selected parameter is dependent upon the purpose(s) of the research. For example, Stockwell has written . . . the precise impact (consequences) of population decline would differ depending not so much on the magnitude of population 1055 as on the processes through which depopulation was being achieved (Stockwell, 1969, p. 552). 66 Accordingly, he advocates the use of the "ratio of total population loss to migration loss" which means, for example, the smaller the ratio, the more migration loss is compensated for by an excess of births over deaths. Any ratio greater than one indicates both natural decrease and outmigration are occurring. The more outmigration is balanced by natural increase the less serious are the conditional characteristics of the source population (Stockwell, 1969, pp. 552- 555). When over a given time period net outmigration exceeds natural population increase, depopulation has taken place (Gade, 1972, p. 16). An important measure of population losses is density, a con- cept especially favored by geographers because it includes the spatial emphasis and is pertinent to the man—land theme in geography. A recent substantial European report of rural depopulation suggests it ". . . occurs when the population density in rural areas falls and . . . there may be a flight from agriculture without rural depopulation in the sense just defined" (Council of EurOpe, 1968, Vol. 1, p. 10). Geog- raphers studying population losses have taken little advantage of this potentially valuable geographical ratio. There are a few exceptions (Field, 1963; Hartshorne, 1939; Pinchemel, 1969, p. 113). As the depopulation concept implies a population change in association with and an effect upon some territory or earth surface area, a more mean- ingful parameter of population decrease could be obtained by taking into account an often variable land area, i.e., the sum total of land resources. Whether a decrease in the farm population is measured in 67 comparison to total county area, farm acreage, or crop-land acreage could result in quite different assessments of population decrease. A small number of students of population decline apparently find population change too gross of a dependent variable; thus, they may analyze the active demographic components of population change, i.e., migration and natural change (Bollinger, 1972, pp. 572-573; Lawton, 1967, pp. 230 and 237-247). As many writers and researchers have viewed depopulation primarily as outmigration, various possible means for computing migration, representing several levels of refine- ment are possible. The residual method calculates the "expected" population for the end of a census period, given the addition of a certain natural change figured from the difference of births and deaths. The difference between the "expected" pOpulation and the actual population counted by the census, the residual, represents the "apparent migration." A ratio, a relative number required for some statistical tests, may be "obtained by relating the number of apparent migrants to the expected population" at the end of a given time period (Szabo, 1971, pp. 25 and 39). A somewhat general and rough measure used especially in the older rural depopulation studies to reveal the relative loss of popu- lation or to infer a loss of population in the rural areas was either the change in the proportion of the total population that was urban or the rate of increase in the urban population during a given time for each areal unit (Lawton, 1967, pp. 227-228; Longstaff, 1893; Ogle, 1889, p. 205). In a like manner, the loss in rural farm popu- lation can be represented by "the rate of decline in the proportion 68 of the population classified as rural farm from one time period to another" (Leuthold, 1968, p. 7). A measure called appropriately the ”depOpulation ratio" recently introduced in the anthropological literature, may have poten- tially wider applicability in the study of declining populations over various periods of time and for different sized areas. Such a value is ". . . established by comparison of relative numbers of a given group at two points in time. One such time should be that when the population analyzed fell to its lowest numerical strength," termed the "nadir" population (Dobyns, 1966, p. 412). The depopulation ratio is particularly useful in gaining some summary or overall quantifi- cation of population decrease over a long time where the population has reached its lowest points in contemporary times and the maximum of the population is unknown. Various estimates are used for the zenith populations for several regions or areas, and the reasonability of these are tested by comparing the several depopulation ratios (Smith, 1970). This situation is analogous to the American farm population whose official high numbers are unknown except for esti- mates prior to the census of 1920, but perhaps this data could be more accurately estimated by using the depopulation ratio method. Classifications: Types of Depopulation The most common basis for classifying populations undergoing losses is by the relative location of the residence, i.e., rural nonfarm, rural farm, rural, and urban. A similar means for classifying is by the dominant physical feature of an area, e.g., Toniolo (1937) 69 investigated "mountain depopulation" and Hunter (1966 and 1967) explored the reasons for and consequences of "uninhabitable riverine areas." Because residential characteristics are so frequently used to generally describe depopulation, strong geographical associations are evident in the phenomenon's conceptualizations. Three kinds of areas of population decline can be listed according to the importance and the extent of the changes in the com- ponents of population change, namely natural change (births minus deaths) and migration. These are: (1) natural decrease of population and outward migration, (2) natural increase of population and outward migration, and (3) natural decrease of population and inward migration (Great Britain Ministry, 1964, p. 34). In a fundamentally occupational based classification by Pinchemel (1969), consideration is also given to the role of natural change and migration in the depopulation process. Pinchemel seems to be the one author who has treated classi- fication of depopulation with respect. His organization of the phe- nomenon is primarily via occupation, but in addition, he gives much insight to the arrangement of causes, listing the general, comple- mentary and accidental, and secondary factors (Pinchemel, 1969, pp. 113-115). He outlines his defined types of rural depopulation as follows (Pinchemel, 1969, pp. 115-116): I. Non-occupational depopulation A. ". . . excess of deaths over births . . ." B. ". . . departure of young people . . ." II. Occupational depopulation A. decline of the farming class B. decline of the agricultural wage earners C. decline in the rural nonfarm occupations 70 Much depopulation is not general in effect, but selective. As population is the central subject of depopulation research, more effective analyses could be achieved if, as in migration, natality, and mortality studies, additional "demographically-relevant" data, i.e., age, sex, race, ethnicity, and marital status, were used to specify and more fully define the dependent variable (Mangalam, 1968, pp. 4-5). The significant role of family migration through the move- ment of the spouse and children with or to join the head of the family and a change in marital status is clear in the large proportion of internal migration explained by these noneconomic factors (Petersen, 1968, p. 288). Some types of depopulation may be expected following the occurrence of a previous series of events. The decrease in people providing services, for example, could be expected when a decline in the agricultural population occurs within an area especially dependent upon farming. Likewise, one could expect to have natural decreases in a population which had experienced a long history of outward migration. In many places "relative" depopulation could possibly occur for years without much societal damage if it was at a low level. The disruptions in a way of life become significant, however, when depopu- lation becomes absolute and selective or "specific" effecting particular areas, special kinds of workers, and certain groups of people. 71 General Locational and Spatial Patterns of Depopulation Places and areas that continue to lose people tend to be alike in general locational and spatial attributes. Knowledge of these geographical generalities could give important clues to the basic cause of depopulation. In addition, an understanding of the comparative characteristics of the source areas Of migrants could greatly assist the authorities in planning successful programs of assistance for "the people left behind" and the displaced outmigrants. Rural areas have traditionally supplied the surplus human resources for urban growth. Because the population composition of the "countryside" and the characteristics of its cultural and physical geography are not the same everywhere, depopulation has varied through time and space. As rural territories have generally experi- enced more pOpulation losses than urban places, the discussion here will be restricted to the particular kinds of rural residences and groups associated with depOpulation. Rurality, rural-farm, and distance. The more rural an area is, the greater its population losses (Saville, 1957, p. 7). The question that immediately follows, of course, is how are degrees of rurality to be measured? In recent United States censuses of popu- lation rural residences are subdivided into two major categories: rural-farm and rural-nonfarm; however, the latter population group is a heterogeneous residual of the farm population, and as such valid generalizations applicable to it are most difficult to make. Con- versely, the rural-farm population is relatively small, having for the 72 most part groups of persons with similar social and economic character- istics, working principally in agriculture and living in comparable settlements. Some distinct characteristics of the farm population justifies the continued separate tabulation and analysis of data for this group (Beale, 1957). Rural-farm populations associated with agricultural areas in a quasi-historical sense are assumed to represent a higher level of rurality than any of the rural-nonfarm population groups. For example, American farm people live primarily on individual and isolated farm- steads; whereas, miners live usually in agglomerated settlements. One measure of rurality is the percent of the labor force employed in farming, but this puts too much emphasis on the contributions of agriculture to the rural society, when interests of the causal factors of population characteristics extend beyond occupation. Measures of rurality based upon the distance of a county from an urban area and the size of an urban area (Hathaway et al., 1968, p. 7) contain the important influence that urbanism exerts upon rural areas and people. Thus, rurality is conceived as the opposite of urbanism. In retro- spect there appears to be two factors which represent maximums of rurality: (l) agricultural settlement and/or agricultural occupation and (2) location away from the nearest urban area tempered by the population size of the urban agglomeration. The initial urban effect on adjacent rural areas. Many studies of rural population decline indicate that the places and areas most remote from cities have experienced the most continuous losses and depopulation tendencies (Bracey, 1958, p. 387; Lowenthal, 1975, 73 p. 39; Newman, 1967, pp. 50-53; Reuss, 1937). Added to the character- istics of these locations is the sparsity or low density of the popu- lation (Beale, 1964, pp. 268-269; Browder, 1963, p. 34; Council of Europe, 1968, Vol. 1, p. 40). However, areas of the nature just described are seldom the first to experience population decreases when a long period of population decline gets underway. The forces effecting the changes resulting in population losses take time to diffuse to the secluded places, i.e., measured from the urban loca- tions. A geographer studying rural depopulation in Ontario, Canada maintains Depopulation started with the migration of the nonfarm population from townships where village functions could no longer compete with the spread of the city idea and with the extension of city services. . . . the regions which first declined were those that first felt the impact of city institutions (Watson, 1947, p. 148). . . . in the Niagara Peninsula it seems clear that first the villages declined then farms were abandoned (Watson, 1947, p. 150). In the states of Vermont and New Hampshire, the earliest loss of population in the late 18th and 19th centuries took place in the southern rural areas nearest to the largest towns and cities, e.g., Manchester and Portsmouth, and of course, adjacent to the more urbanized southern part of the New England region (Wilson, 1936, p. 21). Near Montreal counties lost pOpulation then regained it because of indus- trialization (Biays, 1968, p. 318). The same occurrence was true in England. In Victorian England in the rural localities near the cities, the losses were soon more than overcome through overspill from the cities (Lawton, 1967, pp. 241-246). In like manner similar areas in Vermont and New Hampshire revealed increases in population through most of the subsequent decades of the 19th century as the 74 depOpulation pattern spread north and became generally more ubiquitous during the 19th century (Wilson, 1936, pp. 49 and 105). Similar spatial developments and patterns can be seen at the national level, e.g., the United States. Zelinsky has observed: . . . . Beginning in a large scale fashion by the middle of the nineteenth century in the northeastern region which can fittingly be called the Core Area of Anglo-America, rapid agricultural and industrial progress led to rural attrition; and soon after a concentric crest and trough of rural maximum and decline began to pulsate steadily outward toward the far corners Of the country (Zelinsky, 1962, p. 522). The above quote is an apt description of the relative spatial movement of the depopulation phenomenon; however, more needs to be added about the coinciding role of urbanization. The nation was being increasingly urbanized during the 18005. If the above theory, namely that the rural areas nearest an urban place are the first to experience depopu- lation but shortly thereafter have population gains or little popu- lation loss is correct, then the dominance of the northeastern urban area of the United States could explain the beginning of rural depopu- lation in the Northeast. Therefore, during the initial stages of population declines the places situated nearest to the urban concen- tration experience rural, and particularly nonfarm population decreases; however, there is evidence, for example in Quebec, where rural pOpu- lation losses are comparatively more recent, that there is a rapid spread of urbanization forces via the service industries to the villates (Biays, 1968, pp. 315-316). A significant portion of the rural nonfarm population in the beginning of the depopulation process retain their numbers less well than the agricultural areas (Keyfitz, 1962, p. 311). 75 Farm population decreases in the mid and last phases. During the first phases of rural depopulation process in a region, the state of agriculture and its required physical resources are of secondary importance in causing depopulation, but become increasingly important during the middle and concluding phases of the population decline. The agricultural population suffers the largest relative decline which continues over a long period of time. The generally long and per- sistent nature of the losses is characteristic of farm depopulation. The lag of the agricultural population's reaction to the assumed urban derived force(s) causing population declines is finally overcome and agricultural depopulation continues essentially unabated, increas- ingly feeding upon itself, i.e., initial population losses become an additional cause of population declines. Because of the large percent of the rural population which was historically occupied by agriculture, the decline of the farm pOpulation is a very important phenomenon to understand as many of the changes in the countryside can be grasped only if this significant "transition" is more fully known. A noted Canadian writer reported ". . . the decrease in rural population is chiefly due to the removal from the country community of farmers' households" (MacDougall, 1913, p. 68). In France depopulation occurred in all of the mainly agri- cultural departments from 1851 to the 19605. One review of the 19605 population data of France reveals that the larger the pr0portion of the population dependent on agriculture in rural political divisions, the greater the population decline, except where the percentage of population dependent on farming falls below 40 percent (Council of 76 Europe, 1968, Vol. 1, pp. lS-l6). Such threshold values could prove useful to the planning and implementation of rural development pro- grams. In England in the 18905, thirty or forty years after rural population began to decline, outmigration was ". . . observed in every agricultural district that lies remote from towns" (Graham, 1892, p. 9). Type of farming. Although there are areal variations in the quantity of agricultural population losses, the literature contains contradictory findings as to areal associations with type of farming. Longstaff (1893) found that localities of varying types of agriculture were affected about the same. Another writer, also studying England, determined that the crop growing areas of the eastern parts suffered the largest declines (Roxby, 1912, p. 184). Some of the difference in interpretation may be according to whether the decrease in popu- lation is seen as changes in total numbers of pOpulation or net migration loss (Lawton, 1967, p. 247). Some types of farming appear to retain farm population more adequately than others, for example, dairy, fruit, and tobacco farming (Bracy, 1958, p. 389; Graham, 1892, p. 8; Watson, 1947, p. 146). A good illustration of the retention of farm people in tobacco farming is--because it is difficult to mechanize--the State of North Carolina which in 1960 led all states in total farm population (Hathaway, 1968, p. 27). Also, changes in type 0f farming may result in a loss of farm population (Watson, 1947, P- 15(D. The above tends to suggest that the more specialized the farming the less the loss in farm population, but there are exceptions. 1“ France, crops requiring many workers, e.g., flax, hemp, silk, and 77 vine crops contracted (Pinchemel, 1969, p. 114) as the cities took rural workers for the industries. However, in Ontario, Canada the ". . . general farming districts depopulation has been prolonged" (Watson, 1947, p. 146). Commercial farming. Another important characteristic of farming seemingly pertinent to all types of farming is the degree of commercialization. For the Canadian prairies, a study on farm depopu- lation and the economic conditions of agriculture revealed certain size commercial farms measured by the value of products sold, spe- cifically those so-called farms, "small commercial farms" and "small scale farms," were the best predictors of agricultural depopulation. The simple correlations respectively were: .648 and .647. These two independent variables considered together in a multiple regression equation produced a coefficient of multiple determination or R2 of .577. Interestingly, the inclusion of the two most additionally relevant variables out of the fifteen tested indicators of economic conditions on farms raised the R2 to only .618 (Szabo, 1971, pp. 28, 33, 35, and 39). The extent to which farms are tied into markets is a clue to the preservation of farms and the farm population. In addi- tion, the measures of commercialization are presumed to be more ade- quate measures of the significance of farm size relative to the success of farming than is physical size of farms, e.g., average size of farms in acres. The association of agricultural depopulation with average size of farm, i.e., the acres of land within one farm unit, is unclear because of the contradictory findings of many studies done on various geographical and administrative levels. 78 Traditional farming. In the United States traditional agricul- ture was primarily subsistence. Actually more than vestiges of the traditional agriculture lasted to a very recent time, i.e., post World War II; consequently any historical study of agricultural population declines must consider the relative strength, continuance, and areal variation of the "old agriculture" during the different time periods. It is quite conceivable that where commercial agriculture is strong, the present greater integration of all persons within the market economy disallows maintenance of the traditional agriculture. In other places part-time farming may be an attempt to revitalize the old agricultural ideals or to find a "refuge" (Rohrer and Douglas, 1969). Watson (1947, p. 145) is of the opinion however, that ". where agriculture is developed on a highly commercial basis . . . (it) dispenses with homesteading and part-time farming," especially when urbanization competes for rural labor. The precedence of situation. Although both the static and dynamic states of agriculture are closely related to farm population decreases, the role of situation takes precedence. The degree of agricultural orientation and comparative strength of the stable and changing characteristics of agriculture in an area is tempered by its relative location to the nearest and most influential urban place. In a substantial study of rural depopulation primarily in the last half of the 19th century in England and Wales, the evidence provided lends support to the above generalizations and reveals, ". . . distance from growth points was more important than the quality of soil or the 79 type of farming in accounting for variations in the intensity or duration of loss of population” (Lawton, 1967, p. 243). The major importance of distance in explaining pOpulation decreases may be seen also in regard to mountains whose physical makeup is such to impede transportation and communication; thus, the "effective" distances are greatly increased in such areas. Conse- quently, there is some degree of association of rural depopulation with selected elements of physical geography. In France it is said, Rural depopulation has been greatly influenced by environmental conditions. Impoverished mountain regions and those with poor soil have suffered the greatest depOpulation (Pinchemel, 1969, p. 117). Many studies have documented the tendency of mountain areas to lose pOpulation. Several examples can be cited which serve to illustrate the differences and similarities of mountain depOpulation. In Switzerland, upland areas belong to two mountainous regions, the Alps and the Jura. The Jura have not experienced much population loss because the early local industrialization has acted to retain the people. In contrast, the Alps have had many townships losing persons for the last hundred years, except where tourism and locations on railroads and highways are important. Depopulation is to be noticed more as a local phenomenon, rather than a general one as farms are abandoned and villages slowly decline. Nevertheless, a repopulation movement is nearly equivalent in strength as small industries and tourism are introduced in some localities (Council of Europe, 1968, Vol. 1, p. 43; Perret, 1960, p. 287). Therefore, probably on a local basis in mountainous regions as much diversity exists as to whether the populations are increasing or decreasin-, if not more so, than in 80 any of the other kinds of physiographic regions. Yet, some additional generalizations appear to be possible. Some simplification of the seemingly local diversity in the population change situation is possible through viewing the land in collective categories based upon elevation. Mountain zones of depopu- lation can be visualized (Pinchemel, 1969, p. 117). Generalization may be taken from events in Italy and France. The spatial aspect of the mountain depopulation process may be described as the gradual movement away from the higher elevations, down from mountains to hills, and then to plains (Toniolo, 1937, p. 477). However, the localities of moderate altitude, defined as below 1,000 meters appear to have the least attractions, and continuously lose people. "The middle altitudinal zones have in fact none of the advantages of the plains and none of the assets of high mountains." In the high mountain areas, settlements that have obtained some tourist industry, have survived. Valleys decline in population if the economy is entirely an agricultural one, but show strong growth if industries exist or are introduced (Pinchemel, 1969, p. 117). In respect to the relative time that these losses in population take place in mountainous regions some knowledge may be gained from the English and Welsh experience with rural depopulation during the latter half of the 19th century. In the upland portions of Mid-Wales, southwestern England including the Cornwall Peninsula, and the northern Pennines, there ensued "constant loss" and severe depopulation although moderate population densities prevailed. These upland areas constituted about one half of the total area having net outmigration exceeding natural gain (Lawton, 81 1967, pp. 239 and 242-244). Mountain regions tend to have compara- tively longer periods of population decline of relatively high rates regardless of the existing population densities. The Depopulation Syndrome: Signs, Symptoms, and Characteristics A few omens of depopulation can be stated; if these are known and sufficiently recognized in an area or place, actions may be taken in time to reduce net outmigration of persons, providing this is the desire of the local citizens and it is in the interest of the general public and the welfare of the nation. As has been indicated in the previous discussion, there are special recognizable geographical characteristics of regions, areas, localities, and places which are prone to net outmigration, and subsequent rural depopulation. Pre- sented below are some additional manifestations of pre-depopulation conditions or precursors of depopulation. Density,_overpopulation, and optimumppopulation. Population density and its moderate quality in the uplands of England and Wales did not preclude the continuous significant depOpulation of these highlands during the last half of the 19th century. Density, of course, is very much a relative term, difficult to use comparatively, and is not very meaningful when based only upon physical area. Zelinsky has suggested the substitution of the "pOpulation/resource ratio"--"the relationship between the size and the technical adequacy of a population on the one hand and the quantity and quality of terrestrial resources on the other" (Zelinsky, 1966, p. 102). By using resources as a point of reference some more effective meaning 82 can be obtained from density and the term "overpopulation," a condition said to be ". . . often the precursor of depopulation in the strict sense of the term” (Toniolo, 1937, p. 476). Overpopulation in respect to a static view can be defined in terms of "optimum population," a number when exceeded creates a situation of overpopulation. The optimum population may be inter- preted as the point of maximum return in relation to any or all of the following: land, labor, capital, economic return, food, income, life expectancy, general welfare, standard of living, military power, full employment, and social advantages. Ordinarily, however, "the optimum population of any area is the number of people which, in the given natural, cultural, and social environment, produces the maximum economic return" (Petersen, 1961, pp. 526-530; Pressat, 1971, p. 107). Finally, some overpopulation is called "technological overpopulation," e.g., when machinery introduced in agriculture results in underemploy- ment and unemployment (Pinchemel, 1969, p. 114). Areas exemplify various types of overpopulation. The revealing question is what kinds of overpopulation are particularly indicative of potential depopu- lation tendencies? Unemployment or underemployment is one of the best indicators of overpopulation, but unfortunately, the data for farm population is incomplete and unreliable because of the nature of the agricultural business, i.e., persons may survive on farms via "income in kind" produced on the farms. Undesirable employment conditions exist where technological overpopulation has occurred in agriculture. Farming tends to become more extensive, thus using still fewer workers 83 (Toniolo, 1937, p. 474). As might be expected the farm laborers decline significantly and are the first to be affected by mechani- zation of the agricultural operations (Lawton, 1967, pp. 246-249; Ogle, 1889, pp. 222-223; Pinchemel, 1969, p. 116). Mechanization and farm consolidation. Once machinery is incorporated into the farm system's operations, farms need to be enlarged for the efficient use and achievement of economies of scale in the use of this expensive capital investment. Although mechani- zation rates and levels vary considerably by type of farming, they affect the rate of farm consolidation which is an indication of sub- sequent agricultural depopulation as some farm people will sell their land to an adjoining farmer and reserve use or keep property ownership of the farm house and its lot. Both the degree of mechanization and farm consolidation vary according to the type of farming. Because the types of farming best fitted to a region depend significantly upon a region's particular mix of resources, it is to be expected that farm depopulation, a nationally wide phenomenon, examplifies great vari- ations in the rate and amount of the agricultural population loss within regions and type of farming areas (Browder, 1963, p. 31). In south central Idaho which has two areas with quite different histories of rural farm population decreases, one area's irrigation costs pro- hibit the acceptance of the additional costs of farm consolidation (Bollinger, 1972, pp. 570-571). Community tensions and social disruption. One study fittingly sums up the "atmosphere" permeating a community undergoing the various 84 processes leading to depopulation, including mechanization and con- solidation, as "tensions" (Rundblad, 1957, p. 184). Perhaps because of the use of machinery, farming is increasingly done alone and with less social interaction among farmers (Bausman, 1904, pp. 278-279). Increased dependence on self in farming may mean fewer community duties and "chores" done together. The additional costs of machinery and more land gives a sense of financial insecurity. Community cooperative action dissipates, e.g., in schools and road maintenance. ". The break-down of relations inside the community itself (is) followed by an increased opposition to the outside world" (Rundblad, 1957, pp. 187-190). Social isolation and the need of alternative employment. The inward-looking nature of some of the communities that are destined to be affected by the depopulation process causes an isolation greater than that of physical distance. This attitudinal condition is further exasperated by a serious lack of alternative jobs (Bracey, 1958, pp. 387-388). Some areas are more fortunate in that part-time jobs are available within commuting distance; thus, a healthy exchange of ideas with outsiders is facilitated. Mountain areas especially have an absence of supplementary industries, e.g., forestry may not even be deve10ped in a primarily agricultural area (Toniolo, 1937, p. 477). There is a need for a wider selection of work than that provided by primary activities, e.g., farming, lumbering, and mining, as these industries may decline at the same time. Simultaneous and serious declines in mining and agriculture were experienced in Wales in the late 18005 (Great Britain Ministry, 1964, p. 3) and in the northern 8S Appalachian region of the United States during 1940 to 1960. In the latter, changes in total employment ranged primarily in agriculture from one third to two thirds decreases and in mining from one half to over three fourths decreases (Royen and Moryadas, 1966, pp. 187-190). In addition, profits in primary and resource extractive activities and in farm products tend to be among the lowest because the site of extraction frequently does not share in the returns of processing. The most common factor of areas that will suffer depopulation is declining employment opportunities (Saville, 1957, p. 7). The loss of economic activities in the villages and some towns especially restricts the potential choices of occupation in the countryside. Selectivity of_persons in outmigration. The characteristics of a rural area prone to outmigration and depopulation particularly affect certain components of the areas' population and this is particu- larly true because of the few choices in employment. Career oppor— tunities are limited for young people. The highly educated must move away to use their knowledge and skills lucratively (Great Britain Ministry, 1964, pp. 2-4). But, without doubt, if we assume ". the possibility of earning more money is generally pre-eminent among the several operating motives" (Thomlinson, 1965, p. 225), and ". we consider job Opportunities to be the driving mechanism of migration . ." (Beshers, 1967, p. 142), a rural area having a predisposition for population loss can be expected to have an outmigration which is predominantly female. The feminine preponderance within migrants from rural areas is to be expected as part of the overall tendencies in human 86 migration: the female is more migratory than the male in rural—urban migration and short-distance moves (Ravenstein, 1885, pp. 196-198; Thomlinson, 1965, p. 227). Nevertheless, the greater loss of females from the countryside can be much more fully explained by the fact, given the traditional sex selectivity of jobs, that fewer positions are available in the rural areas for women than men; thus, a high sex ratio is characteristic and ". . . is the most striking demographic conse- quence of rural depopulation." The push or expulsion factor of declining employment opportunities appears to affect women more than men in a rural community predestined to experience depopulation (Saville, 1957, pp. 32-2). The female portion of the population is composed of several groups, each exemplifying somewhat different migrational behaviors. Young single girls commonly leave the rural districts before the boys (Rundblad, 1957, p. 186). Historically, the life of a "country woman" was a very difficult one (Council of Europe, 1968, p. 64). The inadequacy of social, cultural, and domestic amenities and ser- vices in rural areas is believed to have influenced many women to leave rural life in Austria, Turkey, England, and Wales (Council of Europe, 1968, pp. 56 and 86; Saville, 1957, p. 36). The single girl would be somewhat more motivated by the desire to find employment than the married woman; thus, when explanations are given for outmigration based upon employment desires, note should be made of an individual's marital status. A significant amount of internal migration in the United States is associated with individuals as unemployed members of a 87 family and change in marital status (Petersen, 1968, p. 288; Rossi, 1955). Farm land ownership. An interesting question which apparently no one has asked while studying decreasing rural populations is whether the traditional female sway over outmigration from the farms and villages within a nation is associated with land tenure. In farming it is most often a son who gains ownership of the family farm. This is not a question which will be addressed in this study, but is stated here for the implications which perhaps could make the results of this study more meaningful. The changes in land ownership can be indicators of eventual depopulation. The question needs to be asked, as to who are the new owners? Are they nearby residents? Frequently, realtors are interested in out-of-state buyers or urban buyers who either have the ready cash or sufficient credit to meet the listed price of the rural property. Many of these buyers are not interested in a farm for a permanent residence, but use it for speculative purposes, hunting, or as a summer retreat. Therefore, often the land is kept from the local population who needs its resources because they are not able finan- cially to pay a high price for the property. Farmers sometimes rent the land, but this is a less satisfactory way to the land resources as farmers are reluctant to use lime and proper fertilization when the property is not theirs and they might not have the use of the land the following year. Thus, the yields from such land is much less than could be under a farmer's ownership. Absentee owners with 88 nonagricultural objectives may allow the cultivable fields to become brushwood. The passing of land into the hands of absentee owners is a definite characteristic of areas on the brink of depopulation or already encountering population losses; and, such an occurrence and state can be regarded as a "colonial phenomenon" (Council of Europe, 1968, p. 105). The above situation is said to be common in France. One of the best known and documented histories of absentee ownership was in Ireland where Englishmen owned estates. Hart (1963) has written of other similar tendencies in Northeastern United States. Estate and recreational farms are a conspicuous feature of the rural landscape of Megalopolis (Gottman, 1964, pp. 310-319), and the spread of these land uses into the Appalachians displaces the farm population. Within a region or area, however, there may be distinct differ- ences in wealth of the indigenous farm population; some of this vari- ation is associated with the settlement decisions of their ancestors a few of whom chose more favorable sites than did others (Pearson, 1935, p. 214). The farms of the fertile and level plains and valleys contrast with the farms of the less fertile soils of the steep, irregularly shaped, and small fields of the mountains. Wherever financial differentials exist among the farm population of an area, unless there is some governmental attempt to correct the difference in resources, e.g., in special tax provisions based upon value of agricultural production rather than on potential nonagricultural value--a new movement in the United States to preserve open space-- one may expect local depopulation to take place in the hill lands as 89 the good plains and valley sites have standards of economic success, the break even points, higher than the capabilities of the physically disadvantaged lands. The Basic Causes of DepOpulation The Major Demographic Aspects Depopulation may be viewed geographically as the removal of people, either through natural decrease and/or outmigration, from any place or area during any given time. Demographically speaking, depopulation is made up of three basic demographic processes: births, deaths, and migration, through which population decline occurs. To understand depOpulation it is necessary to contemplate carefully the "interaction pattern" among these demographic processes (Stockwell, 1969, p. 555). Births and deaths may be regarded aggregately. A natural change figure results. The type of natural change relevant to depopulation is natural decrease or "biological dissolution," previously associated with some U.S. cities (Born, 1939), but com- paratively a new and rapidly spreading phenomenon in the rural United States (Beale, 1969). Natural decrease in a rural population is a development nearly unprecedented in modern times, and it has affected particularly the farm population, that has ordinarily had high fertility and birth rates. This reversal of an old traditional demo- graphic truism has been brought about by the strong action of a second major demographic process, and that is migration. Outmigration from the farms and rural areas of the United States has continued over a relatively long time. It took place 90 especially in the East, even as settlement of the interior of the country, commonly called the West, was uppermost in the minds of Americans. "Once emigration has begun from whatever cause, factors are introduced which tend to maintain the movement" (Bracey, 1958, p. 390). Consequently, several generations of rural Americans and especially the farm population have had their group replacement capabilities lowered significantly by outmigration, giving evidence of its special selectivity of the younger age groups with the greatest reproduction potential (Beale, 1964, pp. 269-271). The resident popu— lations of many rural areas have had their age structures altered by the aging of the population through the greater emigration of the young (Bracey, 1958, p. 389) and through the loss of the children these people had elsewhere. Eventually outmigration, in addition to causing the actual loss of population from an area, if at high enough level and rate, will be often the primary demographic reason for the occurrence of natural decrease. The ultimate magnitudes of depopu- lation are reached when both natural decrease and net outmigration are taking place at a high rate in an area, resulting in an absolute loss of population. Contributingpforces and a Theory» A number of continuous cultural forces with economic, social, and political facets feed depopulation. Considered here are the most general and universal of the contributing forces. A few conjectures obtained through a temporal approach and 1isted chronologically give support and guidance to the organization of the remaining part of the literature survey. This is in part an 91 exploratory study and the generalities and general hypotheses given should not be regarded as absolutes, but are some conclusions, drawn by this writer from his studies, in need of further study and testing. The major factors that attend and affect rural and agricultural depopulation are thought to be as follows: (1) To be susceptable to moving out of an area, a rural or farm population first has to know what its relative welfare is, and the clues to this come from contacts with outsiders, either through friends, relatives, strangers, mass communi- cations, and/or travel. The process of rural depopulation and its variations in time and space may be thought of as the spread of an idea, namely the idea to leave isolated, low income areas for large villages, towns and cities (Norling, 1960, p. 233; Gade, 1970, p. 77). (2) The pull forces operate as the major causes of outmigration in the beginning, affecting primarily the members of a popu- lation who are dissatisfied and have high expectations. (3) After outmigration has occurred for a few years the various types of conditions, e.g., environmental and socioeconomic, are established; thus, the push forces eventually play the major role in depopulation. As outmigration is the major demographic process bringing about depopulation, some migration theories may be supportive of the chronologically determined importance of pull and push factors. Much of migration occurs by stages with only short distances involved. The areas immediately adjacent to a significantly growing city are first affected by outmigration to the city, and a shifting of rural 92 population evolves inward toward the city, ". . . until the attractive force of one of our rapidly growing cities makes its influence felt, step by step, to the most remote corner . . ." (Ravenstein, 1885, pp. 198-199). One inference is that the pull forces of the city initiate the outmigration and dominate the first phases of the pro- cess. An author in historical demography has written: Some evidence has been found that suggests that "push" factors (that is, poor conditions at the place of origin) play a larger role in long-distance migration than in short-distance migration, while "pull" factors more particularly affect the short-distance movements (Hollingsworth, 1969, p. 4). Thus, it seems that pull forces initiate a period of outmigration as the early movements primarily are over short distances that are more associated with the pull forces. Push factors, however, tend to bring about migrations over longer distances than do the pull For example, the Black migration from the Deep South to Northern cities was caused primarily by push factors. Push forces vary in strength for several reasons: (1) Unlike pull forces which come from various places, push forces come from really only one location, the residence and community of the migrant. (2) The interactions of the earlier population losses with the residential environment result in a particular set of con- ditions for any given area. The push forces affect the source populations differently, culminating in different rates of outmigration from area to area. The contra- dictory findings on the selectivity of migration can be explained by studying the diverse conditions of the home areas. rat re: COL cor As st: of 160* St '30 an, ab 1e; f0 93 If there exists more variation in the rural or farm population losses in the study area during the mid 19005 than the late 18005 and the first two decades of the 19005, this would tend to substantiate the proposed theory: that after the pull forces have dominated for awhile, the push forces exert the predominate influence, and their variant strength is revealed in the differential population loss rates. A major study of rural depopulation in England and Wales revealed that during the first half of the population losses from the countryside, the total decreases were ". . . remarkably general and constant throughout the period 1841-1901 . . ." (Lawton, 1967, p. 229). As England and Wales are about the same size in area as the largest states of the United States east of the Mississippi River, the nature of rural pOpulation declines there perhaps are similar to those of a large or moderately-sized state in the eastern United States. Geographical Level of Analysis If the level of analysis is carried out on smaller adminis— trative units than the state, e.g., counties, more variation in the dependent variables is to be expected. An analysis by township level would reveal generally more variation in population change than a study done of counties. Consequently, in analyzing depopulation it is most important to remember that generalizations from one level of analysis cannot be safely applied to another level without consider- able testing. An interesting effect of differences in area size and length of time on the basic causes of depopulation is summed up as follows: "The smaller the region and the shorter the time-span, the .' l'll‘l. Q-c‘ 94 more migration will preponderate over natural population changes" (Hollingsworth, 1969, p. 2). The demographer's definition of migration relates to the changing of a person's usual residence for a considerable length of time, while traversing a governmental boundary in the process (Thomlinson, 1965, p. 211). The smaller the governmental units, the easier it is to cross a boundary, and thus be classified as a migrant. Yet, another aspect of an analysis of rural population decreases that is closely associated with size of the areal unit of observation is the general nature of the contributing (explanatory) process. Emphasis on small area analysis, e.g., by county or township, means the "push factors" would tend to be more important in explanations of migration because the major spatial variations, i.e., differing conditions from place to place, would be revealed. Processes work differently at the various areal levels (Sly, 1971, p. 29). Therefore, generalizations on migration and thus depopulation must be made in reference to spatial units or administrative areas. Summation of the Temporal Importance of Pull and Push Forces In summary, the contributing processes of rural and farm depopulation are seen as belonging primarily to one of two large, encompassing, and universal processes: the pull and the push develop- ments. Pull forces originate at the place of destination and push forces operate at the place of departure. Although factors influencing outmigration and depopulation are considered in this study as either of One or the other of the basic types, note needs to be taken that 95 some variables may operate as either push or pull factors: technological changes (the factory system pulling peOple to cities and the tractor creating a surplus agricultural popu- lation pushing people out of rural areas) . . . (Thomlinson, 1965, p. 224; Nelson, 1955, pp. 124-125). Other examples may be given that seemingly illustrate that pull and push forces are the extremes of the same force, i.e., the presence of a factor at one site and the slight presence or nonpresence of that factor at the other site. Interaction occurs between a pull force and the potential migrant and he reacts according to the environmental elements and conditions of his place of origin, i.e., the push factors. Pull and push forces are rarely the opposites of the other. Pull forces exert the primary influence in the first part of a depopu- lation era; afterwards the push factors explain most of the popu- lation losses. Lastly, the relative importance of pull factors and push factors is related to some degree to the areal level of analysis; and, generalizations about the causes of depopulation need to be made in relation to a kind of area or place. CHAPTER 3 PROCEDURE OF THE INVESTIGATION: AN APPROACH TO REPRESENTATION AND EXPLANATION OF FARM DEPOPULATION IN THE NORTHERN APPALACHIANS A Specific Statement of the Problem There is general recognition that there are now fewer persons on American farms than in the past and that thousands of farm persons have left the farms never to return; however, the magnitude of the farm depopulation and its implications have been only slightly recog- nized. To some Americans, the farm person is a dispensable individual. Underlying this presentation is the author's view that increasing the public's awareness of the significant and serious extent of the agricultural population decline necessitates much pre- paratory research. The whole depopulation phenomenon has been inade- quately studied. In addition, the adverse consequences of the farm population decreases are among the most potent reasons for such a topical study, along with public policies which at least until recently either encouraged, or by their omissions, worked in favor of the depopulation of American farms; however, the aspects investigated are more fundamental to the eventual clarification of the entire phenomenon of depopulation. The essence of this study is the 96 97 exploration of the general nature and causes of agricultural depOpu- lation; the areal and/or spatial patterns of the farm population losses; the relevant systems, types, and resources of agriculture; the areal distributions of agricultural systems, components, and character- istics; and, the latter's associations with the farm population declines. Changes in agriculture have prompted an unknown proportion of persons to leave farming and dissuaded others from staying who would have ordinarily or naturally entered into farming with the inheritance of a farm. The scope of the analysis thus is primarily limited to the spatial distributions, areal patterns, and interrelationships among selected agricultural variables and farm population changes in the chosen study area, i.e., the northern Appalachians, during the Twentieth Century. Much preferred is the analysis of the farm popu- lation situation within a region as a greater depth of understanding and more plausible explanations can be obtained in a more geograph- ically circumscribed area. Farm Population; Choice of the Dependent Variable The measure selected to represent farm depopulation, i.e., the dependent variable, equals the percentage a county's farm popu- lation is of an earlier time. The superiority of this measure to potentially other useful variables is that aggregation of the total population situation is achieved, taking into consideration effects of migration, deaths, births, and adoptions. Thus, the overall demographic circumstances of the population living on farms at one time can be compared to that of another time. More succinctly, the facts desired 98 to represent the problem are data revealing whether the farm popu- lation of the Region's counties had individually declined, and if so, by how much. Definitions and Criteria of Population Censuses The farm population is defined for purposes of the U.S. censuses of population as that population living on farms (Truesdell, 1960, p. 1). Whether an individual residing on a farm or members of a so- called farm family qualify as persons engaged in agriculture is not easily determined. Prior to 1960 farm residence was primarily resolved by the answer to a subjective question generally referring to whether the house was sited on a farm or ranch--the answer to which the same enumerator for the censuses of agriculture and population often sup- plied (Shryock, Siegel, and Associates, 1973, p. 171). Before 1960, "no specific criteria of acreage operated or value of products sold from a place were used to classify farm population." The pre-196O method of determining farm population is believed to adequately illus- trate the conditions existing in the years of 1940 and earlier, when subsistence farming was widely prevalent (Banks, Beale, and Bowles, 1963, pp. 10-11). A major reason for using a new, objective, and more specific definition in 1960 was to produce a definition of farm residence nearly consistent with the new definition of a farm begun in the Census of Agriculture of 1959. A somewhat condensed version of the definition of a farm than that used in the census of agriculture of 1959 was necessary for enumerating the farm population. 99 In the 1960 census, the farm population includes persons living in rural territory on places of 10 or more acres from which sales of farm products amounted to $50 or more in 1959 or on places of less than 10 acres from which sales of farm products amounted to $250 or more in 1959 (U.S. Census of Population, 1960, Vol. I, Part I, p. xxxvii). Surprisingly, before 1960 ". . . no definition of a farm was provided in the census of population per se," although the farm population has primarily always been defined as the population living on farms (Shryock et al., 1973, p. 171). The essence and major result of the change was ". . . to exclude from the farm population persons living on places considered farms by the occupants but from which agricul- tural products are not sold or from which sales are below the specified minimum" (U.S. Census of Population, 1960, Vol. I, Part 40, p. xv). In other words some persons not engaged in commercial farming, as defined, but who lived on "farms" were excluded from the census of farm population. In addition, farm population in urban areas were not counted. Tests of the effects of the definitional changes revealed that the 1960 farm population was reduced by around 21 per- cent compared to the count which would have been obtained using the old definition (U.S. Census of Population, 1960, Vol. 1, Part 40, p. xv). Thus in explaining farm population change in the 19505 this reduction in numbers of farm persons "on paper" must be noted. Major changes continue in agriculture; and, as agriculture becomes more characterized by commercial, agribusiness, or organic enterprises, the types of statistics needed for manufacturers, traders, buyers, investors, and other businesses and governmental concerns will change as in the past, inducing changes in the minimum criteria to represent a farm. The persons counted as farm population will 100 correspondingly change as long as the definition of farm pOpulation is associated with the definition of a farm. Nevertheless, in 1970 the same basic definition of a farm as given above formed the basis of the census definition of the farm population (Banks and Beale, 1973, p. 7; County and City Data Book, 1972, p. L). The new standards implemented in the census helped to better identify those persons who were closely associated with agriculture. The major changes in the definitions of farm and the farm population in 1959 and 1960 were prompted by the increasing numbers of peOple who resided in the rural and open country on former farms but whose livelihood no longer depended directly upon agriculture. From 1920 when the farm population was first listed separately, as more farm residents were drawn into nonagricultural employment, the assumption that the farm population matched the agricultural population became less correct (Shryock et al., 1973, p. 171). Presently, the farm population, as defined by the Census, is not composed entirely of those individuals economically dependent upon agriculture because some of the persons living on farms are part-time farmers and their dependents, members of farm operators' families living on the farm who work in nonagricultural occupations, or farm wage workers and their families who earn some of their annual income in nonfarm occupations. These groups at least have some connection to agriculture; whereas, the people who live on farms but earn all or nearly all their living in nonagricultural occupations, who have mostly been discounted as farm population, have no other association to agriculture except they live on a farm. As agricultural operations have been abandoned on 101 many farms, nonagricultural people have frequently rented farm houses and houses on the farms formerly the residences of farm laborers, or tenants and blurred the earlier close association between farm resi- dence and economic dependency upon agriculture. Some changes in definitional criteria, and instructions to the enumerators starting in 1950 tightened the relationship between the counted "farm population" and their association with agriculture. Persons on farms who paid cash rent for their house and yard only were to be classified as nonfarm. Also, persons residing in housing of institutions, summer camps, motels, and tourist camps located in the open country were to be regarded as nonfarm. With these changes about 8 percent fewer persons were classified as farm residents than would have been under the former criteria (Shryock et al., 1973, p. 171). An additional reduction in the farm population occurred in both the 1960 and 1970 censuses when unmarried farm youths enrolled in colleges away from home were enumerated as residents of the communities where they were attending college (Banks et al., 1963, p. 10; Banks and Beale, 1973, p. 7). The Census continues to rely on the concepts of farm and farm residence. There have been in recent censuses of population major changes in the criteria for the enumeration of farm persons which have generally made agricultural relationships more characteristic of the designated farm population. A 1966 Census study showed that nearly 40 percent of the employed farm residents worked in nonagricultural activities, but the proportion that such economic pursuits contributed to these farm persons' total livelihood was not reported. Although 102 many farm residents have had some ties to agriculture, the results of the 1960 Census, even with changes in the farm population criteria, displayed a large number of farm families who were not primarily dependent upon agriculture (Shryock et al., 1973, pp. 173-174). Analysis of the percents of farm operators working off the farm data from 1939, 1949, and 1969 for a minimum number of days helped to counteract the nonagricultural aspects of the farm population data and resulted in a relative assessment of the role of nonfarm employment in farm depopulation. It is not the writer's intention to suggest that the farm population should be completely or primarily dependent upon agricul- ture, although if such data were available, the relationships between population and agricultural changes found in this study would be more pronounced. For national policy development, state planning, and the public's welfare, however, the specification of the relative degree of dependence upon agriculture would greatly aid the formation of governmental programs and the enactment of legislation. In fact, regular collections of data on part-time farm persons' characteristics would aid in learning much about persons partially dependent upon agricultural activities. The Quality and Source of Decennial Farm Population Data Most of the above alterations in the criteria for the census enumeration of the farm population resulted in, from the point of view of this study, a higher quality of dependent variable data available for 1950, 1960, and 1970 than for the earlier years. Thus, 103 the more farm population change data represented persons primarily engaged in agricultural activities, the stronger were the relation- ships between the agricultural components and the farm population changes. The sources of the farm population data were of necessity the various U.S. censuses of population. Although additional sources of data were available, such as the farm population estimates made by the Department of Agriculture since 1923 and also the reports of the Current Population Survey, data were given generally on larger geo- graphical areas than the county level desired for this study. The 1910 farmypopulation data. Although no direct count of the farm population data was available before 1920, an accounting of the status of agricultural population before and after World War I was important because of the rapid industrialization, mechanization, and technological changes that greatly affected agriculture. Truesdell (1926) had estimated each State's farm population in 1910 by the use of: (1) the number of farms in 1910, (2) the average farm population per farm in 1920, and (3) the change in the average number of persons per family in the rural population between 1910 and 1920. The average change in the size of the rural family was assumed to be representative of change in the average persons per farm, given the assumption of one family per farm. As the farm family was generally larger in 1910, this difference in the rural family size was added to the 1920 farm family size to obtain the 1910 farm family size which was then multi- .P1ied by the number of farms in 1910. The 1910 county farm popula- tiJJns for the northern Appalachians were calculated by multiplying 104 the number of farms per county by the state‘s average population per farm, 4.79 and 4.27, obtained by Truesdell for Pennsylvania and New York, respectively. The 1920 farmypopulation data. Because the 1920 data by county was unobtainable and comparable data was desirable for the World War I period, averages of the farm population per farm, 4.69 and 4.14, for Pennsylvania and New York, respectively, again were calculated and multiplied with the number of farms found in each county. The 1920 data on farm population by states were not published in the regular volumes of the census of population but appeared in the General Report on Agriculture and in Truesdell's Farm Population of the United States (Shryock et al., 1973, p. 171). Apparently, only special tabulations for the farm population of a few selected counties of the United States were ever published (Truesdell, 1960, p. 2). The 1930 farm pOpulation data. The 1930 U.S. Census data represented the "total rural-farm population" as recorded from the affirmative responses to the question, "Does this family live on a farm?" (Truesdell, 1960, p. 3). The gathering of the data soon after 1929, the beginning of The Great Depression, probably prevented much disturbance of the data by that momentous occurrence. A few counties showed increases in their farm pOpulation based upon the estimates of the county farm populations in 1920. The 1940 farm population data. Again, in 1940 the gathered U.S. Census data pertained to the "total rural-farm population." As ten years previously, either the census respondent and/or the enumerator 105 decided if the residence was located on a farm and in questionable cases the enumerator was guided by the rule stating that places of less than three acres had to "have sold or produced agricultural pro- ducts of a specific minimum value." These types of places however constituted less than one percent of the total farms before 1945. The 1940 census count of the farm population appeared to represent a "substantial overcount" according to the much larger number of "extra" families per 100 farms (Truesdell, 1960, p. 9). The 1950 farm population data. As in several of the previous censuses, the rural-farm population was obtained by affirmative answers to a subjective question, "is this house on a farm (or ranch)? (U.S. Census of Population, 1960, Vol. 1, Part 40, p. xv). The definition for the urban population changed in 1950 which resulted in a 12.2 percent decrease in the rural population nationwide (U.S. Census of Population, 1950, Vol. 1, pp. xv and xvii). The rural population is defined as that population which is not urban. As the farm data used applied only to those persons listed as rural, and as the rural popu- lation declined because of the change in the definition of urban population, the farm population declined some because of the urban definitional changes. In 1950 the Census adopted the urbanized area and unincorporated place concepts to aid in the count of urban persons (U.S. Census of Population, 1970, Vol. 1, Part A, p. x). Both the rural-farm data under the old and new definitions of urban places were collected and used, the former for comparison with 1940 and the latter as a base of comparison with subsequent counts. The change in the urban definition affected the residence of less than 1 percent of 106 the 1950 farm population determined by the old definition. Some of the more specific criteria for excluding some nonagricultural popu- lations as given above accounted for an undetermined percentage of the farm population decline. The 1960 farm population data. Dependent variable data for this year represented the rural-farm population and was calculated by the use of the percent the rural-farm population was of the total pOpulation as given by the Census. As presented above, stricter criteria were implemented in 1960 than 1950 which resulted in a national decrease in the farm population of about 21 percent from that population expected under the accepted definition of 1950. Some so- called "urban townships" were designated by the Census in Pennsylvania, but apparently the effect on population classification was small (U.S. Census of Population, 1960, Vol. 1, Part 1, p. xix). The 1970 farm population data. The 1970 farm population as a proportion of the 1960 farm population was obtained from the 1969-70 change data given in the County and City Data Book of 1972. In both the 1960 and 1970 population censuses the farm population data were based upon samples, but these were large; consequently, the contribu- tion of the sample error to the aggregate error in the data was minor at the national level. Nevertheless, "for certain counties and socioeconomic subgroups, the sampling error is fairly appreciable ." (Shryock et al., 1973, p. 174). 107 Overview of the Areal Patterns of Farm Depopulation The mapping of the representative measures of the relative and absolute magnitudes of a problem is commonly the initial step in geo- graphical analysis of a problem. Once the data are cartographically displayed, gradations in the values of the measured geographical units, e.g., the counties, often reveal contiguous areal units of similar quantities and with much the same geographical characteristics, suggestive of some hypotheses. Retention of the Farm Population Cartographic presentation of the proportions of the farm popu- lation remaining on the farms for the counties of the Northern Appalachians in the periods of 1910-30, 1940-50, 1950-60, and 1960-70 produced meaningful areal patterns of these pOpulation changes (Figures 7, 8, 9, and 10). For comparisons of the more recent time with the situation earlier, it seemed preferable to combine the teen and twenty decades. Because of the emphasis in this study on the declines in the population residing on farms--and as this was the situation in nearly all the Region's counties, for the first three-quarters of the 20th century--the unusual 19305 decade which witnessed a number of counties with increasing numbers of farm persons was not mapped. Some general findings were revealed for all of the periods. Much variation existed in the county depopulation rates of farms in each of the several periods; however, the largest variation and the highest average losses occurred during the 19505 with the 1960s rates somewhat lower. Often contiguous counties with like values formed several areas that differed very noticeably in the retention 108 on oama .mcawgom~aam< :uozuuoz on» :w xuczou >9 :owueaaaom spam as» .oan mo :ofiucouoz ommucoouom--.n .wwm or. :2: a... D g mew 2 on E 8.2. 2 3 a a...» 2 2. E .26 can no a Q o ... ‘ . . ......u............ Ho: Q0 .... .... . ... one o n C O C o. o. to o a. I on I. o e. o o o o O O C 00 0090.0 0.0000 C. O. . CO. 0 I. C 00. 0- 0 ’0 about... I O I. on o o e o o o o o o a no no. ' no 0 o o o o no a O o o . O.- a. o a. o. to C . o. o I... o no... 0 no. . . 00 I. In a o o o no... . o o o! o 0.0 e on o 0 IO. 0 o I no. a o o no.0 0- u on a o I on I... .- o O o O 0.00 Pop). I. on 0. co. 109 .omma on cvma .mcafinowammm< anocuuoz any cw Aucaou x9 cowuaasmod spam onu mo :owpcouom omwucooaod--.m .mflm x x Q no :2: a... min 2 or. we» 2 mm aéo 2 as .26 use no @IID .. .33 . .. ........ .... z \\nwn u” r .......... O c o a new. ooooooo v ”I O I \vu-OIOO A0. Ollolonoo I1. I O o ooooo no .- u-o-ooo.o ii. .000. O .. on o no nnnnnn O o .... . 2...... .. ...-...... a mu . ”scams n: .. n . ..... loo-00.0“ ”I ‘ a .......... 00.. O A. \Ib ..... no. 00 ‘- . . . O .... To 0 Slat Q . 1 .. -.. \L a. .. . . \ I 1‘ o n IIIIIIIII ”.2... h n .. \ .oomH ou omma .mcmfinomHamn< :uonuuoz ecu :« xucaou xn :owumaamom swam may we :ofiucouom ommucouuomuu.m .wwm 110 3”). ) a. W. a; <‘-'; )- 3 =2: :2 new 2 on 3:. 2 no aéo 2 2. ...)O Uco DO EIIID 111 .onmm ou coma .mcmficomfimmm< cuonuuoz can ca xucaou xn :ofiumasmom spam ego mo cowacouom omeucoouom--.oH .mE on :2: 32 ado 2 on méb 0. no mic 0. on =3 23 no "D E 112 of farm persons. Also, areas usually differed as to when the largest farm population occurred. Subregions of Prolonged Farm Depopulation Except for the counties bordering Ohio, Western Pennsylvania has had the most persistent and highest losses. Another area of steadily high rates of farm depopulation coincided with the Pocono highlands in northeastern Pennsylvania. Because of the recurrent low rates of farm population retention in much of the area westward from the Allegheny Front of Pennsylvania, i.e., in the Allegheny Mountains and on the Appalachian Plateau, and in most of the subregion of the Pocono Plateau (Figure 7); the negative effects of adverse t0po- graphical and soil conditions on agricultural operations, and thus, the maintenance of a farm population engaged in agriculture was presumed to be the most fundamental cause of the high loss rates in the agricul- tural population. These subregions particularly lack the physical resources, such as the level and well-drained lands, climate, and soils much favored by commercial farming and the new technological changes in agriculture. As early as the beginning of this century, new agricul- tural techniques had the decided and increasing effect of outmoding a large portion of the sloping, wet, and less fertile lands (Baker, 1921). The costly new agricultural systems partly adopted in the above areas of the northern Appalachians were generally found unfitted to the physical resources of these areas and the traditional farming systems that had gradually evolved to meet the natural constraints of such 113 areas could not compete commercially with the agricultural systems of those areas with much more fertile soils and milder climates. Western Pennsylvania. Largest rates of farm population decreases which were mostly in western Pennsylvania tended to be associated with rural counties. In the last three decades only a few moderately urban counties, specifically McKean, Venango, Jefferson, Beaver, and Cambria, with urban populations of 40 to about 70 percent had continued high loss rates; However, the urban influence in the first three counties was probably less than the rural as their rural populations exceeded their urban. As to highly urbanized counties, only the county dominated by Pittsburgh, Allegheny County, was among the counties with the highest rates of farm population decreases. The strong rurality of counties with high farm population losses for the several last decades indicated the lack of adequate alternative and sufficient working opportunities outside of agriculture prompted some persons to leave farming, particularly given the technologically out- moded physical resources of western Pennsylvania. Resource extraction areas. Some secondary economic activities, using the physical resources obtained, dominated western Pennsylvania's economy. Demand for coal, really the base of the economy in many areas, fell off precipitously after World War II. Coal's chief competitor, oil, increasingly powered locomotives, heated homes, and produced "thermal" electricity. A more efficient process for making coke used in steel manufacturing reduced the amount of required bituminous coal. Steel production declined and the new technology 114 introduced into the industry reduced the number of steel jobs (Stevens, 1964, pp. 348-351). The railroads were heavily dependent on the resource extraction industries, especially coal, and thus suffered attrition which eventually ended in the demise of many rail routes and companies. Railroad mergers were counterproductive for the welfare of many communities as many rail lines, repair, and maintenance facilities were abandoned. Before the I.C.C., Shapp (1963) in sworn testimony with a historical perspective warned of the damage that would be inflicted on many communities and on the greatly reduced chances of industrial redevelopment if the two giants of railroading in the Northeast, the Pennsylvania Railroad and the New York Central, were allowed to join and create yet another monopoly to stifle "active competition” in the Commonwealth. The large rural populations including the farm population of Pennsylvania encountered serious unemployment, low and declining income, shrunken local markets-~all made more prevalent and encompassing because of the undiversified economies. Depressed areas. The decline of the mining industry and the advancement of mining technology displaced thousands of miners and generated a long-lasting stagnation on to the State's coal regions. As more than half of the counties of Pennsylvania mined coal, the downfall of this primary economic activity was a basic cause of much of the distress experienced by many Pennsylvania localities during the mid-Twentieth Century. In 1955 four of the nine large areas with "very substantial labor surplus" (with more than 12% of the working force unemployed) were in Pennsylvania (Stevens, 1964, pp. 348-350). 115 Between 1940 and 1960 the rate of income growth for Pennsylvania placed it forty-fourth among all the states. Income problems for farm people had obviously come before those of other occupational groups, for even in 1929, the Pennsylvania farmers acquired barely more than half the average annual income of that of the urban people and average annual farm income was slightly more than one-fourth that of California (Stevens, 1964, pp. 345 and 359). Nonetheless, the economic situation for farmers worsened with the Depression and the fundamental declines in the pervading economic activities of coal mining, iron and steel manufacturing, and railroading, and also the effects of the earlier forestry work slowdowns reduced the local markets and demand for agricultural products. "The extent to which agriculture overlapped with mining and manufacturing as a part of the economy of many a Pennsylvania county is noticeable" (Stevens, 1964, p. 345). Compounding the recovery of economic health of many of the depressed areas was the widespread devastation inflicted upon the natural environment through the increased use of strip mining and the deposition of waste materials of mining in giant piles or culm banks on the surface. The despoiled natural environment was said to have been a major obstacle in the attraction of new industry to the coal areas (Stevens, 1964, p. 350). Although some recognition has been given to the negative effects of the monetary distractions of coal mining and oil and gas production on farmers' desires to farm (Miller and Parkins, 1928, pp. 114-115; Murphy, 1937, pp. 470-471) apparently little study has been done on the destruction by strip mining of the 116 land resource base of agriculture in Pennsylvania (Miller, 1949, pp. 94-97). The direct and indirect influences of the nonagricultural but rural economic activities have undeniably had an important impact but statistically inseparable effect upon the agricultural systems, and thus also, the farm population. Subregions of Least Farm Depopulation Not all counties with high proportions of rural populations have had high rates of farm population losses. Central Pennsylvania. A contiguous block of Central Penn- sylvanian counties, with no coal mining and with rather high percent- ages of rural populations, have retained agricultural persons on the farms to a greater extent than any other areas of the northern Appalachians. These counties, e.g., Union, Snyder, Mifflin, Juniata, and Perry, located in the heart of the ridge and valley physiographic region of Pennsylvania, have had by necessity, because of the extremely steep slopes of the ridges, nearly all farm land concentrated in the valleys, enriched in places by the underlying limestone. As the modern agricultural systems have had special affinity for level and fertile lands, many farm persons were able to stay on their lands with some incorporation of the new ways of farming. As fewer farms were sited in areas with rough topography than any other part of the northern Appalachians, because of the extreme nature of the relief, farmers of the subregion have seldom given up their farms because topographical conditions did not fit the mechanization of new agricul- tural systems. From the end of World War 11 until about 1960 this 117 part of Pennsylvania along with some of southeastern Pennsylvania had the lowest rates of agricultural land abandonment and some townships had increases in farm land (Slocum, 1969, pp. 45-46). Some of the reasons for these low rates of abandonment and actual increases in agricultural acreages in some localities can be traced to the presence of the Amish and the Mennonites, the more conservative religious sub- groups of whom especially uphold farmlife as the most ideal way to live which coincidentally permits better social control of the sects' members. The Amish are conspicuously present in eastern Centre County and within the Kishacoquillis Valley of central Mifflin County; they came to the latter in 1791 from southeastern Pennsylvania (Klein et al., 1973, pp. 440-441). Both the Amish and Mennonites can be credited with the restoration of good farm land from the state of abandonment in various parts of the Commonwealth; a geographical study of this most beneficial and private development would provide very valuable basic information and lessons for rural development programs. Urban areas. Perhaps the most important areal pattern revealed by the cartographical comparison of several decades of farm depopulation were the relatively low rates of farm population reductions in the most urban counties, except for Allegheny County. The most urbanized counties--Broome, Lackawanna, Luzerne, Erie, Blair, Northumberland, Schuylkill, and Westmoreland--have ng£_had continually high or even moderate rates of farm population decline as have many of the rural counties located in the coal mining areas. Some of the counties however had one decade of major losses since 1940. The fact that the majority of the most urban counties have had some of the 118 lowest rates of farm depopulation indicated the direct urban effects, i.e., those of urban encroachment upon agricultural lands, have not been among the most important reasons for persons leaving the farms and that urban forces alone have not had the magnetic qualities for farm people as so often presumed. Some study of the types of farms in these urban counties possibly would disclose a much higher percentage of estate and semi-urban farms than in other counties somewhat analogous to the wealthy farm holders in the Philadelphia suburbs and the renowned horse farms adjacent to Lexington, Kentucky, both types given credit for much of the preservation of open space in the urban environs. As Pennsylvania has had an exceptionally low population growth rate and a modest increase in the urban pOpulace with little change in the percentage rural and urban composition in the last decades, the urban effects have undoubtedly been less than in rapidly growing regions of the United States. Recommended study areas. A few counties although losing farm population since 1940 achieved the best records for keeping persons in agriculture and should be the study areas of future field investi- gations to obtain answers as to what can aid farm families to continue farming. The counties meriting an observational examination are Chautauqua, Cortland, Union, Juniata, Perry, and probably Tioga (in Pennsylvania), Centre, and Somerset, as the farm population retention rates were also high for the latter three in the 19505 and the 19605. 119 An Overview of Farm Depopulation The 1910 to 1930 Period During 1910 to 1930, the southern third of the northern Appalachians had very little farm depopulation; whereas, to the north across a narrow transitional belt of counties, stretching nearly all the way east and west in Pennsylvania, were counties of heavy losses except for a few scattered counties primarily in New York (Figure 7). The farm population changes in this period produced areal patterns that suggested considerable spatial correlations. Some speculative thoughts as to the basic causes of these patterns follow. The southern subarea. Southwestern Pennsylvania is a part of the Allegheny Plateau that composes much of the northern and western portions of West Virginia and as such, resembles the central Appalachians more than any other part of the northern Appalachians. This area of Pennsylvania has had significant losses recently in population as have the central Appalachians; and likewise, its economy has long been based on the vast and rich soft coal deposits. However, during 1910 to 1930 the coal industry in this subarea was still economically healthy although there were some indications of decline. World War I greatly increased the demands on the railroads; the steel and related heavy industries concentrated in this part of Pennsylvania. Such an economy generated demand for local agricultural production. Within the Allegheny Plateau of the Commonwealth small farms predominated in the more rugged southern locations and larger farms appeared in the more northern areas (Pillsbury, 1971, pp. 4-5). At this time, 120 probably the subsistence characteristics of many of the small farms permitted easier adjustments to the rapidly fluctuating economy than did the larger farms located mostly outside of the main soft coal production areas of southwestern Pennsylvania. The larger farms with more capital invested had much more to lose from the decreasing prices of farm products. The northern subarea. In the northern portion of the northern Appalachians, the high rates of depOpulation on the farms from 1910 to 1930 may have reflected the effects of the broad acceptance of a new system of agriculture, commercial dairying. The change to dairying often required farm consolidation which displaced subsistence farm families. Some of the agricultural experiences of New England (Wilson, 1936) probably affected the northern part of the Region earlier and more intensively than the southern because this area had both physical and cultural characteristics akin to those of New England. The 1940 to 1950 Period In the 19405 the counties with the highest levels of farm population stability had peripheral locations in the region partially reflecting the availability of defense jobs and the presence of dairying, both of which allowed young men under certain rules to be deferred and remain on the farm (Figure 8). The farm persons of the more central parts of the Region inflicted with inadequate transpor- tation, physical and social isolation, much agriculturally marginal land, and limited working opportunities had few alternatives except to 121 seek work elsewhere, to find a part-time job if available, and/or to continue subsistence farming. In northeastern Pennsylvania the side effects of the decreasing demand for anthracite coal were first experienced in the counties surrounding the northern field in Lackawanna County. Many farmers of the northern Appalachians were still very much dependent on the demands of the local populations. The 1950 to 1960 Period The areal patterns of farm population change in the 19505 resembled somewhat those of the previous decade. A large cluster of counties with high farm depopulation rates was revealed in western Pennsylvania (Figure 9). The location of the counties with good farm population maintenance rates in southern New York, most of northeastern Pennsylvania, and central Pennsylvania connoted an association with the importance of dairying; whereas, in most of western Pennsylvania where widespread depopulation of the farms prevailed, dairying was of much less significance and general and mixed farming dominated. The less commercialized nature of agriculture in much of western Penn- sylvania, partly because of the less availability of favorable physical resources, motivated many farm people to seek part-time off the farm employment that eventually led many out of farming. The continued decline of anthracite coal mining in the northern and central fields may have adversely affected the farm population of Luzerne, and Carbon counties. 122 The 1960 to 1970 Period In the 19605 less variation in the farm population change rates characterized the Region than in the 19505; thus, the possible meanings of the county areal patterns of farm depopulation were less evident (Figure 10). Farm depopulation was more general in geographical occurrence, but there were a few contiguous counties with similar con- ditions and rates of farm population losses. The counties with the best records of holding farm families on the agricultural land in this decade tended also to have held on to their farm populations better than most counties in the other decades since 1940. These counties' characteristics favorable to agriculture were discussed earlier. Several blocks of counties known for particular economic pursuits had relatively high farm population losses, prompting queries as to the effects of these operations on the conduct and nature of agriculture. The oil and coal areas of western Pennsylvania. The counties known for their oil production--McKean, Venango, Warren, Elk, Forest and to a lesser degree a few adjacent counties had proportionally high farm population declines. The area immediately south of these major oil counties, in west-central Pennsylvania, and particularly in Clearfield, Jefferson, Butler, Clarion and Armstrong, had only somewhat lower loss rates and were the counties of major coal-strip mining operations. Both the extraction of oil and coal have led numerous times to local "boom and bust" economies, leaving behind ghost towns and abandoned farms. 123 The anthracite miningareas of eastern Pennsylvania. The additional areas of high relative decreases in farm persons were in extreme eastern Pennsylvania and the eastern fringe of New York. From 1940 to 1970 high farm depopulation rates occurred in the hard coal counties, first in the northern ones, then the central, and lastly the southern. The southern hard coal counties with heavy loss rates of farm persons in the 19605 decade were Carbon, Northumberland, and Schuylkill. The differential areal and temporal declines in anthracite production apparently effected the farm pOpulation loss rates accordingly. The recreational areas. Lastly, the counties in the recre- ational areas of the Pocono and Catskill highlands had relatively large reductions in numbers of persons living on farms, partly caused by residents of megalopolis who purchased a second home and some land in an altitude presenting cool temperatures where they could escape from the oppressive heat and air pollution of the big cities in the summer. The establishment of "residential farms" changed the pre- existing types of farming, e.g., by increasing the renting of land, and brought changes to agricultural systems in the Region. Methodology [gelationship_of the Problem to .332 Method of Analysis In research, about the first task is to choose a topic and/or problem. The outcome of this first step in research depends upon the researcher's experiences, interests, and education; all of these 124 played an important part in the selection of the farm depopulation problem for this dissertation, but perhaps, the first influence was basically the most persuasive. The author knew through years of living and working on farms in Susquehanna County, from seven years of study- ing and residing in Monroe and Centre counties, and from traveling in northeastern, central, and west-central Pennsylvania and southeastern New York that the degradation of agriculture and the depopulation of the farms constituted two general and serious problems in most rural communities of the northern Appalachians. Thus, there arose the question: how much are the two phenomena related? The loss and deterioration of farms and the agricultural and human resources really constitutes one problem in numerous localities but is nevertheless a multifaceted problem. Earlier the author studied one major aspect of the total problem: the abandonment of Pennsylvania agricultural land in the post World War 11 period (Slocum, 1969). The retrenchment of agriculture and the general agricultural changes in the northern Appalachians, have had numerous negative effects, including changes in land use; but, the consequences for farm persons have received generally insufficient attention in public sponsored research. An agricultural administrator of the United States' Economic Research Service in mid-1973 said, Much of our past research has been based on an objective of efficiency in the land-labor-capital context. This has little relevancy to this group of small farmers and rural residents (West, 1973, p. 2). The farm resident deserves much more emphasis than that provided by the view that he is just a resource or input similar to the other resources necessary for agricultural production because he, in himself, 125 combines many resources, including management abilities, in addition to the fact that he is a human, a quality around which many of our values coalesce. Thus, to think of the farm person only in respect to his potential contribution to agriculture, without regard to other aspects of his life and the effects which agricultural changes may have on his welfare is an approach of analysis that produces results of limited applicability to his welfare. To avoid the treatment of persons engaged in agricultural pursuits only as agricultural pro- duction variables, it is desirable to regard them as constituting a separate system, i.e., a farm population system interacting with either an agricultural system or subsystems; then, the effects of one upon the other may be seen from a wider perspective. A Systems Perspective A broad perspective is often preferable in research because of the frequency of problems having multiple causation. Causes are often composed of sub-causes originating from various systems. Solutions with effective results have been achieved infrequently in the past because of an overly-specialized approach to a problem--the analysis of the workings of only one system or a subsystem. The repercussions often only appear in observational form in the long run, and thus go unrecognized, increasing the intensity of problems. For example, the errors of not regarding the effects of man-made systems on natural systems finally became obvious as environmental problems became serious. The human system is as fully complex as the natural; we cannot pretend to know of all the influences and their consequences 126 upon the human being, but we need to know as a minimum, that systems often interact with each other and elements of one system interact in varying degrees with those of other systems. An example of a number of interacting systems pertinent to the general theme and approach of this study could be given as follows. The acceptance of at least some parts of agricultural systems, e.g., tractors with auxiliary power machinery, generally made life easier for farm families. In the northern Appalachians, however, much of the land was unsuited to the new agricultural technology. Consequently, the potential production and competitive position of many farms frequently became less, lowering many a family's economic standing and affecting the local economic system. Yet, many agricultural innovations, like the automobile, were considered essentials of modern living and of the new social systems. Another example of the interrelationships among systems rele- vant to the study area supplies support to the use of the systems approach. The farmers held a high proportion of land which gave only minimal returns. Agricultural changes caused more land to be unsuit- able for modern farming. Nevertheless, taxes had to be paid on this outmoded land by farmers and as the costs of upgraded school systems and of other public services expanded, the farmer paid higher taxes on more lands that could no longer produce given the new agricultural technology. The major cost that many individuals had to pay for the privilege of living in the countryside and/or on a farm was the acceptance of a lower standard of living, a consequence produced by the interaction of several systems. 127 The System Concept Thus, the general approach chosen for this research is most closely related to systems analysis, more in a conceptual way than an operational or applied sense. The meaning or concept of "system" is a mental construct and gives orientation to this study. The idea of systems is attractive to researchers because of the large fundamental accumulation of knowledge about the major classes of systems, i.e., open and closed, and on the similarities in structure, behavior, and states. "A system may be simply defined as a whole composed of parts which interact. The parts are called elements of the system" (Carey, 1970, p. 179). The element is the basic unit, but at certain reso- lution levels of analysis it may be regarded as a system. Two broad classes of systems are the closed and open systems. The open system is the most representative of the real world and unlike the closed system will permit matter to pass its boundary as well as energy (Carey, 1970, p. 180). Through time (history) Boulding (1970) sees systems operating in one of four ways: randomly, mechanically, teleologically, or ecologically (selectively). Although there exists a large collection of literature on systems analysis, the discussion is restricted primarily to works in geography to correspond to the perspective of the discipline in which this study is done. Systems theory in geography. Harvey (1969), among the geog- raphers, perhaps best analyzed the concept and "subconcepts" of systems analysis and concluded ". . . that methodologically the con- cept of system is absolutely vital to the development of a satisfactory 128 explanation" (Harvey, 1969, p. 479). Ackerman (1958) in his review and assessment of geography as a basic research discipline and in his emphasis on process was among the first geographers to advocate the use of systems thinking in geographical research. James (1972) wrote, each kind of process is also modified by the presence of other things and events of unlike origin that exist together in mutual interaction in earth space. The interconnections among things and events of unlike origin on the earth form systems of functionally related parts (James, 1972, p. 459). A system that contains at least one important spatial variable, e.g., location, distance, direction, extent, density, and succession is considered a spatial system (Wilbanks and Symanski, 1968). Writings on systems related to geography have become common in the last fifteen years (Chorley, 1962; Foote and Greer-Wootten, 1968; Langton, 1972; McDaniel and Hurst, 1968; Walmsley, 1972). Agricultural Systems Much has been written on agricultural systems, but there exists no real universal agreement on the meaning of the concept. The International Geographical Union for several years has had a sub- committee addressing the problem under the direction of Kostrowicki who has written extensively on the topic. Some notable research on agricultural systems has been carried out in the past (Grigg, 1974; Duckham and Masefield; Harris, 1969; Helburn, 1957; Olmstead, 1970; Spencer and Stewart, 1973). The General Model Consideration of all the systems influencing and interacting with the farm population would present a study beyond the necessary 129 scope of this work. The point of reference thus selected for this study is that changes in agricultural systems due to the forces (the mechanisms) from the ever-enlarging and encompassing technological system or environment are basic to an explanation of the overall state of the farm population system at any given time. The ultimate effects of these change forces on the interrelationships of agricul- tural and population systems and the subsequent displacement of farm population within an area varies however according to the site and situational characteristics of a place and an area (Figure 11). The technological derived forces shaped by these geographical factors result in differing removal rates of persons off the farms. The primary resources of agriculture were envisioned in bundle-like groups, termed components of agriculture, e.g., land, capital, labor, mechanization, and off the farm inputs. The changing nature of agri- culture was thought to be a reflection of the increasing inputs on to the farms from the technological system and environment. General Procedure of the Analysis Once the dependent variable was selected to represent the problem, several decisions had to be made on what variables would represent the tentative explanations and would be tested. The chosen variables (Appendix A) generally fitted the subsystem categories of an agricultural system as shown in Figure 11. These variables as attributes of the subsystems or elements of the agricultural systems were hypothesized as having been likely influenced by the spread of technology to the farms. Some of the variables, e.g., tractors, 130 .maoumxm 3.5332: 98 3.332 smack—3 Eoumxm camouflaged a 5“: #832335 Rouge—Each. me Home: 033--.: .m: Emkm>m 4o g 148 .3 \ 3 E \ 2 2° 4: O \ a \ 3 l5 3 O \\ .C s- \ ° IO “A g s 9 \\ S 5 *f‘ ~D~--- . 0 l I9lO l920 1930 1940 1950 l960 l970 II-- Horses and mules — TrOCIOfI Fig. l4.--Replacement of Horses with Tractors. Millions of tractors 149 working 100 days off the farm; (2) the percent of total farms, dairy; and (3) labor productivity or agricultural sales per agricultural person. The part-time farmer was defined as working at least 100 days per year in off-farm employment (Rogers and Burdge, 1972, p. 138). The numbers of part-time farms expanded from 15 percent of all U.S. farms in 1929 to 31 percent in 1949; thus, nationally these types of farms represented nearly one-third of the total farms in the post World War 11 period and their relative numbers continued to increase in the 19505. During this period rural sociological research found that this kind of farming did not represent a means into or out of agriculture for most persons engaged in it, but surprisingly a "permanent status" (Rogers, 1960, p. 5). The negative correlation, r = -.74 between the measure of part-time farming and the proportion of farms that were dairy revealed that in 1949-1950 the two kinds of farming were for the most part mutually exclusive. Because of the yearlong demands on the dairy farmer's time and the necessity of his maintenance of a two or three times-a-day milking schedule, the dairyman is very much restricted to farming a seven day work week--and there is little time or energy remaining for significant off the farm work. Dairymen hesitate to hire labor to care for valuable milkers each of which demands special individual attention and careful feeding so to remain a maximizing producer and a good investment. By using their own and family labor, dairy farmers experienced increases in their labor productivity. In the study area counties with a high percent of dairy farms had high levels of agricultural labor productivity. Unlike in dairying, a 150 part-time farming enterprise can only have a part of the Operator's time; thus, when the farmer works elsewhere, his farm labor productivity declines according to the extent of time he spends off the farm. The average value of implements and machinery per farm acre as a measure of farm mechanization had increased to the point in the late 19405 and early 19505 that it had a large positive effect on increasing production per unit of land. By 1950 the northern Appala- chians had achieved an r = .72. Mechanization permitted farming operations to begin and end on time. Machines gave the farmer more time for management, planning, and additional productive work. The increased productivity of the land made possible by the expanded mechanization rates, which has been called "the most important change in farm technology of the Twentieth Century," also had an important indirect influence on releasing the forage and grains consumed by the displaced horses to the production of marketable agricultural products (Wilcox, 1947, pp. 289-290). The major contribution of machinery came particularly with tractors, after World War II in the late 405 and early 505 in the northern Appalachians when thousands of horses were replaced by tractors. Horse—drawn equipment eventually was traded in on tractor-drawn or powered auxiliary machines. Basic agricultural characteristics in 1969-1970. What were the major differences in the basic appearance of the Region's agricul- ture in the late 19605 and the early 19705 from that of the previous time (Table 11)? The disbursements for gasoline became the variable with the highest intercorrelations with other farm attributes. The intercorrelation of the average dollars spent for gas per farm with 151 Table 11.--Highest Linear Correlations Among Agricultural Variables, 1969-1970. Variable ID Variable Name r Number 4:12 Avg. 3 spent for gas per farm .81 Avg. cropland acreage per farm reporting 4:8 Avg. $ spent for gas per farm .79 Avg. expenditures for fertilizer per farm reporting 14:15 % harvested cropland, hay .77 Pasture-cropland ratio 21:16 % farm operators working 100 days off the farm -.76 % of total farms that are dairy 6:9 Avg. value of implements and machinery per farm acre .74 Farm persons per 100 acres 6:11 Avg. value of implements and machinery per farm acre -.71 Avg. size farm in acres 11:16 Avg. size farm in acres .71 3 of total farms that are dairy Source: Matrix of Correlation Coefficients, Appendix F. 152 the average crOpland per farm rose from r = .78 to r = .81, the highest association among the agricultural variables in 1969-70. In contemporary times there is thus within the study area a relatively high positive relationship between how much cropland a farm possesses and the quantity of gasoline needed. The second highest association also involves gasoline expendi- tures, related to the average expenditures for fertilizer per farm reporting with an r = .79. As fertilizer is used relatively more on the crOpland today than in the past, i.e., r = .55 in 1949-50 and r = .66 in 1969-70 and gasoline expenses are closely related to the amount of cropland per farm, the emphasis on the intensity of cropland use partly explains the important interrelationship between the two pur- chased inputs, of gasoline and fertilizer. Once again the use of an off the farm input seemingly establishes a need for the use of addi- tionally purchased inputs. An increase in use of fertilizer, as long as the point of diminishing returns is not reached, increases production to the point that fixed costs become accordingly decreased per unit produced and large amounts of fuel costs can be more easily rational- ized. Tillage of larger cropland acreages absolutely requires greater gasoline use, and to raise the probability that this major expense will be covered, the use of fertilizer significantly increases the probability of a money-making crop. In 1969-70 the negative association, r = -.76, between part- time farmers and dairy farms remains as important as in 1949—50. Many of the counties with low rates of farm operators working off the farm 100 days or more are counties with a high proportion of dairy farms. 153 The association, r = .71, between the proportion of the total farms that are dairy and average size of farms is exactly the same as in 1949-50. As the average size of farms increases the proportion of farms that are dairy rises. UMWERS ANALYSIS OF THE COMPONENTS OF AGRICULTURE: SOME DOMINANT CHARACTERISTICS AND SYSTEMS OF FARMING IN THE NORTHERN APPALACHIANS, 1909-10 and 1929-30 Identification of the Region's Agricultural Components The most important characteristics of the northern Appalachian agriculture were thought to be different from one time period to another, and this hypothesis was substantiated for the most part after the above analysis of the major agricultural variable intercorrelations in each era. However, there was little attempt to relate these attributes to subareas in order to identify the geographical extent of agricultural systems or types of farming. Previously, it was hypothesized some of the areal variation in farm population changes could be attributable to certain types of farming, and changes that occurred within these. As data for the relative number of farms of each type, except for dairy, were not included in the selected data some indication of the dominance of certain agricultural systems in each county is necessary to the eventual formation of theoretical statements explaining the farm population losses. Thus, a major objective of this part of the study is to associate the essence of 154 155 the agricultural data with a geographical unit which in this research is the county. It is preferable to establish the geographical patterns of the most important historical components of agriculture found in the data analysis of this study. If some of the agricultural population losses correspond with one particular component found in all farming systems, e.g., investment, than a greater study of this economic characteristic of agriculture would be warranted first to discover exactly how the component results in a decrease in the farm population and thus effects the human resources. Using the intercorrelation matrices of the data, principal component analyses were carried out for the several decades since 1910. This type of analysis gives the correlations of the variables with the components. The components are the new independent variables and the main underlying dimensions of the data. Through Observing the general nature of the listing of variables that correlate highly with each component (the correlations are called the loadings), it is possible to identify the components. Of special significance to a geographical approach are the component scores of a principal component's analysis which relate the composite variable formed from the intercorrelations of the original data matrix to each geographical unit or observation. In this study, therefore, each county has a value for each of the com- ponents chosen by the principal component's program. When these new composite values, i.e., component scores, are mapped a double-check on the interpretation of the components deduced from an analysis of the loadings is made possible by whether the tentatively identified 156 component could logically have the geographical distribution that is revealed by its mapped component scores. The Components of Agricultural Systems in 1909-1910 The components selected by the principal axis analysis accounted for nearly 87 percent of the total variation in the agricul- tural data. The principal component in 1909-1910. The major component extracted explained more than 30 percent of the variance and corre- lated primarily with implements, machinery, land, and building values (Table 12). Generally speaking the spatial pattern of the high posi- tive principal component scores, representing collectively the associ- ated variables, support an intensive agricultural interpretation (Figure 15). The counties of Erie, Lackawanna, Luzerne, Allegheny, Cambria, Schuylkill, and Westmoreland had high land prices because of the urban use demands near such cities as Erie, Scranton, Wilkes-Barre, Pittsburgh, Johnstown, and the other industrial and coal oriented cities. Also, mechanization probably began near the cities. Addi- tional positive values occurred in counties on the eastern and southern portions of the Region, an area adjacent to the progressive Pennsyl- vania Dutch agricultural domain of southeastern Pennsylvania. The lowest negative values, meaning low relative investment, tended to be those counties both isolated and the poorest in agricultural develop- ment, e.g., Fulton, Bedford, Huntingdon, Potter, Perry Sullivan, Pike, Juniata, Cameron, Forest, and Green. Most of these counties are among Pennsylvania's most mountainous. Thus, quite unexpectedly, capital 157 Table 12.--Major Loadings from the Principal Components Analysis of Agriculture, 1909-1910. Positive Loadings Negative Loadings Variable Loading Variable Loading Component One: Intensive Agriculture Value of hired labor .80 Avg. value of implements and machinery .79 Land productivity .72 Avg. value of land and buildings .72 Avg. expenditure for feed .63 Avg. expenditure for fertilizer .58 Explained Variance 30.06% Component Two: Cattle and/or Dairy Farming Labor productivity .81 Farm persons/100 acres % harvested cropland, hay .76 Avg. size farm in acres .66 Avg. expenditure for feed .66 Corn yields .53 Explained Variance 28.06% Component Three: Cropland Availability Avg. cropland harvested .81 Ratio farm expenses to Land productivity .48 value of agri. products Explained Variance 16.60% Component Four: Diversified Farming and Specialized Crops and Livestock Avg. expenditure for Corn yields fertilizer .53 Cropland harvested .52 Explained Variance 12.10% -.76 -.65 158 .OAmaumoma .ousumsofium< O>Hmcounm .mouoom Ono unocomaou mo :Owuspwuumwa ~aonfimcoucu .mououm 0:0 «cocomaou mo cowuanwnumwn Huonskmccoa mean.mavv www.mmv Hmo.vov mae.mm3 mmfi.aov somumn zoz Romm.vma.ov mm~.wse.v 025.2oo.¢ aka.mvw.m mam.m~o.e xao> zoz Ammm.mvo.afiv wo~.omm.m mmm.oma.o voo.~oo.o now.amm.o ofiocmfiu< ofleeflz ”Nam.mmAU Nem.ome mam.24m mmo.oom Acm.~om osofipumccou ammm.omv mam.mv ooh.ov o~2.Nm omm.~¢ camHmH mecca amam.mmev mov.oam mmm.mv~ ~ma.oo~ Nae.w2~ moommscummmmz “mo~.vmmv om~.ofim ~wm.mmv Nan.~ov wwo.m2m ucosno> Ammv.vv~v cam.m~2 mo~.mo~ oom.2m Nmm.am coagmasa: zoz fim-.~2mv www.mmH o-.mm nom.an mmm.2m mama: nemm.vmn.mv HHN.aom.H «Ho.mmN.~ cow.mcH.H www.mHN.H ecmfimcm zmz am:0uv owmaww How you :uou «wa.v~m.m ~mm.mom.m oom.~mo.e omo.2wv.n oo~.~vo.m aficm>~sm==oa com.wa~ omo.52m mom.omv wan.a~v mav.omm Ammumn :02 o-.wvo.v nan.a~o.m wma.omm.m mma.~mv.m 0mm.~o¢.¢ x00» 302 ooq.~m~.m mmm.mqw.m new.mom.m mmfi.oov.m 2mm.mmw.a odocmsu< mane“: Han.ows av~.~mm mom.o~m emv.~am Nwo.mem pauflaumccou oom.a~ omm.v~ o~m.am 3-.mm 430.4m ecmsma oeogm www.mNN moa.vm~ Amm.mom ~wm.wov NNm.wmm mouomszummmmz mma.onw "ma.mmm ov~.~vo.H moq.ao2.~ mmo.kam ucosuo> www.202 mam.mm2 m4~.mm~ o~e.om~ omm.~om onazmesm: zoz ~oa.mmm o-.amv 2-.mom mN~.Hoo mom.m~o mafia: emo.mma.2 om~.202.~ vom.mwm.~ mzo.m~w.w ~mv.~mo.~ camamcm zoz Am:Ouv cofluosvowm xm: ~H< moms coma mmms «mmfi mems .aomaumvmfi .ummozuuoz on» :w mvcouh :Owuospoum :Hmuu paw owmfifim show can xmr--.wa anah 207 .ucwfloz :oowm mucomoumon momfi cw numb owmfiwm new use :Houm .momH OH mvma .uomnnsm kn mowumwumHm .muuomom HNHOCOU .QHDHHDUHHw< mo momsmcmu m30mhm> "moohsom 033.004.25 oao.omq.om moo.vmm.mm cam.¢~m.mv ova.mom.mv mfl=m>sxmccma How.woo.m aoo.ov~.m mmw.am~.c ooq.aov.o mmm.~mm.¢ somumn :oz omo.vfiv.- Hm2.mmm.22 mmm.mm~.- amm.~o~.~s oa~.moa.a x00» :02 amm.mmm.am 42~.Amm.mo omv.mo~.mh mHa.vmm.oo Hov.mom.ao aflocmfip< oHeefiz nom.fim nem.om a~m.om~ amo.om2 mHo.HmN asufiouoccou no~.m~ owo.sm wao.oH Amo.m2 mom.mm namfima oeonm aqm.moq oao.qm oso.oe2 som.wm2 HAN.QH~ modemszummmaz o-.~m~ Amm.mm ~ma.~m eom.~m wmo.~o~ acoe~o> 3mm.aq amm.va ave.m2 o22.m~ cm~.oc ouagmaea: zmz nmm.~o~ moN.HH wwm.m om~.AH oHA.mm mafia: va.moo ode.mw~ www.mmm mma.mmv mmo.moa ecafimcm 302 mEhmu— HH< co Amaofimsnv CMMHU Mom fiQumo>hmI CHOU mom“ woos smog «was megs .eoscmucoo--.w2 manmh 208 new inputs, including mechanization, were often abandoned. With the presence of inexpensive energy the large increases in yields per acre, e.g., for corn and alfalfa, farming activities could be carried out profitably using fewer acres. Agriculture in general became much more intensive in the Region. Although there was a general movement toward capital—intensive agriculture affecting most agricultural systems within the northern Appalachians during the late 19405 and 19505, certain areas and types of farming were especially characterized by the capital component as reflected in the areal patterns of the component scores (Figure 23). The highest positive scores coincided with areas with more grain and specialty crop production than for the study area as a whole. The counties with considerable capital inputs clustered together in east- central Pennsylvania, and within Allegheny County and north and north- west of Pittsburgh. Several urban dominated counties, including Erie, Luzerne, Blair, and Schuylkill in Pennsylvania scored high on the use of capital inputs presumably because of the presence of vegetable and other intensive agriculture. Two counties, Centre and Tompkins, within which are located Land Grant universities, The Pennsylvania State University and Cornell University, respectively, had high measures for the capital component which could mean that the capital oriented agriculture advocated by these institutions particularly influenced the nature of agriculture on the nearby farms. Moderate and negative component scores were found for most of the dairy counties of New York; yet, Tompkins County was very much the deviant county with a relatively high capital input quantity, suggesting perhaps, that the 209 .ommunmva .ousuH:OfiHm< O>wmcoucu fiauwmmu .mouoom 03h ucocomaou mo :ofiuonfiuumwo Hmowmeccom "oopsom ooo.~om.mfi ooo.am ooo.~n~.~ av com mmm sucsou ecmHmROEHmoz ooo.omo.aa ooo.ov ooc.wmm we 025 Sam saasou couwcagmmz ooo.wo~.o~ ooo.omv ooo.oam om ova ac“ sucsou Homeosom ooo.vm~.o ooo.om ooo.o~n Ho owe me“ sucsou «accuses ooo.m-.m ooo.ov~ ooo.amm we oo am“ socsoo camuommaw ooo.NmA.o2 ooo.~mm ooo.mom.H am 000 «an sucsou acaaeca ooo.mmv.n ooo.om ooo.m~ an oN cam sucaou acumen ooo.w~m.v ooo.mm~ ooo.a~ om oo NmH season couasm ooo.mmo.m ooo.~o~ ooo.~m~ ma own mo" sucsou ouooxaa coo.am~.o ooo.mm ooo.mam an can com sucsou «apogee ooo.~mm.m~ ooo.ww ooo.mao.o Ho omo Ham sucsou “mausm ooo.wov.- ooo.~mm ooo.m~ am omN and seesaw euoeeom ooo.-a.m ooo.m~ ooo.mom Hm owfi NHN socsou po>aom ooo.mmm.m ooo.am ooo.o- as cm” and aucsou mcoeameu< ooo.mmo.m ooo.m~ ooo.qmm.v Ne oNH.H awn seesaw scoemoaa< ooo.~mm.~wa» ooo.oom.om ooo.omo.wam no m cam.mo com a venom use mommm muUSfiOhm macho Hahn» Ohu< hon Bumm>um2 ouu< Hon HQHSHHDUMHmd. umvhou mo ..HDUMHHOI mo mQOHU ~5on mmanmuomo> moanmuowo> Hauoh. mofiwm Happy. mofiwm Hmuoh. mo o=~m> mo m0k0< mo 03Hm> .moQH was vcmfl .mfl:m>meccom spoumozcusom a“ macho mo wofimm pew o:~m>--.n~ ofinmh 252 Specialization in selected crops was associated positively with capital dependent agriculture. The labor oriented agricultural component in 1969-1970. The third most important component of agriculture in 1969-70 explained more than 12 percent of the total variance found in the agricultural variable data. This dimension of agriculture had a strong labor input as labor productivity had an r = .73 and the value of hired labor had an r = .58 with the component (Table 19). The two labor variables had the highest association with the component; thus, this dimension of agriculture was primarily under the influence of labor_ inputs. Other variables somewhat related positively with the labor component were the pastureland-cropland ratio and the ratio of farm expenditures to the value of agricultural production. Therefore, counties with a major type of farming favoring the use and retention of pasture would have higher labor component scores. Where this com- ponent is of high value, the margin between farm expanses and gross farm receipts would be narrow. In 1970 there were fewer counties in the category of highest component scores and there was less positive association with urban areas than in 1950 (Figures 24 and 27). Fundamentally, however, there were few changes in the basic patterns of component scores. A large area in the easternmost part of the northern Appalachians retained moderately high and positive scores. Because of the strong position of specialized dairying in the northern two-thirds of this area, most of the labor had to be concentrated in milk production, especially in the New York State counties and in Pennsylvania's 253 .oama-moo~ .ousufisufiuu< eopcouuo Hones .mououm mouse ucmcoaeou mo coaosnuuumae aaou<--.a~ .maa Love: 23 00.0- a med 2 2.6. U 2... 2 one s .25 can on; a 254 Susquehanna and Wayne counties. Dairy farming characteristically uses labor more efficiently than do most types of farming as there is necessary work to be done on the farm the year-around, regardless of the weather and under careful management nearly an even percentage of the cows can be producing at any time of the year, establishing an even flow of income throughout the year. Because of the necessity of working a seven day week for fifty-two weeks in a year, some dairymen and their families find it impossible ever to be away from the farm for a few days of recreation and vacation unless they have obtained and trained hired help for a period of time. Traditionally there were few occupations anymore confining than dairying. Many of the smaller dairy Operators who were less able to afford or to obtain hired labor probably vacationed with their families only once or twice in a lifetime, if at all. Very few dairy farm families even had the short time required for eating out at a restaurant. These are a few of the many peculiarities of the general way of life as lived by the traditional dairy farm family that undoubtedly the many urbane people of today would find difficult to understand and accept, but for the most part were accepted by the farm folk as just a part of regular living. The largest change in the patterns of the scores between 1949 and 1970 involved the counties of the southern tier of New York and the northern tier of Pennsylvania where the increasing importance of the labor component, presumably to a major extent because of the use of more hired labor, was nearly all inclusive and quite striking (Figures 24 and 27). The north central portion of the Region had an 255 increase in the labor scores but a decrease in the capital input scores. In this northern area of the northern Appalachians it may be that capital inputs have been found not to be as adaptable to the systems of farming, the agrarian way of life, and the physical environment as labor, and there has been a resultant movement back to the use of more labor compared to capital, but final conclusions on this subject can not be rendered in this analysis because the composition of the labor component in 1970 varies somewhat from its counterpart of 1950. One way of testing such a hypothesis would involve correlating the changes in the different measures of labor and capital inputs for similar areas and the Region from 1950-70. It is quite likely that different results would be obtained for various capital inputs and that the correlations of these with labor for certain groups of counties would be higher than the Region's corre- lations. The Regions' linear correlations among three variables included in this study as listed in Table 28 are suggestive of some of the thoughts given above. Although the correlations are low, there appears to be some increase in hired labor as mechanization rises. Nevertheless, there is an extremely low relationship between invest- ment in implements and machinery, and labor productivity which is contrary to theory. Writings frequently state or imply that because of modern equipment and science applied, the average American farmer has very high values of productivity enabling him to produce enough to feed several dozen people. This low degree of association may 256 Table 28.--Linear Correlations between Capital and Labor, 1949-50 and 1969-70. Numbers 6 and 10: Numbers 6 and 19: Numbers 10 and 19: Year Value of Implements Value of Implements Value of Hired and Machinery/Value and Machinery/Labor Labor/Labor of Hired Labor Productivity Productivity 1950 .403 .133 .301 1970 .369 -.006 .610 Source: Basic data for variables calculated from United States Censuses of Agriculture, 1950 and 1969. mean that in general the Region is not well suited to the successful application of the American type of mechanization and the full bene- fits of using such cannot be achieved; therefore, where over- mechanization has occurred in relation to incompatible physical resources, negative returns occur or relatively small amounts of increased production take place as a result of the additional use of mechanization. Perhaps, for the use of machines, the point of dimin- ishing returns was confronted some time ago. Another possible explanation is that although many persons have left the farm, mech- anization may not have displaced relatively as many people in this Region as in other areas of the Nation, thus, production would be divided among more persons. The third and last association reveals the increasing importance of hired labor in expanding labor pro- ductivity, and supports the relatively greater use of labor than capital inputs in the northern portion of the Region. Some kinds of farming are particularly demanding of both capital and labor inputs. In the northern Appalachians this seems 257 to be true for the Specialty crop and vegetable farming areas, especially near large cities, e.g., Pittsburgh, Johnstown, Altoona, Scranton, Wilkes-Barre; but also near the numerous smaller cities and in the anthracite coal fields. The major changes from 1950 took place in Luzerne and Blair counties, the sites of Wilkes-Barre and Altoona, respectively, where labor declined in importance but capital inputs remained high (Figures 23 and 26; Figures 24 and 27). Two areas, differing physically and agriculturally, had most of the negative component scores in 1950 and 1970. The majority of the counties in the Ridge and Valley province of central and eastern Pennsylvania had negative labor scores in both years. Many of these counties, however, utilized many capital inputs as they gave good returns from the valley soils. This part of Pennsylvania had a concentration of poultry, dairy, and a combination of the two; nevertheless, here is located the largest concentration of grain crop production within the State outside of southeastern Pennsylvania. Grain cultivation was mechanized very early in the agricultural revolution beginning around Civil War times, and has become more capital oriented than most types of farming. This area has the largest area of poultry farming within Pennsylvania and a large area where both dairying and poultry are common. Poultry raising has significantly expanded in Pennsylvania through several decades; and ”probably in no other branch of farming has progress been so rapid and changes so revolutionary." Intensive Specialization and contract farming has especially been characteristic of this type of farming and generally the distribution of poultry corresponds with the major 258 distribution of corn and grain production (Wrigley, 1946, pp. 29-31; Alexander, 1963, pp. 202-203 and 208-209). Since the late 19405 and the early 19505 much labor saving technology has become a part of poultry farming as considerable production research resulted in the deve10pment of new production systems and techniques (Bressler, 1950). The other area which has had relatively low labor use is in the western portion of the Appalachian Plateau, excluding the environs of Pittsburgh. This is an area of semi-agricultural activity and general farming where one-third to a half of the land is cultivated but there is an intermixture of pasture and woodland subject to grazing (Pa. Stat. Abs., 1967, pp. 278 and 281). This is an inter- esting part of Pennsylvania agriculture because it is a system of farming that was neither depicted as capital intensive or labor demanding in either 1950 or 1970. Thus, in this area the major input into farming continues to be the land, and the farming is therefore of an extensive nature. The residential farming component of 1969-1970. Although this type of farming is still of little importance overall in the Region's agriculture as it accounts for less than 10 percent of the total variance in the agricultural data, it is nonetheless a slowly growing segment of the total agricultural picture in the northern Appalachians (Tables 14 and 19). Probably much of the importance of this general kind of farming with its variant subtypes is concealed by the fact that some can meet the minimum qualifications to be classified as commercial farms. Some of these farms presumably fall into the 1969 Census of Agriculture's part-time and part retirement 259 classes. Others are undoubtedly "tax write-off" farms. For example in 1950 the variable with the highest correlation with this component was the ratio of farm expenditures to the sales value of agricultural products, meaning the margin between farm expenses and gross farm sales was rather small. This variable was of second importance in 1969-70. The variable with the highest association with this com- ponent was the percent of farms reporting trucks with an r = .64; thus, the truck ownership variable was the second most important characteristic of the component in 1949-50. The possible role that trucks perform in this agricultural system is suggested by the patterns of the component scores. Several clusters of contiguous counties are depicted. Counties of northwestern Pennsylvania, Cameron, Elk, McKean, and Potter and also the counties of Monroe and Pike of east central Pennsylvania had above average positive scores. These two areas, the first in the Appalachian Plateau and the second on the Pocono Plateau are similar in many ways. Good agricultural land is at a premium; soils are generally poor; outdoor recreational activities prevail; there is a large proportion of the land that is wooded; the average value of farm production per farm is low. In such a rural economy, the truck is an all-purpose and utility vehicle. The farms of the northwestern portion of the Appalachian Plateau in Pennsylvania are among the largest in physical size within the State, e.g., Potter had the largest farms in 1969; thus, a truck can serve as transportation on the farm. The area also has the most isolated situation in respect to accessibility to cities; consequently few cityward trips are made and a car is not as necessary. 260 Two other groups of counties had high component scores: one located in the eastern portion of central Pennsylvania and the other in south-central New York. Both areas are similar because of signifi- cant crop and grain production. The Pennsylvania area has considerable poultry farming. Apparently, these farming activities have a greater use for trucks than other kinds of farming. Lastly, in the eastern portion of the Appalachian Plateau in southern and central Pennsyl- vania several counties had moderately positive values for this com- ponent, and here as in the above mentioned areas, it is surmised that the presence of trucks on the farms is a good indicator of the relative importance and dominance of local markets. The "Corn Belt” agricultural component of 1969-70. Although this component explained only somewhat more than 6 percent of the total variance in the agricultural data matrix, there are indications to the effect that if additional variables had been included in the analysis, especially ones related to crop production, the value of this component in describing the Region's agriculture would have significantly increased. Given the above and the writer's conviction that this component will become a much more important part of the northern Appalachian agricultural activities, some of the recent development associated with it, the facts, and the implications involved are presented in this section of the study. The agricultural variable most related to the component is the number of bushels of corn per harvested acre or corn yields with an r = .64. Other variables correlated with the component are the county average value of hired labor per farm reporting and the county 261 average of labor productivity per farm, both with an r = .43. Naturally as yields are significantly increased, more hired labor can be afforded and the relative amount of sales per farm person, i.e., the measure of labor productivity, will considerably increase as it is highly unlikely that additions in labor will come very near the rate of the growth in cr0p yields. One of the curious aspects of this components characteristics is the total restriction of the moderate and high scores to the western half of the northern Appalachians. Corn requires a large input of land and the land tends to be more available in the western part of the Region than in the eastern portion. The highly specialized dairy counties are poorly represented by this component. Appalachian farmers seem to be adopting more of the ill- advised practices of some of cash grain farming found in the mid- western United States. Cropland is relatively scarce in the East and particularly so in the Appalachians; and yet, the new agricul- tural techniques are primarily fitted for the best land, the crOpland. Thus, to maintain production using the new costly technology, extra- ordinary pressure is exerted to increase yields on the limited crop- land. Most farmers have become market oriented; consequently, they are today more production minded. In the shortrun a farmer can "mine" his soil and obtain more production. This questionable practice may become an increasing problem in the Region if corn is grown continuously on the same fields for several years. The use of large amounts of fertilizers 262 replace most of the nutrients; but, additional erosion, sedimentation, and excessive fertilization may cause pollution and eutrophication. Why is there a new emphasis on growing crops in the Region? As nothing as Spectacular as production increases in crop farming have come from the use of the new technologies in livestock farming, except for that associated with fowl production (Durost and Bailey, 1970, p. 5), it appears that farmers are adopting the crop technologies that can be fitted into their farming Operations. Unfortunately, the traditional benefits of conservation associated with dairy farming would be significantly reduced if the technologies first used in crop agriculture were primarily accepted. Although on a necessarily lower scale, as the physical resources in Appalachians are much less favorable than those of the Middle West, it appears the farmers of the northern Appalachians are adopting Corn Belt agriculture. Concerning that most important crop of the Middle West, it has been recently written: Up to World War II, corn typically was grown in a three year rotation of corn-oats, clover, without fertilizer . and the Corn Belt yield was about 38 bushels an acre. Today, corn is seldom rotated. Leading growers typically fertilize with 150 pounds of nitrogen . .'. . The average Corn Belt yield is now 90 to 100 bushels (Durost and Bailey, 1970, p. 3). Granted this is a brief capsule of Mid-western agriculture, but some useful insights evolve when an analysis is made of this kind of agriculture. With the emphasis on all out production, regardless of the basic motivations, the Corn Belt syndrome of farming exemplifies how easily rotation practices, and anti-erosion and conservation techniques, may receive secondary consideration. 263 Years of agricultural policy based upon the political expedi- ency of providing the cheapest possible food have greatly encouraged the emphasis on agricultural productivity. Thus, most concerns have been suppressed, including conservation, to the achievement of the highest possible production; but, because the per unit return to the farmer was generally at a minimum, the farmer has had to produce a large number of units to earn enough to be financially solvent. Systems of agriculture have been forced on many farmers by exogenous factors. What evidence is there that northern Appalachian agriculture is becoming more like that of the Corn Belt? Most surprisingly, yields of corn as grain in two-thirds of the counties of the Region in 1969-70 corresponded very favorably with the average yields per acre in the Middle West. The yields doubled between 1949 and 1969 (Table 29). Total production has continued to increase, expanding by nearly 42 percent in the short and recent period of 1964 to 1969 for all farms within Pennsylvania (Table 30). Total acreage of corn in the State declined by about 20 percent, but this decrease was very nearly made up by increases in sorghum and soybean acreage. Acreage of field corn for grain however actually advanced. Nearly half of the loss of acreage in corn for forage was made up by the increase in the acreage of sorghum forage and there was a growth in the aggregate quantity of corn forage produced in Pennsylvania in 1969. Two major facts are important in indicating the direction of Pennsylvania agriculture in general, i.e., on all farms. Production from special forage crops, e.g., corn and sorghum was maintained and 264 Table 29.-~Average Yields of Corn Harvested for Grain per Farm in Selected Counties, 1949 and 1969. 1949 1969 1949 1969 (bushels per acre) (bushels per acre) Pennsylvania 49 91 _ New York 47 88 County County Armstrong 46 95 Allegany 48 105 Beaver 46 93 Broome 46 96 Bedford 45 100 Cattaraugus 51 97 Blair 51 110 Chautauqua 44 103 Butler 51 100 Chemung 52 94 Cambria 48 94 Chenango S4 98 Cameron 44 110 Delaware 41 98 Centre 49 102 Otsego 50 99 Clarion 46 110 Schoharie 46 118 Clinton 45 102 Schuyler 44 109 Crawford 54 102 Steuben 48 92 Elk 41 108 Tioga 46 96 Erie 51 97 Fayette 49 91 Forest 44 110 Huntingdon 47 100 Indiana 46 99 Jefferson 47 106 Juniata 48 105 Lawrence 50 111 Lycoming 48 96 Mercer 50 112 Mifflin 52 108 Perry 46 94 Potter 41 105 Snyder 47 97 Somerset 47 101 Tioga 43 94 Union 48 94 Venango 48 99 Warren 46 95 Westmoreland 47 93 Wyoming , 42 105 Source: Data calculated from the U.S. Censuses of Agriculture, 1950 and 1969. 26S .ousufisofinw< mo mamceu ecu x3 vocwmov we wanna HmweueEEoo xHHmuocow .mspmm mufi mmmfiu pom mane n V "ouoz .momH paw mvma .ou:u~:u«um< mo momSmcou .m.: com: women m:0wumH:onu ”mousom Aam.eav Ame.oav ma.m meow .ommafim o~.~m om.qo ofi.me mm.mm Anny cameo mhu< Rom mvfiowr Aomo.aav.qv flawo.mmm.mv mmc.aoH.~ Acooam mcouv umaaam ovv.oov.aa oom.omv.om ovm.mom.ma Ham.2mo.mm ”any cameo cowuusghm HmuOH Aeo~.momv Amma.mamv mo~.a~m oom.mqm nmm.mm~ mamamm new use kmm.~wa msa.oaa moa.aoo.m aam.mdm.2 semen emamm>uaz mo~.ooa.a amm.~ma.a oao.~em.a Nmo.amv.~ ammouu< swoop mo.m~ nm.m~ mm.m2 on.aa :eoo eoumo>emz vcmfimowu « omn.amm.m osm.~vo.o omm.vvm.a mmm.zoo.w ecmaaouu Sauce mead seam mama veaa .mom~-vom~ ace mvmz-vvaa .ma=a>~xmccoa ca composeoua :uou as magmas--.om manna 266 increased while both acreage and tons of hay declined, 24.7 percent and 5.2 percent, respectively, and the census category of "all other land," primarily being pasture but including woodland pasture, declined by 38.8 percent. Secondly, two market valuable grains, corn and soybeans, were produced in record quantities. Corn raised for grain yielded almost 42 percent more bushels in 1969 than 1964. Soy- beans for beans expanded. Acreage went up by over 42 percent and bushels produced climbed 156.23 percent. These figures clearly dis- closed a major intensification of agriculture in Pennsylvania with specially significant growth of corn and soybeans for grain. These two crops, of course, form the base of the Mid-western agriculture. As would be expected the five largest classes of farms, the most commercialized and with the largest sales, became more like Corn Belt farms than did farms in general in the State in the late Sixties (United States Census of Agriculture, 1969). These farms accounted for 93.4 percent of the corn produced for grain and 92.3 percent of the soybeans for beans produced in 1969. Between 1964 and 1969 total production of corn for grain rose 48.4 percent compared to nearly 42 percent for all farms and the acreage of corn for grain expanded 6.3 percent. Unlike all farms together, the five largest commercial farm groups had an increase of 3.3 percent in acreage devoted to field corn for all purposes. The tons of corn used for Silage however grew by nearly 30 percent. The production of soybeans for beans at 161 percent of the 1964 harvest was similar to the production growth of all the farms. The very large recent increases in two favored feed grains on the most commercialized farms in Pennsylvania quite interestingly 267 confirms the general trend of agriculture in the area away from dairy farming and the raising of cattle--what in the past has been called the best use of the land because the physical elements of the area were especially fitted to the production of large quantities of grass and hay. Yet, these resources, once available in superabundance are being used less and less. The acreages of all major types of hay declined. Only clover and timothy showed an increase in quantity produced, but this was more than cancelled out by the fall in alfalfa and alfalfa mixtures. There was 9.7 percent less pasture acreage in 1969 than in 1964. Pastureland and rangeland (other than cropland and woodland pasture) acreage shrunk 49.4 percent while cropland used only for pasture or grazing climbed by 76.7 percent, but this increase in the latter only made up somewhat more than two-thirds of the acreage lost in permanent pasture. The increasing intensification of agriculture in Pennsylvania and the northern Appalachians which make up a large part of the State raises some exceedingly important questions. Soil erosion has been given as a local cause of migration to the city (Bowman, 1934, p. 92). In the history of United States agriculture, some cropping systems have been especially associated with erosion, e.g., in the South the crOps cotton, tobacco, and corn contributed to serious soil losses. Corn, as a row crop permits considerable erosion. In the 19505 some monoculture was being practiced and advocated, but one of the biggest disadvantages in addition to serious erosion problems was the need for a complete and timely application of nitrogen to the soil (Allaway, 1957, pp. 386-395). 268 In the 19305 much concern over the considerable erosion taking place in the central prairies or the Corn Belt was being voiced (Bennett, 1939, pp. 685-694; Leighly, 1967, p. 152). Because of the depth of the soil in many parts of the Middle West, the rather large loss of soil, although in relative terms of a moderate quantity, has been perhaps overlooked. Although much is known on how to control erosion in the Middle West, less than the desired quantity of these practices have been implemented because ". . . of a lack of sub- stantial monetary return to a farmer" for doing so (Council on Environ- mental Quality, 1971, pp. 31-33). Given that rotation practices have fallen short of their ideal use in "Corn Belt" agriculture, what are the probabilities that serious erosion will occur before the alter- native methods of conservation now known will be applied by the farmers (Barnes and Blakely, 1971, pp. 289-293). If erosion is greatly reduced, will not continuous and double cropping exhaust the soil? Is an agriculture practiced on the relatively level terrain of the Midwest feasible on a topography as uneven as that of the study area? Before generally answering the preceding questions, it should prove helpful to ask: is intensification necessary and why is the trend of agriculture in this direction? The following gives some answers: POpulation explosion; more mouths to feed. Rising labor costs; a narrow margin between cost and returns of production. These are some of the reasons why many farmers intensify their operations and crop their land more intensively (Barnes and Blakely, 1971, p. 289). In the East, in addition to corn, more of some feed grains are now (in the 19705) being raised because of the skyrocketing prices of 269 feed grain and high-protein feeds which have doubled and nearly tripled, respectively (Figure 28). The milk-feed price ratio has now reversed its direction; a pound of milk is no longer worth more than a pound of concentrate ration fed to milk cows (Figure 29). These are ominous signs for the future of dairying in the northern Appalachians. Undoubtedly, a major cause for the expansion in certain grain production within the Region in the late 19605 was the general and major increases in feed consumption (Figures 30 and 31). This general intensification of grain feeding probably motivated many of the Region's farmers to expand grain production in the late 19605. AS has been noted, large quantities of purchased feed grains have a long history in the study area. Nearly a half century ago this tendency was observed . . . with the increased demand for dairy products and with better facilities for transportation there has been a decided tendency on the part of farmers to buy western grain and raise only the necessary roughage on their places. This has enabled them to carry much more stock. The same writer also expressed the seriousness of the neglect of pastures in the Northeast which resulted in many "worn-out pastures" (Colby, 1926, pp. 208-209). The abandonment of nearly 50 percent of the permanent pastures in the Study area during the late 19605 sug- gests that relatively little has been done since the 19205 to improve these lands. As much of the Region has considerable relief, pasture was often restricted to the hilly and rough lands. The question of the effect the adoption of tractors and additional machinery had on the preservation of pastures should be investigated. Consideration 270 350 300 250 200 Percent of l967 I50 IOO 50 1960 1965 l970 l975 --- Feed grains(receivad by farmers) — High-protem feeds(wholosala at principal markets) A Preliminary Fig. 28.--Price Trends of Feed Grains. 271 200 I40 Percent of |967 l20 80 P........OmQII'QOO.......... I960 I963 I966 I969 I972 I975} A — Milk-feed price ratio (pounds at ratio equal in value to one pound at milk sold to plants) -—--— Ration vaiue(concentrate ration ted to milk cows) Farm milk price A Forecast Fig. 29.--Milk-Feed Price Ratios. 272 ZOO ISO ’__“__,sa”iflhul;;flua. g:\»/’. m C ..~.. '. ...... N. 2 r- .'/ .1000...0.O..... I ’./.’ ~0\‘e‘ .... .. 0‘, . I ..e “5 100 h-'°“"-m_.. .' "" ‘J 5v": .... -.-.’. . ,w--—. 2 F \'-....I .9 5 50 O 1960 1965 1970 I975 A Wheat Sorghum grain ..--.. Byproduct feeds -.-... Corn -—---— Oats and barley A Forecast Fig. 30.--Trends in Feed Grains Consumed. 273 25 20 I‘ V\ ’”--- ---~‘ ”‘~‘ KT—j--1-“-”'T .._...—... l/w~ 7 '5 2" T /,.’... ..............., “......vli-th." ” ' ,. °' ‘* . .’s. -‘.‘4’.’ \ p"’:......... ..- boee..¢° ' 0......0’y .49‘.‘.-.-.-.-. IO 7 Millions of tons O 1964 |966 1968 l970 1972 1974 Grain proteins(gluten feed 8 meal; brewer 8 distiller dried grains) -- -- Other animal prateins(tankage, meat meal 8 milk products) _.._ FIshmeal Other oilseed meais(cottonseed,linseed, peanut 8 copra meals) ---°-° Soybean meal Fig. 31.--Trends in Sources of Protein Consumed. 274 of this question is given in the following discussion of the last component of northern Appalachian agriculture in 1969-70. The mechanization and motorization component of 1969-1970. This component accounted for more than five percent of the total variance, a small sum, but the association r = .68 of percent farms owning tractors with it and the areal distribution of the component scores, requires at least some description (Table 19). Interestingly, the variables having the next highest correlations with the component, percent farms reporting trucks and also autos, gave the component a strong motorization appearance. As the highest and moderate component scores are primarily located in counties of the northwestern and northeastern portions of Pennsylvania and a few adjacent counties of New York where specialized dairying dominates the agriculture, a major implication is the exceedingly important position that the tractor occupies in the dairying activity. The tractor is heavily relied upon in the rising harvests of forage crops on such farms. Certainly it is the single most important investment on dairy farms; however, as previous analysis of the data has illustrated, the expanding range of possible inputs into farming has lowered the relative importance of the tractor as a single indicator of capital investment. As the dairy areas of the Region had traditionally a large percentage of its farm lands in pasture, interaction between pasture land continuance and use and tractor power necessitates consideration. Did the expanded use of tractors and other machines distract from the proper utilization of pastures? Some evidence suggests that the 275 abandonment of pastures was partially related to the increasing adoption of the tractor. Correlations between percent of farms reporting tractors and the pasture-crop land ratio give a change of relationship starting with a moderately negative relationship in 1929-30 to really no relationship in 1969-70 (Table 31). This may mean that the pastures with the most adverse topography and with the poorest grass production were abandoned to the greatest extent soon after the initial phases of tractor adoption. The use of tractors released some cropland that had been used for horses, but it also resulted in some growth in production from cropland. Therefore less pasture land was needed. In general it appears that value of machines and implements on farms has continued to be associated with a decreasing pasture-crop land ratio (Table 31). Table 31.--Temporal Correlations Between the Degree of Tractor Acceptance or Mechanization and the Pasture-Cropland Ratios. Average Investment of Machinery and Implements Per Farm with Pasture- Crop.Land Ratio Percent of Farms with Year Tractors with Pasture- Crop Land Ratio 1929-30 -.57 -.21 1949-50 -.37 -.53 1969-70 .08 -.48 CHAPTER 8 EXPLANATIONS FOR PERIODIC AGRICULTURAL POPULATION CHANGES IN THE NORTHERN APPALACHIANS, 1909-1970 Form of the Explanation To what degree through time for the several areas of the northern Appalachians are the major characteristics of agriculture associated with the relative preservation or loss of farm population? The agricultural characteristics are derived from a principal com- ponents analysis of selected agricultural variables for each of four time periods. In the preceding chapters these grouped characteristics, i.e., components, are interpreted and geographically analyzed with the expectation that such analysis of the Region's agriculture would lead to additional hypotheses for the varying farm pOpulation changes. Relative change in agricultural population is measured as the percentage each corresponding agricultural population of a county is of that of a previous census. Thus, an indicator of the relative degree of "farm population retention" for each county within the study area results. Principal component analyses of agricultural data for each of the 1909-10, 1929-30, 1949-50, and 1969-70 periods produced component scores which are the inputs for stepwise multiple regression 276 277 and serve as independent variables. The farm population retention percentages represent the dependent variable, i.e., farm population change, for each of the major time periods. The results of several exploratory stepwise multiple and linear regressions using component scores of one period with agricultural population change of a subse- quent period are presented in a special effort to determine if changes in farm populations occur relatively a short while after alterations in the agricultural systems or whether in actuality a considerable time lag in reaction represents reality. If significant high or moderate coefficients of multiple correlation (R's) are obtained from a stepwise multiple regression, the results of an analysis of mapped residuals are presented. The Nature and Application of the Stgpwise Regressions First, a brief and general statement of the function of a stepwise multiple regression is necessary to the clarification of the analysis applied in this study. For each selected time period, each independent variable, as in this Study an agricultural component, is chosen sequentially according to its importance, that iS in the reduction of sums of squares (Wittick, 1971, p. 6). For each particu- lar time period, the independent variable first chosen is the one that has the leading simple correlation with the dependent variable (King, 1969, pp. 145-148). The independent variables, i.e., the agricultural components, are considered and entered into the regression in turn and sequentially according to the one which most 278 reduces the unexplained variation remaining in the dependent variable, i.e., the farm population change. Temporal Associations of Agricultural Components with Farm Population Retention Ratios A series of multiple regression analyses using component scores of agricultural data as independent variables to explain dif- ferences in county farm population retention ratios produced a wide range of findings. In general, farm population changes in the first third of the Twentieth Century were only slightly related to the basic dimensions and nature of agriculture within the northern Appalachians. Later changes in the agricultural population of the northern Appalachians were associated with some major characteristics of the Region's agriculture. Several possible explanations for this major finding, i.e., the farm population was only recently much affected by the nature of agriculture, are logically possible. Because of the considerable specialization taking place within agriculture, more diversity pre- vailed in farming activities and operations in the last several decades than in the earlier years of this century. The wide pre- ponderance of self-sufficing farms and general farms in the past, the majority of which had very similar characteristics, precluded most effects that different agricultural establishments would have had on different farm population change rates. In addition, major worldwide disruptive forces, e.g., World War I and its carry over effects into the early 19205, the Great Depression in the 19305, and World War II in the 19405 all but submerged any influences that a few varying 279 characteristics of agriculture could have had on the decisions of individuals to stay in or leave farming. Major Agricultural Attributes Corre- lated with EnsuingLFarm Population Changes Decades of 1910-1940. The availability of cropland from 1910 to 1930 was more related to farm population changes than any other agricultural characteristic included in this study (Table 32). Nevertheless, the degree of correlation was low. Although there was a decline of farm population during the decades making up this time period, the loss was at a modest rate compared to the more recent decades as illustrated in Chapter 111. As such, given the higher birth rates on the farms during the earlier time, there must have been some recognizable pressure upon young men to obtain available farm lands. Land during this time was relatively more important as an input than after World War II when a number of inputs, including fertilizer and higher yielding varieties, made possible greater pro- duction on fewer acres. Traditionally, the major means for increasing total production was by expanding the acreage cultivated. As the farms became an the average smaller, e.g., from 1920 to 1925 the average size of farm declined in nearly every county of Pennsylvania, some farm persons had to live at a lower standard of living. There occurred also an increase in the number of farms in most counties of central and western Pennsylvania, while at the same time the number of acres in farms declined in nearly all Pennsylvania counties (see Census of Agriculture for Pennsylvania). Although there had been 280 Table 32.--Correlation Matrices for Selected Agricultural Components and Farm Population Retention Rates, 1909-1930. 1909 to 1910 agricultural components with 1910 to 1920 farm population change Retention rate 1.00 .13 -.17 .39 .05 Investment .13 1.00 -.01 .00 .00 L1Y°St°°k ?“d -.17 -.01 1.00 .00 .oo dairy farming Crepland availability .39 -.00 .00 .00 .00 Component 4 .05 -.00 .00 .00 .00 1909-10 agricultural components with 1910-30 farm population change Retention rate 1.00 .19 -.22 .36 .22 Investment .19 1.00 -.01 .00 .00 L1Y°St°°k ?“d -.22 -.01 1.00 .00 .oo dairy farming Cropland availability .36 -.OO .00 .00 .00 Component 4 .22 -.00 .00 .OO .00 281 some abandonment of agricultural land within the Region prior to 1910 (Hoglund, 1953; Vaughan, 1929), acres given up in the first decades of this century were usually quite marginal for agriculture. Thus, where additional cropland could be obtained, such areas tended to hold upon the land a relatively greater number of the farm folk. In both 1910-1920 and 1910-1930, livestock and dairy farming, the second most important component of agriculture within the Region had a low and negative relationship with farm population change (Table 32). The greater the presence of this component within an area, the greater the chance that more loss of farm individuals would occur than in the Region as a whole. During the early decades of this century, the cumulative relationships of the agricultural components to the total variation of the pOpulation changes were low. Although a wide choice of agri- cultural variables were selected and utilized in the analysis, including measures of technology, labor, fuel, investment, productivity, type, and physical size of farms, the quality of data was undoubtedly not as good as the data of later decades. In addition, not all vari- ables placed in the analysis to represent the agricultural situation in the later decades were available for the 1909-1910 year. Also, 1910 farm population had to be estimated. Consequently, not as a complete view of agriculture was feasible in the earlier decades as at mid-century. Possibly if a more complete and accurate portrayal of agriculture had been obtainable, agricultural characteristics would have been more related to farm population changes. 282 Although a major hypothesis of this study Stated that gen- erally increasing acceptance of new technology forces more persons off the farms, the investment component containing a summation of various investment variables--regarded as representations of new technology--was generally the most poorly associated with the agri- cultural population changes from 1910-1940 (Table 32), and gave also generally the least amount of explanation to the total variation found within the dependent variable. Investment had however for the period of 1909-1910 been selected by the principal components analy- sis as the aspect most accounting for the variations within agriculture (Table 12). During this period the use of capital was of much less importance relative to other major inputs than it was later to be and this fact may help to explain its poor performance in accounting for population losses from the farms. To be remembered also was the amount of capital tied to the use of draft animals, particularly horses, whose numbers really did not decline significantly in the Study area until after World War 11; thus, the general nature of technology remained primarily unchanged. The 1950 decade. Contrary to the rather meager correlations of agricultural components with farm depopulation in the 1910 to 1930 era, a high degree of correlation, r = .854, was found between the first component, dairying:and off-farm work dichotomy of 1949- 1950 and agricultural population change of 1950-1960. Although labor oriented agriculture and residential farming_(Table 14) were only Slightly but inversely associated with farm population retention 283 (Table 33), both added to the above and to the stepwise multiple regression analysis resulted in an R2 of .826. Table 33.--Correlation Matrix for Selected Agricultural Components, 1949-50 and Farm Population Retention Rates, 1950-1960. Retention rates 1.00 .85 .01 -.26 0.17 -.02 Dalrylpg and °ff'f?rm .35 1.00 -.00 -.oo -.oo .00 work dichotomy ratio capital 1nte"51ve .01 -.oo 1.00 -.00 .oo .oo agriculture Lab?’ °rlented -.26 -.oo -.oo 1.00 -.00 .oo agriculture Residential farming -.17 -.oo .oo .oo 1.00 -.00 Extensive farming -.02 .00 .00 .00 -.00 1.00 Some general agricultural characteristics can be connected to farm depopulation through some inferences made from the above statis- tical analysis of the 1950-1960 decade. A large part of the Region was agriculturally typed as dairy and where this was true, off-farm work was less than in other areas. As the intensity of dairying increased, work by farmers off the farm decreased and farmers remained in farming much more than did farmers in general. Studies have indicated that off-the-farm work by agriculturalists has led many farm people away from farming (Hathaway and Waldo, 1964, p. 45; Hathaway and Perkins, 1968, pp. 185-189) and the findings of this study support this contention. During the 1950—1960 time period the more dairying in an area the less was the loss of farm pOpulation, but in contrast where dairying was of little importance and 284 off-the-farm work was high, farm depopulation ordinarily was Signifi- cant. Comparison of the farm population change map (Figure 9) and the agricultural component map (Figure 21) revealed that the dairy counties of eastern and northern areas best retained farm persons; whereas, the counties of high off farm employment in western Pennsylvania had high depopulation. Other than the high association of work off the farm with decreases in the farm population there appeared to have been a number of minor influences or factors contributing to the loss of farm people during the 1950-1960 decade. This observation is supported by the components that had low negative correlations with retention of farm population (Table 33). Interestingly, the expected adverse effects of relative proximity of a farm population to an urban area upon the retention of the agricultural population was found in a study not to hold during the last two decades in the Northeast where the highest off-farm mobility rates for employment were found in the counties most distant from a SMSA (Hathaway et al., 1968, p. 188). Thus, during this period of off-the-farm employment opportunities, especially near cities, may have tended generally to keep farmers in farming somewhat longer in the northern Appalachians than if few "outside" opportunities had existed, although such work eventually led to many leaving farming. This writer found, for 1945-1959 in Pennsylvania, in a Study of agricultural land abandonment, a phenomenon closely associated to farm depopulation, that off-farm employment with an r of -.77 gave the most explanation and that rural isolation accounted for some of the agricultural land abandonment (Slocum, 1969, 285 pp. 65—66). Because farm depopulation and agricultural land abandon- ment are similar, understanding of one could perhaps aid in exploring the nature of the other. Inferential evidence however generally suggests tha- the conditional or push forces were more important in accounting for the very large decrease in the Region's farm population during the 19505 decade than the pull forces. Again the indice of agricultural investment, formally called intensive agriculture, had no relationship with either the retention or loss of farm population. Increases in the amount of capital invested in a farm did not permit more persons to work on the farm nor did the relative decline in the amount of capital on the farm decrease the number of farm persons. In the former instance, many farmers expanded the scale and size of their enterprise(s) by using more capital, but often did not hire any more workers. Also, as capital could be substituted in many situations for labor, there was less obvious need for a farm couple to have a large family to provide labor on the farm, when the size of the farming operation was expanded. In the second instance, stage of life of the farmer or the family cycle however often influenced his use of capital (Bennett, 1969, pp. 228-230). The median age of farm operators rose in recent years (Nikolitch, 1967). Older farmers, many of whom had grown families, did not experience the necessity of expanding their farming operations so as to better provide for and support their family. Older farmers in general desire to "slow down" and are less interested in having a large and efficient operation as their own health and physical conditions decline with the advancement of age. As the 286 competition from larger commercial farms increased, small scale farmers found that the prices and earnings of agricultural products remained depressed and/or stabilized and tended to give a marginal existence that some farmers, particularly the older ones, improved upon and adjusted to through disinvestment. Cattle were generally sold off. Perhaps, a second tractor was sold. Possibly several house lots were surveyed and disposed of with varying degrees of compensation. Maintenance of buildings often were relinquished. Other examples could be cited. Much of the capital orientation of agriculture came following World War II and took a number of years to reach a high level of intensity; therefore, the effects of these inputs in the full decade after the 19405 upon the demographic com- ponent of the general agricultural system had not had enough time to have an impact. The adverse effects that capital and technological inputs have on the endurance and the persistence of the numbers of farm persons accumulate gradually and are stretched out over a rela- tively long time. Net Status of Agricultural Attributes Correlated with Preceding Farm Population Changes The conditions of agriculture at the end of a decade were thought to be possibly better representations of the general trend of the agricultural changes occurring during the same period that the farm population was declining than the conditions at the start of that time. The relationships found in this part of the analysis would represent the more concurrent effects of agricultural changes 287 upon agricultural population changes. AS technological effects have been assumed to take a relatively long time to affect the population numbers, as concluded in the above analysis, no important correlation of the agricultural component representing the input of technology was expected with farm population change, but correlations with the other components were thought likely. The 19205 decade. During this decade of an expanding economy and off the farm Opportunities, changes in agriculture could be expected to explain a very small proportion of the total decrease in the farm population. Nevertheless, self-sufficing farms and dairy farm systems (Table 13) were negatively related to the preservation of the farm people (Table 34). These two types of farming contributed statistically about 14 percent to the total possible explanation. In contrast intensive agriculture had some influence on keeping people in agriculture, possibly because during this decade average size of farms and the amount of land in farms decreased in many of the counties, and through this type of farming, persons were able to con- tinue in agriculture. Table 34.--Correlation Matrix for Selected Agricultural Components, 1929-30 and Farm Population Retention Rates, 1920-1930. Retention rates 1.00 .23 -.20 —.32 -.01 Intensive agriculture .23 1.00 -.02 .00 .00 Dairying -.20 0.02 1.00 -.00 -.OO Self-sufficing farming -.32 .00 -.00 1.00 -.00 Part-time farming -.Ol .00 —.OO -.00 1.00 288 The 19405 decade. Dairy farming in contrast to earlier decades had a positive effect on retaining farm population and was associated more with population change than any other agricultural system or attribute (Tables 14 and 35). Truck farming and/or specialty crop systems perhaps were associated with losses of agricultural pOpulation because of their intensive characteristics and demand for labor which, because of the War, was scarce. As labor availability problems existed in agriculture, there apparently was some return to more extensive farming which helped to retain the population engaged in it. Table 35.--Correlation Matrix for Selected Agricultural Components, 1949-50 and Farm Population Retention Rates, 1940-1950. Retention rates 1.00 .39 -.07 -.24 .05 .24 Dairy and off-farm work dichotomy ratio .39 1.00 -.00 -.00 -.00 .00 cap?tal Inte"Sive -.07 -.oo 1.00 -.00 -.oo .oo agriculture Lab?’ °rlented -.24 -.oo -.oo 1.00 -.00 .oo agriculture Residential farming .05 -.00 .00 -.00 1.00 -.00 Extensive farming .24 .00 .00 .00 -.00 1.00 The 19505 decade. With 1970 components of agriculture as independent variables a multiple correlation (R) equaling 0.785 with the farm population change of the 19505 appeared. The characteristics of 1950 agriculture retained in 1970 agriculture were closely con- nected to the changes in the numbers of persons on farms during the 19505. Interestingly, dairy farming without work off the farm 289 accounted for a large part of the relative retention of population within agriculture (Tables 19 and 36). The other agricultural com- ponents selected by the stepwise regression all worked against the retention of agricultural labor or encouraged farm depopulation. The selection of a technological summary indice, mechanization and motorization, with an explanation of nearly 9 percent of the total variation found within the agricultural population change data was the first indication of support for the basic technological-related hypothesis in this study. Technology seemingly has had a small negative effect nearly coinciding with population change upon the short term maintenance of the farm population, but apparently these effects are not readily measured in a direct sense. Heavy reliance on capital representative variables as indices of technological developments probably was not sufficient to give indicators of separate Table 36.--Correlation Matrix for Selected Agricultural Components, 1969-70 and Farm Population Retention Rates, 1950-1960. Retention rates 1.00 .70 .02 -.13 0.15 -.09 -.29 Dairying and off-farm work dichotomy ratio '70 1'00 "02 '00 '°00 ‘-00 -.00 Capital intensive agriculture .02 -.02 1.00 -.00 .oo -.oo .oo LaPPT °rlented -.13 .oo -.oo 1.00 -.00 -.oo .oo agriculture Residential farming -.15 -.oo .oo -.oo 1.00 -.00 -.oo C°TP Belt -.09 -.oo -.oo -.oo -.oo 1.00 -.03 agriculture Mechanization and _ 29 _ 00 00 .00 - 00 - 03 1 00 motorization 290 and entire technical influences. For example real estate values have often reflected the proximity of an area to a city, and evaluation of the property has been based upon its potential worth for urban residential, industrial, and/or commercial land uses. These higher values are weakly related to actual improvements made on the lands and buildings. The two additional agricultural farming systems, residential farming and labor oriented agriculture, had only low corre- lations of a negative nature, with farm depopulation. Because of the human resources demanded for World War II, labor scarcity was trouble- some on into the 19705. Other new developments arose, including residential farming, some of which, because of the newness of the interest, presumably resulted in disillusionment on the part of some new arrivals on to farms who stayed for only a short time. The 19605 decade. Agriculture in the northern Appalachians appears according to the principal components analysis to have become more heterogeneous recently as more components were extracted from the agricultural data than for the earlier decades. A stepwise regression analysis indicated that nearly all of these basic aspects of agriculture were either related partially to farm population retention or decline (Tables 19 and 37). A multiple correlation R = 0.733, suggests that in contemporary times the increasingly varied nature and most important characteristics of agriculture are increas- ingly related to the growth or decline of agricultural population numbers. As in the 19505, during the 19605 those areas with much dairying and little off the farm employment tended to keep farmers on 291 Table 37.--Correlation Matrix for Selected Agricultural Components, 1969-70 and Farm Population Retention Rates, 1960-1970. Retention Rates 1.00 .38 .34 -.37 -.29 -.23 .05 Dairying and Off Farm Work Dichotomy .38 1.00 -.02 .00 -.00 -.00 -.00 Capital Intensive 34 _ 02 1 00 _ 00 00 - 00 00 Agriculture ° ° ' . O . . Lab?’ Orlented -.37 .oo -.oo 1.00 -.00 -.oo .00 Agriculture Residential Farming -.29 -.00 .00 -.00 1.00 -.00 -.00 Cor? 381‘ -.23 .oo -.oo -.oo -.oo 1.00 -.03 Agriculture ”9°hanlzat1°" and .05 -.oo .oo .oo -.oo -.03 1.00 Motorization the land but much less so, however; the dairy farming and the off fa rm work dichotomy component explained more of the variation in the farm population change than any other agricultural variable, R2 = 0.146. The second most important agricultural variable associated with population changes on the farms was labor oriented agriculture. Tho enterprises using comparatively more labor than other types of agri- culture especially lost persons employed in their operations. This aspect of agriculture nearly equaled the most important agricultural variable for power of explanation and gave the most important farm- related reason for farm depopulation. Labor continued to be among the most expensive inputs of agriculture and where cheaper substi- tution could be made, often with an aspect of technology, this often occurred. However, labor received considerably lower wages on the farm than in nonagricultural employments; many such farm employees 56 292 and owners were thus attracted to nonfarm work. A curious relation- ship is that between capital intensive agriculture and the relative increase and strength of the retention of persons upon the farms. Possibly during the 19605, the use of additional capital and tech- nological inputs, especially low cost energy, became important for doing the necessary work for keeping on the farms the residual popu- lation that was left after several decades of heavy outmigration. The farmers with savings to invest in their farming operation or with access to capital may have been the farmers who most successfully adjusted to the migration of labor. Those individuals who engaged in residential farming may have only temporarily prevented farms from going out of production as this system of agriculture formed a negative relationship with farm population changes. "Bona fide" farmers sometimes sold to persons who would farm for a short time, and then retire; using the farm as their residence, these persons would rent out the farm land to nearby farmers. In addition to residential farming and labor oriented agri- culture, "corn belt agriculture," 3 type of agriculture only delineated in the 19605 agricultural data, was associated with agricultural depopulation. Interestingly two labor variables, value of hired labor and labor productivity, along with corn yields, were found earlier in this study to be basic characteristics of this farming system. Geographical Analysis with Residuals The residual represents the difference between the dependent variables actual value for each observation, i.e., of a county, and the value estimated by the regression equation. A regression using 293 the independent variables may overpredict an observation's dependent value or underpredict it, producing a negative or positive residual, respectively. Thus, the retention of farm population is overpre- dicted if the value of the residual from regression is negative and underpredicted if this value is positive. Therefore, the values nearest to zero in certain counties mean that these counties' farm population retentiveness is very nearly explained by the selected independent or explanatory variables. Mapped residuals present an excellent measurement of the effectiveness of a regression model in accounting for areal variation in a phenomenon (Thomas, 1968), which in this study is the prOportion of the agricultural population staying on the farm per county from one time to the next selected time. For some of the time periods studied, the initially selected agricultural component is well correlated to the dependent variable. In these instances, it is thought especially revealing of the rela- tionships of certain types of farming to agricultural population changes to delete all of the components which reduce the unexplained vari- ation, except the one most related to the population variable. Then, the residuals produced for the counties are associated only with the component giving the most explanation, and the areas made up of counties particularly affected by this most influential agricultural component can be determined. The same procedure may be followed with the second most important component in explaining the population changes. Only after some components have proven their worth by accounting for an important proportion of the total variation in a regular and undirected run of the stepwise multiple regression, are 294 they tested individually and the areal distribution of their resid- uals analyzed. In this manner the areal significance of individual agricultural components in changing agricultural population is dis- covered. Comparisons of those counties whose agricultural populations are associated with single components as opposed to a collective number of components, presents a portrayal of areas with perhaps a major cause of farm population change and those areas with multi- tudinous causes. Once the above areas are recognized, then the patterns of the counties whose retention of agricultural populations or depopulation are unassociated with any of the agricultural com- ponents, serve as the basic clues for generating additional hypotheses and for selecting sites for field studies. The Relative Areas Associations of Farm Population Changes to the Selected Agricultural Components By mapping the residuals of the multiple regressions and of a few selected linear regressions, basically three different county groups appear--namely those counties whose farm population changes are well accounted for, either by a composite of agricultural components or by an agricultural component, i.e., with values between -4.9 to +4.9 and those two county groups whose farm population changes are significantly "under explained" or "over explained" and which have values of -l0.0 and less or +10.0 and over, respectively. Generally, the higher the multiple correlation, the more the number of counties whose values of agricultural population change are explained by the agricultural components. Often the counties with dependent values, 29S i.e., farm population retention or change, closely associated with the independent variable(s) will be contiguous counties as will the group of counties with values most deficiently explained. A cluster of counties so depicted often will have similar agricultural and geo- graphical characteristics; a knowledge of these data may aid in the establishment of new insights in describing and explaining the associ- ations between the dependent and independent variables. In a like manner, the geographical nature of the subareas with farm population changes most inadequately explained by the regression model, may give clues for the establishment of new hypotheses that when tested may provide additional explanations for the farm population changes. Ideally, each county's farm population change should be related to one or more agricultural components or attributes. Those counties with farm population changes unassociated with agricultural develop- ments merit intensive field study. Areal Relationships of Subsequent Farm Population Changes to Agricultural Components The areas of explained agricultural population change, 1910- .12§9° For the 1910-1930 period the agricultural component of cropland availability was the attribute most related to farm population changes, explaining about 13 percent of the farm population varied changes. Two other agricultural attributes, livestock-dairy farming and diversified farming added together accounted for about 10 percent more of the variation in the dependent variable. 296 What counties had farm population changes closely associated with these agricultural characteristics? The counties best explained by cropland availability and also the additional agricultural character- istics that had residuals between -4.9 and +4.9 were generally located within the central part of the northern Appalachians, except for a few counties in the northwestern area and a set of contiguous counties in the Triple Cities sub-region on New York State (Figure 32). Interestingly, these "explained" farm population change sub-regions correspond with the areas that had average farm pOpulation losses from 1910-1930. Consequently, in the counties with average farm population decreases, farm persons had a tendency to stay in farming where there was an adequate supply of agricultural land. Most of the counties with average losses tended to have diversified farming. Possibly where dairying was on the increase, it was replacing some small general farms as dairying required larger acreages (Ellis, 1967, p. 277). Some counties, e.g., Broome, Erie and Luzerne with large cities such as Binghamton, Erie, Wilkes-Barre, respectively and some adjacent counties had significant decreases in their cropland acreages; however, the most urban counties retained more farm population than would have been predicted by only the presence of potentially useful cropland. Nonurban counties whose farm population changes were reason- ably associated with the selected agricultural components had a con- siderable concentration of farming activities within fertile valleys and/or in general and diversified farming. 297 m - ICC and loss ’71 -9.9 to -5.0 CI] -4.9 to 4.9 m m 5.0 to 9.9 I0.0 and over Fig. 32.-Residuals from Regression of Retention Rates of Farm Population, 1910-1930, on Scores of Associated Agricultural Components, 1909-1910. 298 Areas with high negative residuals, 1910-1930; some hypotheses. What explanations can be given for those areas with unexplained farm population changes? Because the multiple correlation of the com- posite agricultural components with county farm pOpulation change was of only moderate size, i.e., R = .52, it was to be expected that there would be many counties whose population changes on the farms would not be associated with the selected agricultural characteristics. Two such areas existed: one in the north central sub-region and another in the eastern and southern margins of the Region. In the north- central sub-region at the end of two decades, there were far fewer persons left on the farms than would have been predicted by the effects of the agricultural components found to be generally related to the population changes. Some nonagricultural factors were obviously important in lowering the labor force on the labor force on the farms in these counties. In the north central area of Pennsylvania was located many marginal areas for farming. Much of the land, because of infertile soils and short growing seasons, was unfavorable for economic pro- duction. Numerous acres should not have been cleared for agriculture. The area had many townships settled relatively late as did also "the infertile and rugged Pocono Plateau" (Gould, 1969, pp. 46-47; Florin, 1965). The area had less than an hundred years ago a major lumbering industry. The widespread removal of the forest cover sub- jected the soils to erosion and deprived much of the area from a dependable livelihood from either forestry or farming. Therefore, 299 there were many failing attempts of farmers to farm land which was frequently marginal for commercial agricultural production. A farmer's increasing efforts commonly result in worsening the erosion problem. "Soil erosion takes place so gradually that most people, including the farmers themselves, are unaware of it." From 1900 to 1931 in Pennsylvania more than 51,000 farms containing at least four million acres were lost to agriculture because of soil erosion. Some areas had upwards to 25 percent of all farms abandoned. By 1930 more than a million additional acres were considered sub- marginal due to the serious damage of erosion (Stevens, Cordier, and Benjamin, 1953, p. 358). Some areas were naturally affected more than others. The north central counties of Pennsylvania were particularly adversely affected by soil erosion (Patrick, 1938, p. 10). In addition, some public land policies and decisions deter- mined the availability and use of agricultural land. Pennsylvania at the end of the last century began the purchase of vast acreages of diverse types of land, including tax delinquent parcels, abandoned farms, and cutover forests. The exact proportion of each of these types of lands remains obscure, but most lands appear to have been mostly deforested or nonagricultural. Until about 1870 Pennsylvania was the leading producer of lumber in the U.S.; and seemingly, it was not until after the leading lumber companies had moved out of the State in the late 18005 that the great importance of harvesting locally produced timber was recognized (Stevens et al., 1953, pp. 349-353). The State was then regarded as the only organization large 300 enough to own and manage forest lands and to promote the conservation and reforestation programs. Most of these new public lands became State Forest, State Game Lands, and/or State Parks. The north central sub-region, the largest area of significant farm population decreases from 1910 to 1930, had in the first decade of the 19005 at least one half of the approximately half a million acres owned by Pennsylvania (Pa. Dept. of Forestry, 1902, pp. 11-25). The acquisition of forests and potential forest land for public control was ongoing for many years, partly because of farm abandonment. All the lands at first were designated as "reservation lands" and no formal division existed; however, given the general location of these lands in much of the roughest terrain of the State, most of the acreage became State Forests. Nevertheless, it is interesting to note, that the areal pattern in the State lands was established in the beginning; nearly all of the lands purchased by the public were situated in the central third of the State, coinciding with the highlands and the ridge and valley section of the State. The major exception to the location of these lands was the Commonwealth's purchase of lands in the Pocono Highlands area of Pike, Monroe, and Lackawanna counties; and New York State established a forestry preserve in the Catskill uplands, a part included in the extreme northeastern counties of the northern Appalachians (Ellis, 1967, pp. 504-505). Possibly, the public ownership of lands deprived some of the area's agricultural population of supplemental earnings from the land, resulting in some financial difficulties for farmers. Of 301 importance was the interdependence of some farms and the lumbering industry. Local agriculture supplied the forestry workers with food stuffs and other services. Farmers raised horses, oxen, and mules; sheltered the draft animals in the nonlumbering periods; and provided the hay and some of the grains to the woodsmen's animals. Once lumbering folded, the local markets contracted. Insufficient trans- portation and the marginal nature of farming within the sub-region prevented successful participation in the national market. Agricultural activities and agricultural systems, before the development of widescale transportation facilities and refrigeration, were an integrated part of the overall rural system and owed their existence especially in the areas of least favorable physical con- ditions to the presence of other well-known rural economic activities, such as forestry, mining, and additional mineral resource workings. As to the latter, an interesting question is the effect that the oil discoveries must have had on agriculture in western Pennsylvania. Agriculture in the more physically unfavorable areas, where forestry and mining prevailed, existed often to provision the woodsmen and miners. Note has often been made of the countless men attracted to the American West by gold discoveries who found the raising and pro- ducing of crops and agricultural products for the local mining popu- lation more lucrative than mining. Few words appear to have been written on the importance that rural primary economic activities have had on the existence of agri- cultural systems in the eastern United States. As Pennsylvania had historically a strong forestry and mining economy, the above discussed 302 influence greatly affected the nature and longevity of farming in some of those areas particularly characterized by such rural primary economic activities. When these basic industries declined farming after some lag in time did also, particularly after the passing of the older generations whose persons were often "trapped" in farming as they were often too old and/or without skills to qualify for other occupations. Areas with high positive residuals, 1910-1930; some hypotheses. According to the map of residuals for the period 1910 to 1930, south- western Pennsylvania with the exception of Allegheny County, retained considerably more of its agricultural population than predicted by the agricultural components, e.g., cropland availability, dairy farming, etc. (Figure 32). This sub-regions farming possibly was not well defined by the agricultural primary data and the agricultural com- ponents of 1910. The agricultural characteristics varied in importance from county to county within the area; suggesting that generalization on agriculture for this subdivision of the northern Appalachians was unrewarding. There was a higher retention of the farm population in south- western Pennsylvania and in the anthracite area than any other parts of the Region, e.g., the counties of Lackawanna, Luzerne, Carbon, Schuylkill, and Northumberland (Figure 7). In 1930 most of these coal counties had large farm populations compared to other counties and also, large urban populations. Because of the more locally focused economies at this time compared to those that were to follow, 303 the farm population found readily available markets for their farm products and services in the heavily populated mining communities. As in the earlier positive hypothesized association of forestry's successful development and decline with the respective fortunes of farming, a very similar postulate can be stated for the overall relationship between mining and farming before the advent of modern mechanization and techniques which greatly reduced total employment in mining. Farming apparently benefitted as long as coal production increased, was labor intensive, and the mine population expanded. World War I boosted the mining economy of Pennsylvania and it remained relatively healthy until after 1930 when Pennsylvania lost its leadership in soft coal production (Stevens, 1964, p. 348). Coal mining however saw some omens during the 19205. The number of miners employed in coal mining and the price of coal declined. Agricultural prices decreased also, but where farms were more accessible to markets, as in southwestern Pennsylvania, agriculture became more diversified, and agricultural income stabilized. Although mining jobs declined, "unemployment in mining areas usually entailed fewer working days rather than a complete discharge, which kept miners from looking for jobs elsewhere" (Klein and Hoogenboom, 1973, pp. 396-397). In addition, the declines in mining late in the 1910- 1930 period had only negligible effects on the agricultural population as sufficient time had not passed for the reactions in agriculture to be noticeable. The map of residuals revealed for both hard coal counties of northeastern and soft coal counties of southwestern 304 Pennsylvania that the farm population stayed much more than would have been expected with the agricultural characteristics present. The areas of explained agricultural population change, 1950 to 1960. The variation in the county pOpulation retention ratios was highly accounted for, during the 19505 by agricultural characteristics; the dairying and off farm work dichotomy component of agriculture alone formed an r of .854 and statistically accounted for nearly 73 percent of the total variation found in the farm population change values. With the inclusion of truck and/or crop specialty farming, plus residential farming R equaled .909 and the amount of the total variation explained by agricultural components rose to nearly 83 percent. With the above pertinent agricultural components acting in concert, two-thirds of the counties of the northern Appalachians had residual values near to zero (Figure 33). These counties formed an areal pattern that took in two quite different agricultural areas: (1) the main dairy counties of the eastern half of New York's Southern Tier and of the far northern counties of Pennsylvania, and (2) the counties of western Pennsylvania where prevailed an agricultural economy characterized by working off the farm. In the dairying area the low rates of farm depopulation in the 19505 was much associated with the significant tendency then for dairymen families to stay on their farms, to shun off farm employment, and to devote full time to the pursuance of a successful dairy Operation. The major exodus of dairy farmers and their families was to occur a decade later. A major contrast existed between the relatively low rates of pOpulation 305 .omm~-mea~ .mueocoaaou Hausuaaufium< eouaauomm< mo monoom :o .oomanommu .cowumasmom sham we moved cowucouom mo cowmmouuom scum maesvwmomnn.nn .uflm r .26 23 0.0. g ad 2 0.... Hm. a... 2 9?. D 0.»- 2 ad- S .8. 2a 0.9. a o 00. I... 00. 000.- . ...... O... C o a... o o o to .0 eon ccccc 0000.000 to 0. coo-Io. coo...- eeeooo-oeeoo 0.090... o 0.000 a. on a... e \ 00 o 00 o o o e e 0 00.0000. 0 0 0.0000 0 o o . o .00 O. O 0 O. o 0. O ‘ . coo. co... co ‘ OHM”. O O 01.0! 010. \ o e a 0 0. one...- 00 0.00000 0 I on o H e .m. munvw oeoouoeeoeoo one. no I 306 losses from the farms in most of the eastern portions of the northern Appalachians and the major losses in the agricultural population of western Pennsylvania (Figure 9) where part-time farming was much in vogue (Figure 34). Overall these two quite different farm population change areas during the 19505 were closely related to the relative importance of dairying. Some counties farm population changes were related to dairying, but not collectively to all the agricultural components chosen by multiple stepwise regression program (Figure 35). The counties of Centre, Clearfield, Clinton, and Juniata in central Pennsylvania and the counties of Lackawanna and Wayne in northeastern Pennsylvania had population changes primarily related to dairying or off the farm employment. Additional counties so explained were Beaver and Warren. Nevertheless, the map of the dairying or the off the farm work residuals contained basically the same areal patterns as did the map of multiple residuals. The mapped residuals of labor oriented farming revealed some clear areal patterns (Figure 36). Much of the Allegheny Plateau of western Pennsylvania, an area where extensive farming practices have had much use, had farm population declines that were much larger than of course could be explained by labor intensive types of farming that were areally limited. Central Pennsylvania with its special crops and higher labor needs contrastingly retained more of its agricultural population than would have been expected. The northeastern part of the Region appeared as an area with relatively low losses of farm people because of its strong orientation to dairying. Likewise, 307 .33 Jane". 30:... mmo name one: no con mad—Ho: 339890 Emu 3< mo acouuomuném .mE o\om.mm ounces $66 82:: 0\OO¢ cm>O 308 .oa-aeaH .oaaaa ado: gnaw ado-acaxaaaa mo mouoom co .oo-omm~ .cowuufisnom sham me «one: newucouom mo nowmmonmoa scum madsvwmom-.mn .mwm a .26 see 0.0. g ad 2 ad a a... 2 ad- U ad. 2 ad- s .8. 2a 0.9- E O . . - ... Q \ # .... . a \ no... 0 u e e n e e. o O 0 oo o o 0.00.... also...» 00.... O eon-...... o an... O. O ”’01.... O I I... O. 0 e... ... o O 309 .omaa-aaaa .ousuaaoaua< voucaauo gonad mo mouoom :o .oo-ommH .cowudaanom spam we mound noducouom mo scammouuom aowm mgasvwmom-.om .mfim r .26 use 0.0. g ad 2 06 mm ad 2 ad- U od- 2 ad- 3 .8. a... 0.9- a .. .-.\ a d Q . ‘ o. o e c \\ 310 several counties of northeastern Pennsylvania with a strong dairying orientation had low farm depOpulation. The relationship of the labor component to truck farming or specialty cr0p farming helped explain the population changes in some southern tier counties of New York State. The counties particularly affected were: Broome, Cattaraugus, and Steuben. The farm population changes of several counties in Pennsylvania were importantly related to this agricultural component which added to the multiple relationship in primarily the north- central, south-central, and the Pennsylvanian western boundary areas of the Region. Occasionally the combined components in the multiple regres- sion seemingly worked against each other for some counties. Although one component could be related to the farm population change when additional components were added a residual from multiple regression for some counties was produced which did not exemplify an important total explanation. Areas with high negative residuals, 1950-1960; some hypotheses. The counties with the lowest residuals from the multiple regression included several counties of north-central Pennsylvania and several scattered counties. Steuben County in New York State had the lowest score; however, this was presumably related to the importance of grain farming there as was that of north-central Pennsylvania in part, especially in Centre and Clinton counties. North-central Pennsylvania probably lost more of the farm population than predicted by the multiple regression because some subsistence farms and marginal farms had been retained longer due to the area's isolation than in 311 some other parts of the Region. Three of the remaining counties, Broome, Allegheny, and Carbon were under the influence of significant urbanization. Cattaraugus County with a moderately negative score had agricultural characteristics that associated much of the northern part of the County with the intensive dairying-grain growing agricul- ture of western New York and most of the southern part of the County with the more extensive type of agriculture of the Allegheny province of Pennsylvania. Dairying with much of the required grain grown on the farms as in western New York State was not especially associated to farm pOpulation changes as revealed by the mapped linear residuals of the dairying and off the farm work component (Figure 35). Like southern Cattaraugus County most of Warren County, although its farm population change was illustrated by the mapped linear residuals to be associated with the most important and first selected component, presumably had some agriculture closely related to the extensive and subsistence kind found in the north-central counties of Pennsylvania. Areas with high_positive residuals, 1950 to 1960; some hypotheses. Only a few counties had relatively high positive residuals. Some of these counties had a major type of agriculture rather unrepresented by the agricultural data included in the multiple regression program. Both Chautauqua and Schuyler counties in New York had an important grape and fruit production. Greene County in Pennsylvania raised sheep. Sullivan County was rather atypical of the Region's general agriculture because of the low level of economic and rural development (Gamble, 1967). As stated above some of the counties had population changes related to the dairying 312 and off-the-farm employment component when its effects were considered separately; thus, the counties with high multiple residuals, e.g., Lackawanna, Wayne, and Beaver, were accounted for. Lastly, a number of counties in the central portion of Pennsylvania, Mifflin, Snyder, Perry, and Northumberland, found within the Ridge and Valley physio- graphic province, had farm population changes unexplained by any statistical measure of agriculture. Probably, the agricultural activities and beliefs of the Amish and Mennonites in this sub- region of the State tended to keep the relatively high proportions of persons in agriculture on the farms during the period. Areal Relationships of Preceding Farm Population Changes to Agricultural Components The analysis of residuals discussed below sought in part to discover whether measurements of agriculture at the end of and following a period of farm population change could explain the rela- tive retention of persons in agriculture. The areas of explained agricultural population change, 1940 to 1950. During the years of the 19405, agricultural populations were greatly affected by World War II; nevertheless, there were as a con- sequence some secondary effects. A significant demand for labor attracted persons away from the farms and farm persons joined the armed forces, forcing an increase in the rate of capital input sub- stitution and mechanization. Labor intensive types of agriculture lost workers; whereas, extensive farming types of farming and dairying tended to retain persons on farms more than did agriculture in general. 313 Support for the above observations were given in the compari- son of the residuals produced from the multiple regression of the selected 1950 agricultural components and the county farm population changes (Figures 37 and 38). Except for the northwestern part of the northern Appalachians and the western half of the New York southern tier, counties with the least losses in farm population had residuals close to zero; and therefore, had the farm population changes most related to the agricultural components. These counties were primarily located in the major dairying areas of southeastern New York; north- eastern, central, and northwestern Pennsylvania. Dairying held agri- cultural p0pulation on the farms the best of any agricultural attribute and accounted for the majority of the farm population changes associ- ated with farming characteristics. The areas of unexplained agricultural population change, 1940 to 1950. The counties with high negative residuals had in common the highest rates of farm depopulation which were minimally associated with the chosen agricultural components (Figures 8 and 37). Most of the areas so depicted had physical conditions unfavorable to agricul- ture, i.e., associated with the Pocono Plateau and the Allegheny Highlands. In addition, several of the counties, particularly those of the Pittsburgh, Scranton, and Binghamton metropolitan areas were represented by high negative residuals. Only a few counties, Allegheny County and the counties of northeastern Pennsylvania had the highest rates of farm labor utilization; however, this agricultural character- istic could not explain an important proportion of the large farm 314 .omaH-aema .aucoaomaoo Hausuaaouum< udaoomm mo aouoom :o .omaH-oea~ .eowuaasmoa spam we nouns :owueouom mo :owmmonmom.aowvaHawvaaoM- an .aaa .26 .23 0.0. g ad 2 ad a a... 2 a.¢- U ad. 2 ad- a .8. as 0.2- a 315 .oan-momfl .aucocomaou HaH3uasuwum< vouawooaa< .cowpeaamom sham mo mound cofiueouom mo :owmaoumoa scum aaaavamo¢-.wn .mflm . m0 monoom :0 .chmH-oomu .26 .23 0.0. g ad 2 od mmm ad 2 ad- D od- 2 ad- s .8. as 0.9- a IIHIOIII .0 O. O. I II I II 0...... DIIIOII I III III IIIIIII I. O O I I .. o I .o II .... I ... I I I-um". ............ H c” ......... I II- uuuuuuuuuuuuuuuuuuuu I I .- eeeeeeeeeee I 0 IO. ........ IIIIIII II I III IIII II 0 II OIIIIOIII .00. CI IO. 0.0 .0. IIII IIooIIII .- I IIIIII IIII IIIIIII IIIIA OIIIOO IOIII III III II IIII IIII IIIII. I a . use. II In. .I II I . I. I O. O IOIOOI I” II 0 O O I .......... IIII ‘ IoII H”... II I . I I IIIII I I .0 O. ...... coco... ... u . III . ... II 0.... OIIIIII Int 0 O 0 ....I O ..I III I O O ... O C .0 I. I II. D III .0. O. I... I IIIIIII I III-II . II. II I00 I IIIOOI- II III I III cIIIII . coco...oIoII vcoon. 0...... I I III. I I I II IIII II I D I IbIb IIIIIIIIIIIIII I I IIII II ..‘I .‘.C U“‘ . I .- 0 O. IIIIIIIIII IIIIIII IIIIO II II IIIIII I II IIII I IIII I I. I II II .0 0. III I 316 population declines of those counties during the 19405, possibly because of the demands of industrial production during World War II. Some of the low agricultural depopulation rates of counties were associated with dairying according to the above areal analysis of the near zero residuals. Nearly all the remaining counties with low depopulation of farms had moderately high and high positive residuals. Southwestern, east-central and parts of northwestern Pennsylvania, and the western counties of New York's southern tier retained more persons in agriculture than predicted by the stepwise multiple regression, possibly because farming in these areas had large economic and pro- ductivity growth rates and possessed types of farming unrepresented by the agricultural components. Except for Forest, Elk, and Cameron counties in northwestern Pennsylvania the areas had numerous small cities and towns that provided an increasing local demand for farm products and supplemental employment. The areas of explained agricultural population change, 1960 to 1970. The decreases in the farm population within the northern Appalachians from 1960 to 1970 were associated with the overall nature of agriculture of the Region in 1970; the coefficient of multiple correlation (R) equaled .733. More attributes of farming were related to farm depopulation than during any other period analyzed in this study. Farm persons engaged in dairying and capital intensive agri- culture were less inclined to leave farming than farm individuals who worked off the farm, labored intensively in some types of farming, used their farms primarily as a residence, or adopted a type of agriculture emphasizing high unit production such as "corn—belt 317 agriculture." The adoption of types of farming with a high production emphasis may have at times overtaxed a farmers financial resources. The counties with moderate farm population declines generally coincided with the counties with near zero residuals. Areas of average agricultural population changes well explained were the dairy counties of northeastern and southcentral Pennsylvania and the eastern part of the southern tier of New York (Figures 10 and 38). A number of central Pennsylvania counties with strong measures of capital intensive agriculture but with only average dairy scores had moderate agricultural depopulation and residuals near zero. Apparently either the predominance of dairying or capital-intensive agriculture was associated with average farm population change rates (Figures 25, 26 and 10). A comparison of the linear residuals of the dairy off the farm work and the capital intensive agricultural residuals revealed that the moderate population losses of the eastern part of the northern Appalachians were associated with dairying; whereas, the average losses in counties of western Pennsylvania were related to both the dairying or off the farm work component and capital intensive agriculture. Counties with high population loss rates that associated with the composite of agricultural components were Allegany, Delaware, Tioga, and Tompkins in New York State; and Allegheny, Armstrong, Bedford, Cameron, Carbon, Elk, Fulton, Jefferson, Lycoming, Northumber- land, and Washington in Pennsylvania. Cameron, Carbon, Elk, Fulton, and Allegheny counties, often among the anomalies in studies of Pennsylvania, had unexpected explanations for farm depopulation. 318 The areas of unexplained agricultural population change, 1960 to 1970. There was a rather close correspondence between the counties with the largest farm depopulation rates and the locations of the lowest negative residuals (Figures 10 and 37). These counties had much higher losses of farm people than could be explained by the agricultural components selected by a stepwise multiple regression and were found in the eastern most part of the Region and in north- western Pennsylvania. Some of these counties had a high dependency on labor oriented agriculture which was not conducive to the retention of farm workers (Figure 27). The eastern fringe of the northern Appalachians had high farm population losses probably related to the spillover effects of Megalopolis. Some counties' farm population changes were explained by linear regressions when there had been no explanation given by the multiple regression. The western Pennsylvanian counties of Butler and Lawrence had farm population changes associated with dairying versus off the farm work component and capital intensive agriculture, respectively; while, Greene and Crawford agricultural changes were related to both agricultural components. Potter and Cortland's farm population retention rates were associated with dairying versus off the farm work component. The rates of Columbia, Mifflin, and Schoharie related to the capital intensive agriculture. Apparently in the multiple regression some of the other agricultural subsystems interacted with the farm population changes of these anomalous counties in a different manner than for the majority of the counties. The 319 whole of agriculture is really greater than the sum of the subsystems of agriculture. The counties with the best retention of agriculturally employed persons had most of the highest positive residuals. There existed reasons other than the selected agricultural components for the lowest rates of farm depopulation in these counties. Partly because plentiful cropland and good soils lend a basic advantage to farming operations, the counties of central Pennsylvania were somewhat more successful than most counties of the northern Appalachians in holding persons on farms. More persons could stay on farms through their employment in new local industries. Some of the counties were among the most economically viable in the Region. The Importance of Time Lag Effects of Agriculture From this exploratory study of the possible associations of agricultural systems, components, and characteristics with farm depopulation, some assessments were formed. The changing nature of the various aspects of agriculture in general have had a role in the relative stability and change of the population engaged in agricultural employment on the farms. The study of agricultural qualities as associated with farm population declines has given insufficient attention to the lag of the farm population responses to agricultural changes. Prior to the post-World War II era, the farm population had less inclination for moving than any other major group of Americans. Farm persons have had stronger ties to the land and family than probably any other group; therefore, they have traditionally been 320 more reluctant to leave farming than in general other employed persons have been to leave their work or residence. Many farm families post- poned their departures for many years. The man-land theme in geography was especially pertinent for studies of the agricultural populations; but, unfortunately, the peOple have been studied much less by geographers than the features of the rural landscape. The general tendency of a large proportion of the population on farms to respond to various causes and stimuli belatedly has had very limited and insufficient recognition. In an attempt to illustrate statistically the continued sig- nificance of the lag effects of changes in the agricultural components on the farm population, a stepwise multiple regression using the 1950 agricultural component scores with the change rates in the farm popu- lation was carried out. In explaining the farm population changes during the 19605 in the northern Appalachians, the 1970 agricultural component scores produced an R of .733 compared to an R of .630 with the 1950 agricultural component scores. The latter correlation indicated an important continued relationship between the nature of agriculture ten to twenty years prior to the losses of persons from the farms in the 19605. The past status of agriculture had nearly as much influence on the agricultural population changes in the 19605 as did the concurrent status of agriculture (Figures 38 and 39). One would expect if the recent past agricultural history has had such an important relationship to the contemporary farm population changes, then in past American history the importance of the lag effects on 321 .ommu-mvma .mueocomaou unusuasuwww< noumwooam< mo mouoom :o .onmu-oomfi .aowuafismoa spam we mouam :owucouom mo cowaaoumoa scum aaaspwaom-. .26 .26 0.0. E ad 2 od 5 ad 2 ad- U od- 2 ad- s .8. as 0.9- a w an .aaa (I Id I II I II I IIIIII I .0 O I C O. IIIII I I I I IIIIIIII III IIIII IIIIIII I IIIIII . . I IIIII IIIOII II IIIII I I I I I IIIII I I I I IIIIII I IOIOIIIOO III III I I I IHIIIIIQ1I O I C IIIIIIIII I I... I... IIII . O... I IIIIII I II II 322 later farm population changes were even greater, given the relatively rapid pace of change in modern economic societies. CHAPTER 9 SUMMARY, CONCLUSIONS, AND THE IMPLICATIONS FOR FUTURE RESEARCH Summary The Problem The fundamental task of this research project was to determine if there was in fact important relationships between the large decreases in a regional farm population of the United States, i.e., of the northern Appalachians, and the rapid changes occurring within the agricultural sector of the study area. Agriculture in the United States is a complex of interacting parts; and, although it may be perceived as one system, it is composed of several subsystems, most represented in varying degrees in each type of farming. It was generally hypothesized that agricultural technology has intersected these segments of agriculture differently and the impact has varied depending upon the existing natural and geographical setting. The subsystems of mechanization, capital investment, other off farm inputs, labor, land and production were derived from agricultural variables that seemingly would have been significantly altered by new technology. Similarities in the basic resources used in each type of farming however permitted considerable generalization on the changes 323 324 within agriculture. The major questions that guided the research and analysis for each time period, 1910 to 1930, 1950 to 1960, and 1960 to 1970 pertained to: (l) the patterns of farm depopulation (2) the general nature of agriculture (3) the importance of certain subsystems in the several areas (4) the relationships of subsystems as indicators of technological forces to farm depopulation. The Procedure of Analysis A choice of study area was selected that would represent a wide range of natural environmental conditions in addition to the manifestation of serious farm depopulation over several decades. These requirements were met by the chosen area of study, the northern Appalachian region of western and northern Pennsylvania and the southern edge of upstate New York. This area coincided with the northern portion of the Appalachian Region. The region's boundaries were delineated by the various Appalachian acts and amendments of the United States Congress. The regional research was aided by the writer's considerable familiarity with the chosen area. To represent and measure the magnitude and extent of the decreasing farm population problem, termed in this study as farm depopulation; ratios and/or proportions of farm populations of an earlier base year were calculated. The structure of the data avail- able from the several United States Censuses of Population influenced to some extent the selection of the time frame of the study. The farm population was first directly enumerated in 1920; however, 325 because of some basic calculations of average farm family size in 1910, estimates of the farm population for this year were made. Around this time the population of farms began a notable decline. The time periods of this research corresponded with the census decades from 1910 to 1970, except the 1930 decade and 1940 decade received little attention because of the staggering influence of "extraneous forces" of the Great Depression and World War II which disrupted the usual situation. Counties served as the geographical unit of analysis and data consisting in the form of percentages, ratios, and averages represented both the dependent and independent variables. Analysis at this geographical scale was thought preferable because of the wide availability of published data on the county level and the need to aggregate the data in such a manner that would easily convey a general assessment of the natural environment. The mapped farm depopulation or retention percentages for the periods of 1910-1930, 1940-1950, 1950-1960, and 1960-1970 produced some areal patterns that prompted both natural and agricultural hypotheses for the farm population losses. Thus, agricultural vari- ables were chosen that appeared to best reflect the overall tech- nological hypothesis. The selected variables primarily fitted the resource inputs and types of farming aspects of agriculture. Twenty-one agricultural variables (Appendices A and B) were submitted as inputs into a principal components analysis for each time period and the basic dimensions of the Region's agriculture interpreted from the major extracted components. The principal 326 components were defined by the way each agricultural variable loaded (correlated) with each component. The principal components met the assumption of independence and were then placed with the farm popu- lation retention ratios in a stepwise multiple regression program. The amount of association each agricultural component had with the farm population change and the amount of the total variation in the dependent variable each additional component accounted for were obtained. Finally, residuals of regression were mapped for certain com- ponents that had important statistical association with the farm population ratios in addition to the collective groups of components. The areal patterns gave clues for additional hypotheses and explanations. Findings and Results Throughout the time periods of study, parts of western and west central Pennsylvania had the highest farm depopulation rates. Another subarea that tended to have considerable relative farm depopulation was the Pocono Plateau of northeastern Pennsylvania. The ridge and valley physiographical province of Pennsylvania, especially in south central Pennsylvania, had the least farm depopu- lation. Parts of southern New York, particularly the southeastern dairying counties and adjacent eastern areas of Pennsylvania retained their farm populations somewhat more than the Region as a whole. The average retention or depopulation rates were generally the most accounted for statistically. The cropland availability com- ponent was associated positively with farm population retention in the 1910 to 1930 period. In the time periods of the 19605 and 19705, 327 dairying counties best sustained their population on the farms; whereas, off farm employment of farm operators led to higher rates of farm population losses, especially in western Pennsylvania. For the 1950 decade the proportion of the total variance accounted for by the agricultural components of 1949-50 equaled an R2 of .826. There was somewhat less statistical explanation given by 2 the 1969-70 agricultural components for the 19605; R equaled .733. Conclusions and Implications for Research Many minor findings resulted from this research. Both the findings not listed in this chapter and the major findings listed above lead to certain generalizations and conclusions. Technological forces had only a moderate influence on the farm population via agri- cultural changes prior to World War II; however, after this War, the technologically-based changes in agriculture have been much more associated with agricultural depopulation. The analysis of the Region's agriculture for the 19605 revealed that agriculture is now more heterogeneous in its characteristics; therefore, generalization about agriculture in the future will be more difficult. Areas dominated by primary economic activities had farm depopulation not well explained by this study. Urban areas had less farm depopulation relatively than expected which might suggest that conditions in the rural areas were more associated recently with farm population losses. Future studies of farm depopulation should be carried out in the field in those areas where explanation was not given by the technological thesis. BIBLIOGRAPHY BIBLIOGRAPHY Abler, Ronald, John S. Adams, and Peter Gould. Spatial organization; the geographer's view of the world. Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1971. Ackerman, Edward A. Geography as a fundamental research discipline. Department of Geography Research Paper No. 53. Chicago, Ill.: The University of Chicago, 1958. Alexander, John W. Economic geography. Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1963. Allaway, W. H. "Cropping systems and soil." Soil: The 1957 yearbook of agriculture. Washington, D.C.: U.S. Government Printing Office, 1957. Appalachian bibliography subjects. Morgantown, West Va.: Appalachian Regional Commission, 1967. Appalachian bibliography. Morgantown, West Va.: West Virginia Univer- sity Library, 1972. Baker, Oliver E. The increasing importance of the physical conditions in determining the utilization of land for agricultural and forest production in the United States. Annals of the Association of American Geographers 11:17-46, 1921. Baker, 0. E. Land utilization in the United States: Geographical aspects of the problem. The Geographical Review 13:1-26, 1923. Banks, Vera J., and Calvin L. Beale. Farm population estimates, 1910-1970. Statistical Bulletin No. 523. Washington, D.C.: Rural DevelOpment Service, United States Department of Agriculture, 1973. Banks, Vera J., Calvin L. Beale, and Gladys K. Bowles. Farm popu- lation estimates for 1910-62. Washington, D.C.: Economic Research Service Bulletin 103, United States Department of Agriculture, 1963. 328 329 Barger, Harold, and Hans H. Landsberg. American agriculture, 1899- 1939; a study of output, employment, and productivity. New York, N.Y.: National Bureau of Economic Research, Inc., 1942. Barnes, R. C., Jr., and B. D. Blakely. "Conservation in Modern Farming." Ip_A good life for more people: the 1971 yearbook of agriculture. Washington, D.C.: U.S. Government Printing Office, 1971. Bausman, J. Beaver county. The Knickerbocker Press, 1904. Beale, Calvin. "Farm population as a useful demographic concept." 12 Application of demography: the population situation in the U.S. in 1975. Chicago, 111.: Scripps Foundation and Population Research and Training Center, pp. 39-45, 1957. Beale, Calvin L. Rural depopulation in the United States: some demographic consequences of agricultural adjustments. Demography l(l):264-272, 1964. Beale, Calvin L. Natural decrease of population: the current and prospective status of an emergent American phenomena. Demography 6(2):91-99, 1969. Beale, Calvin L. The revival of population growth in nonmetropolitan America. Washington, D.C.: Economic Research Service Bulletin 605, United States Department of Agriculture, 1975. Benedict, Murray R. Farm policies of the United States, 1790-1950; a study of their origins and development. New York, N.Y.: The Twentieth Century Fund and American Book-Stratford Press, Inc., 1953. Bennett, Hugh Hammond. Soil conservation. New York, N.Y.: McGraw- Hill Book Co., 1939. Bennett, John W. Northern plainsmen; adaptive strategy and agrarian life. Chicago, Illinois: Aldine Publishing Company, 1969. Bercaw, Louise Oldham. Bibliography on land settlement, with particular reference to small holdings and subsistence home- steads. Washington, D.C.: U.S. Government Printing Office, 1934. Bertillon, Jacques. Le probléme de la depOpulation. Paris, France: A. Colin and Cie, 1897. Bertillon, Jacques. La dé pOpulation de la France. Paris, France: 1911. 330 Beshers, James M. Population processes in social systems. New York, N.Y.: The Free Press, 1967. Biays, Pierre. "Southern Quebec." Ip_Warkentin, John (ed.), Canada; a geographical interpretation. Toronto, Canada: Methuen Publications, pp. 281-333, 1968. Bishop, C. E., ed. Farm labor in the United States. New York, N.Y.: Columbia University Press, 1967. Bogue, Donald J. Principles of demography. New York, N.Y.: John Wiley and Sons, Inc., 1969. Bollinger, W. LaMar. "The economic and social impact of the depopu- lation process upon four selected counties in Idaho." 12. Mazie, Sara Mills (ed.), U.S. commission on pOpulation growth and the American future. Commission Research Reports, Vol. 5. Washington, D.C.: U.S. Government Printing Office, pp. 561-596, 1972. Borgstrom, Georg. Too many; a study of earth's biological limitations. New York, N.Y.: The Macmillan Co., 1969. Borgstrom, Georg. Focal points; a global food strategy. New York, N.Y.: Macmillan Publishing Co., Inc., 1973. Boulding, Kenneth E. A primer on social dynamics; history as dialectics and development. New York, N.Y.: The Free Press, 1970. Bowman, Isaiah. Geography in relation to the social sciences. New York, N.Y.: Charles Scribner's Sons, 1934. Bracey, Howard E. "Some aspects of rural depopulation in the United Kingdom." Rural Sociology 23:385-391, 1958. Bracey, Howard E. "Rural housing contrasts, a European view." 12_ A good life for more people: the 1971 yearbook of agriculture. Washington, D.C.: U.S. Government Printing Office, 1971. Bressler, G. 0. Labor saving on Pennsylvania poultry farms. Pennsylvania Agricultural Experiment Station Bulletin 532. 1950. Browder, Gordon. "Agricultural depopulation in eight Montana counties." Montana Business Quarterly l(3):31-46, 1963. Brunger, Eric. "A chapter in the growth of the New York State dairy industry, 1850-1900." New York History 36:136-145, 1955. 331 Bryn Greer-Wootten. A bibliography of statistical applications in geography. Commission on College Geography Technical Paper No. 9. Washington, D.C.: Association of American Geographers, 1972. Butz, Earl L. "Agribusiness in the machine age." Ip_Power to produce; the yearbook of agriculture, 1960. Washington, D.C.: U.S. Government Printing Office, 1960. Carey, George W. Systems, model building, and quantitative methods. Ip_Bacon, Phillip (ed.), Focus on geography; key concepts and teaching strategies. Washington, D.C.: National Council for the Social Studies, 1970. Carol, Hans. Stages of technology and their impact upon the physical environment: a basic problem in cultural geography. £2. Dohrs, Fred E. and Lawrence M. Sommers, Cultural geography: selected readings. New York, N.Y.: Thomas Y. Crowell Co., pp. 284-299, 1967. Carroll, William M. "Political and economic development of rural areas." Farm Economics, November, 1971. University Park, Pa.: The Pennsylvania State University Cooperative Extension Service. Carver, T. N. "Rural depopulation." Journal of Farm Economics 9(1): 1-10, 1927. Chisholm, Michael. Rural settlement and land use. London: Hutchinson University Library, Hutchinson 8 Co., Ltd., 1962. Chorley, Richard J. Geomorphology and general systems theory. Ip_ Dohrs, Fred E. and Lawrence M. Sommers, Introduction to geography: selected readings. New York, N.Y.: Thomas Y. Crowell Company, 1967. Chorley, Richard J., and Peter Haggett (eds.). Socio-economic models in geography. Parts I and III of Models in geography. London, England: Methuen, 1968. Clifton, Ivery D., and William D. Crowley, Jr. Farm real estate historical series data: 1850-1970. Washington, D.C.: Economic Research Service Bulletin 520. United States Department of Agriculture, 1973. Colby, C. C. Source book for the economic geography of North America. Chicago, 111.: University of Chicago Press, 1926. Commonwealth of Pennsylvania. Pennsylvania statistical abstract. Harrisburg, Pa.: Department of Internal Affairs, 1967 and 1975. 332 Conklin, Howard E., and Broder F. Lucas. An economic classification of rural land, Cattaraugus County, New York. Cornell Economic Land Classification Leaflet 3. Ithaca, New York: New York State College of Agriculture at Cornell University, June, 1952. Council of Europe. Problems caused by rural depopulation and rural revival in the balance between town and country. Vol. 1, Rural Depopulation and Vol. II, Rural Revival. Strasbourg, European Conference of Local Authorities. Seventh Session, October 28-31, 1968. Cowden, T. K., and E. G. Fouse. The supply and utilization of milk in Pennsylvania. University Park, Pa.: Pennsylvania Agri- cultural Experiment Station Bulletin 327, 1936. Cummins, David E. Resource use and returns for grade A dairy farms, 1968-69, based on studies in southeastern Wisconsin and central New York. Washington, D.C.: Economic Research Service Bulletin 466, United States Department of Agriculture, 1971. Cunningham, L. C. Commercial dairy farming in New York. Ithaca, N.Y.: Cooperative Extension Service, New York State College of Agriculture, Cornell Miscellaneous Bulletin 65, 1968. Dahlberg, R. E. "The concord grape industry of the Chautauqua-Erie area." Economic Geography 37:150-169, 1961. Davis, John H. "From agriculture to agribusiness." Harvard Business Review 34:107-115, 1956. Davis, Kingsley. "The theory of change and response in modern demo- graphic history.” POpulation Index 29(4):345-366, 1963. Davis, Kingsley. The urbanization of the human population. .EE Breese, Gerald, The city in newly developing countries. Englewood Cliffs, N.J.: Prentice—Hall, Inc., 1969. DeJong, Gordon F. Appalachian fertility decline; a demographic and sociological analysis. Lexington, Ky.: University of Kentucky Press, 1968. Dent, J. B., and J. R. Anderson. Systems analysis in agricultural management. Sydney, Australia: John Wiley and Sons Australasia PTY LTD., 1971. Dobyns, H. F. "Estimating aboriginal American population; an appraisal of techniques with a new hemispheric estimate." Current Anthropology 7:395-416, 1966. 333 Dorman, Eliot F. "Livestock Review." Ip_Pennsylvania craps and livestock annual summary 1966. Harrisburg, Pa.: Pennsylvania Crop Reporting Service Bulletin 36, Pennsylvania Department of Agriculture. Dorn, Harold F. "The natural decrease of population in certain American communities." Journal of the American Statistical Association 34:106-109, 1939. Duckham, A. N., and G. B. Masefield. Farming systems of the world. London, England: Chatto and Windus, 1970. Dumont, Arséne. Depopulation et civilsation, etude démographique. Paris, France: Lecrosnier et Babé, 1890. Durand, Loyal, Jr. "The American dairy region." Journal of Geography 48:1-20, 1949. Durost, Donald D., and Warren R. Bailey. "What's happened to farming." Ip_Contours of change: the 1970 yearbook of agriculture. Washington, D.C.: U.S. Government Printing Office, 1970. Ellis, David M., James A. Frost, Harold C. Syrett, and Harry F. Carman. A history of New York State. Ithaca, N.Y.: Cornell University Press, 1967. Eversley, Lord. "The decline of the numbers of agricultural labourers in Great Britain." Journal of the Royal Statistical Society 70:267-319, 1907. Feuer, Reeshon, W. L. Garman, and M. G. Cline. Chautauqua county soils. Soil Association Leaflet 3. Ithaca, N.Y.: New York State College of Agriculture at Cornell University, January, 1955. Field, N. C. Land hunger and the rural depopulation problem in the U.S.S.R. Annals of the Association of American Geographers, pp. 465-478, 1963. Fippin, Elmer 0. Rural New York. New York, N.Y.: The Macmillan Co., 1921. Firey, W. Man, mind, and land: a theory of resource use. Glencoe, 111.: The Free Press, 1960. Florin, J. The advance of frontier settlement in Pennsylvania 1638- 1850. Unpublished Master's Thesis. Pennsylvania State University, 1965. Foote, Don C., and Bryn Greer-Wootten. "An approach to systems analysis in cultural geography." The Professional Geographer 20(2):86-91, 1968. 334 Foote, Richard J. Concepts involved in defining and identifying farms. Economic Research Service Bulletin 448. Washington, D.C.: U.S. Department of Agriculture, 1970. Ford, Thomas R. (ed.). The southern Appalachian region; a survey. Lexington, Ky.: University of Kentucky Press, 1962. Frederick, R. H., E. C. Johnson, and H. A. MacDonald. Spring and fall temperatures in New York State. Cornell Miscellaneous Bulletin 33. Ithaca, N.Y.: New York State College of Agri- culture, 1959. Froelich, Patricia J. "The trend in milk consumption in the New York-New Jersey metropolitan area." Ithaca, N.Y.: Cornell University, Department of Agri. Economics. Farm Economics #196, 1954. Fuguitt, Glenn V. "Part-time farming and the push-pull hypothesis." The American Journal of Sociology 64(4):375-379, 1959. Fuguitt, Glenn V. "A typology of the part-time farmer." Rural Sociology 26:39-48, 1961. Fuller, Theodore E., and E. L. Baum. Employment, unemployment and low incomes in Appalachia. Agricultural Report 73. Washing- ton, D.C.: U.S. Department of Agriculture, 1965. Gade, Ole. "Geographic research and human spatial interaction theory: a review of pertinent studies in migration." 13_ Spencer, Robert F (ed.), Migration and anthropology; pro- ceedings of the 1970 annual spring meeting of the American Ethnological Society. Seattle, Wash.: University of Washing- ton Press, pp. 72-92, 1970. Gade, Ole. Spatial process and change in a depressed area; a study of migration in Nordland County, Norway. East Lansing, Mich.: Michigan State University. Unpublished Ph.D. dis- sertation, 1972. Galloway, Robert E. Part-time farming in eastern Kentucky. Lexington, Ky.: Kentucky Agricultural Experiment Station Bulletin 646, 1956. Gamble, Hays B. The economic structure of Sullivan County, Pennsylvania. Pennsylvania Agricultural Experiment Station Bulletin 743, 1967. Gans, Herbert J. More equality. New York, N.Y.: Pantheon Books, 1968. 335 Garlock, Fred L. "Financing capital requirements." Ip_Power to produce; the yearbook of agriculture, 1960. Washington, D.C.: U.S. Government Printing Office, pp. 375-379, 1960. Gates, Paul W. The farmer's age: agriculture, 1815-1860. Vol. III The Economic History of the United States. New York, N.Y.: Harper and Row Publishers, 1960. Gee, Wilson. "A qualitative study of rural depopulation in a single township: 1900-1930." American Journal of Sociology 39: 210-221, 1933. Gee, Wilson, and J. J. Corson. Rural depopulation in certain tide- water and piedmont areas of Virginia. New York, N.Y.: D. Appleton-Century Company, Inc., 1929. Goldthwait, James Walter. "A town that has gone downhill." The Geographical Review 17:527-552, 1927. Gottmann, Jean. Megalopolis, the urbanized northeastern seaboard of the United States. Norwood, Mass.: The Plimpton Press for The Twentieth Century Fund, 1961. Gottmann, Jean. Metalopolis, the urbanized northeastern seaboard of the United States. Paperback edition. Cambridge, Mass.: The M.I.T. Press, 1964. Gould, Peter R. Spatial diffusion. Resource Paper No. 4. Washing- ton, D.C.: Association of American Geographers, 1969. Graham, P. Anderson. The rural exodus; the problem of the village and the town. London, England: Methuen and Co., 1892. Great Britain Ministry of Housing and Local Government. Depopulation in Mid-Wales. H.M.S.O. London, England, 1964. Greeley, Roland B. ”Part-time farming and recreational land use in New England." Economic Geography 18:1146-1152, 1942. Griffiths, John C. "Unit regional-value concept a global strategy for development of natural resources." 13 Research in the College of Earth and Mineral Sciences, Circular 80. Univer- sity Park, Pa.: The Pennsylvania State University, pp. 13-14, 1970. Grigg, David B. The agricultural systems of the world; an evolu- tionary approach. London: Cambridge University Press, 1974. Haddon, A. C. The wonderings of peoples. Cambridge, England: Cambridge University Press, 1912. 336 Hampe, Edward C., Jr., and Merle Wittenberg. The lifeline of America: development of the food industry. New York, N.Y.: McGraw-Hill Book Company, 1964. Harris, D. A. The ecology of agricultural systems. 12_Cooke, Ronald U. and James H. Johnson, Trends in geography; an introductory survey. Oxford, England: Pergamon Press, 1969. Hart, John Fraser. "The three R's of rural northeastern United States." Canadian Geographer 7(1):13-22, 1963. Hartshorne, Richard. "Agricultural land in proportion to agricul- tural population in the United States." Geographical Review 29:488-492, 1939. Harvey, David. Explanation in geography. New York, N.Y.: St. Martin's Press, 1969. Hathaway, Dale E., and Brian E. Perkins. Occupational mobility and migration from agriculture. 13_Rural poverty in the United States. Report by the President's National Advisory Com- mission on Rural Poverty. Washington, D.C.: U.S. Government Printing Office, 1968. Hathaway, Dale E., and Arley D. Waldo. Multiple jobholding by farm operators. Research Bulletin 5. East Lansing, Michigan: Michigan State University Agricultural Experiment Station, 1964. Hathaway, Dale E., J. Allan Beegle, and W. Keith Bryant. People of rural America, 1960 census monography of the U.S. Bureau of the Census. Washington, D.C.: U.S. Government Printing Office, 1968. Heady, Earl 0., and Luther G. Tweeten. Resource demand and structure of the agricultural industry. Ames, Iowa: Iowa State Univer- sity Press, 1963. Heady, Earl 0., Howard C. Madsen, Kenneth J. Nicol, and Stanley H. Hargrove. "Agricultural and water policies and the environ- ment: an analysis of national resource use, food supply capacity and environmental quality." CARD Report 40 T, Rural Development Series. Ames, Iowa: Iowa State University Press, June, 1972. Helburn, Nicholas. "The gases for a classification of world agri- culture. The Professional Geographer 9(2):2-7, 1957. Hennefrund, Helen Emma. Part-time farming in the United States; a selected list of references. Agricultural Economics Bibliography #77. Washington, D.C.: Bureau of Agricultural Economics, United States Department of Agriculture, 1939. 337 Higbee, Edward. The squeeze; cities without space. New York: William Morrow and Company, 1960. Hirsch, Fredrick A. "Population peaks of United States counties." The Professional Geographer 22(2):88-91, 1970. Hoglund, A. William. "Abandoned farms and the new agriculture in New York State at the beginning of the Twentieth Century." New York History 34:185-203, 1953. Hollingsworth, Thomas H. Historical demography. Ithaca, N.Y.: Cornell University Press, 1969. Hood, Kenneth. "Part-time farming near industrial areas." Journal of Farm Economics 17:67-75, 1935. Hoover, Edgar M. An introduction to regional economics. New York, N.Y.: Alfred A. Knopf, Inc., 1971. Hunter, John N. "Ascertaining population carrying capacity under traditional systems of agriculture in developing countries." The Professional Geographer 18(3):lSl-154, 1966. Hunter, John N. "Population pressure in a part of the West African savanna: a study of Nangodi, northeast Ghana." Annals of the Association of American Geographers 57:101-114, 1967. Hutchinson, Bertram. Depopulation and rural life in Scotland; a summary report of three inquiries for the Department of Health for Scotland in part of rural Scotland as to the causes of rural depopulation. London, England: Central Office of Information, June, 1963. James, Preston E. All possible worlds: a history of geographical ideas. Indianapolis, Ind.: The Odyssey Press, A Division of the Bobbs-Merrill Co., Inc., 1972. Jaubert, P. Des causes de la depopulation et des moyens d'y Remédier. London, England, 1767. Jensen, Merrill (ed.). Regionalism is America. Madison and Milwaukee, Wis." The University of Wisconsin Press, 1951. (Paperback edition, 1965) John, M. E. Part-time farming in six industrial areas in Pennsylvania. State College, Pa.: Pennsylvania Agricultural Experiment Station Bulletin 361, 1938. Kasdan, Leonard. "Introduction." Ip_Spencer, Robert F. (ed.), Migration and anthropology; proceedings of the 1970 annual spring meeting of the American Ethnological Society. Seattle, Wash.: University of Washington Press, 1970. 338 Kauffman, Nelson M. "Pennsylvania weather summary." Ip_Pennsy1vania crops and livestock annual summary 1966. Harrisburg, Pa.: Pennsylvania Cr0p Reporting Service Bulletin 36, Pennsylvania Department of Agriculture. Keyfitz, Nathan. "L' exode rural dans la province de Quebec, 1951- 1961." Recherches Sociographiques 303-315, 1962. Kimmell, Donald C. Agricultural development in the Pittsburgh District, 1950-1945. University Park, Pa.: The Pennsylvania State University. Unpublished Ph.D. dissertation, 1950. King, L. J. Statistical analysis in geography. Englewood Cliffs, N.J.: Prentice Hall Publishing Co., 1964. Klein, Philip S., and Ari Hoogenboom. A history of Pennsylvania. New York, N.Y.: McGraw-Hill Book Co., 1973. Klimm, Lester E. "The empty areas of the northeastern United States." Geographical Review 44:325-345, 1954. Kostrowicki, Jerzy. Agricultural typology; summary of the activities of the ICU Commission for the years 1964-1968. Geographia Polonica 19:11-29, 1970. Langton, John. Potentialities and problems of adopting a system approach to the study of change in human geography. 12_ Progress in Geography 4:125-179, London: Edward Arnold (Publishers) Ltd., 1972. Lasley, Floyd A., and Charles N. Shaw. Economic aspects of dairying in the northeast. Agricultural Economic Report #188. Washington, D.C.: Economic Research Service, U.S. Department of Agriculture. Washington, D.C.: U.S. Government Printing Office, 1970. Lawton, Richard. Rural depopulation in 19th century England. In Steel, Robert W. and Richard Lawton (eds.), Liverpool éEans in geography; a jubilee collection. London, England: Longmans, Green and Co., Ltd., pp. 227-255, 1967. Leighly. Land and life. Berkeley, Calif.: University of California Press, 1967. Leuthold, F. D. "The rural exodus." Forest Farmer 28(2):6-7, 52-52, 1968. Linton, R. E., and H. E. Conklin. Economic viability of farmsareas in Tompkins County. Cornell Economic Farm Classification Leaflet l4. Ithaca, N.Y.: New York State College of Agri- culture and Life Sciences, 1972. 339 Longstaff, G. B. "Rural depopulation." Journal of the Royal Statis- tical Society 56:380-433, 1893. Lord, Russell. The care of the earth; a history of husbandry. New York, N.Y.: Thomas Nelson and Sons, 1962. Lowenthal, David and Lambros Comitas. "Emigration and depopulation: Some neglected aspects of population geography." Geographical Review 52:195-210, 1962. Lowenthal, David. "The return of the nonnative: new life for depopu- lated areas." 12_Kosinski, L. A. and J. W. Webb, Population and scale; micro-population. Edmonton: IGU Commission on Population Geography, 1975. MacDougall, John. Rural life in Canada; its trend and tasks. Toronto, Canada: Westminister Co., Ltd., 1913. MacDougall, John. Rural life in Canada; its trend and tasks. Toronto, Canada: University of Toronto Press, 1973. McDaniel, Robert, and Michael E. Eliot Hurst. A systems analytic approach to economic geography. Commission on College Geography Publication No. 8. Washington, D.C.: Association of American Geographers, 1968. McLin, Jon. Agriculture 1980: The Mansholt plan. American Univer- sities Field Staff Reports. West Europe Series Vol. IV, No. 6 (General), 1969. Mangalam, J. J. Human migration: a guide to migration literature in English, 1955-1962. Lexington, Ky.: University of Kentucky Press, 1968. Marsh, George P. Man and nature. Reprint of 1864 edition. Cambridge, Mass.: Belknap Press of Harvard University Press, 1965. Mathis, Anthony G. ”Dairying in the 19705." A reprint from Dairy Situation, DS-329, March, 1970. Washington, D.C.: Economic and Statistical Analysis Division, Economic Research Service. Metzger, Homer, and C. W. Pierce. Milk marketing by producer- distributors. University Park, Pa.: Pennsylvania Agricultural Experiment Station Bulletin 544, 1951. Miles, Rufus E., Jr. Awakening from the American dream; the social and political limits of growth. New York, N.Y.: Universe Books, 1976. Milk, Richard G. "The new agriculture in the United States: a dis- senter's view." Land Economics 48:228-239, 1972. 340 Miller, E. W. Strip mining and land utilization in Western Pennsylvania. Scientific Monthly 69:94-103, 1949. Miller, George J., and Almon E. Parkins. Geography of North America. New York, N.Y.: John Wiley and Sons, Inc., 1928. Miller, Reid S. "The Pennsylvania dairy industry." .12 Pennsylvania crops and livestock annual summary 1966. Harrisburg, Pa.: Pennsylvania Crop Reporting Service Bulletin 36, Pennsylvania Department of Agriculture. Morrill, Richard L., and Ernest H. Wohlenberg. The geography of poverty in the United States. New York, N.Y.: McGraw-Hill Book Co., 1971. Murarka, Dev. "How to keep 'em down on Soviet farm." The Christian Science Monitor, pp. 1, 4, March 14, 1975. Murphy, Raymond. Pennsylvania, a regional geography. Harrisburg, Pa.: The Telegraph Press, 1937. Nelson, Lowry. Rural sociology. New York, N.Y.: American Book Co., 1955. Newman, Monroe. A regional research program for Appalachia. Paper presented at the Sixty-Fourth Annual Meeting of the Associ- ation of American Geographers. Washington, D.C., August 19, 1968. Newman, Monroe. The political economy of Appalachia; a case study in regional integration. Lexington, Mass.: Lexington Books: D.C. Heath and Co., 1972. Nikolitch, Radoje. A comparison of age levels of farmers and other self-employed persons. Economic Research Service, Agricul- tural Economic Report No. 126. Washington, D.C.: U.S. Department of Agriculture, 1967. Nobe, K. C., E. E. Hardy, and H. E. Conklin. The extent and intensity of farming in Western New York State. Economic Land Classi- fication Leaflet 7. New York State College of Agriculture at Cornell University, June, 1961. Norling, Gunnar. "Abandonment of rural settlement in vasterbotten Lappmark, North Sweden." Geografiska Annaler 42:232-243, 1960. Ogle, William. "The alleged depopulation of the rural districts." Journal of the Rural Statistical Society 52:205-232, 1889. Olmstead, Clarence W. The phenomena, functioning units and systems of agriculture. Geographia Polonica 19:31-41, 1970. 341 Parodi, D. "A propos de la depopulation." Review de Metaphysique et de Morale 5(3):390-398, 1897. Patrick, Austin L. Soil erosion survey of Pennsylvania. Pennsylvania Agricultural Experiment Station Bulletin 354. State College, Pa.: Pennsylvania State College, 1938. Pearson, S. Vere. "Rural depopulation." lp_Growth and distribution of population. New York, N.Y.: John Wiley and Sons, Inc., pp. 209-228, 1935. Pennsylvania Department of Forestry. Statement of work done during 1901 and 1902. Wm. Stanley Ray, State Printer of Pennsylvania, 1902. Perret, Maurice E. New trends in the depopulation of upland areas in Switzerland. Abstracts of the 20th International Geographical Congress, 1960. Petersen, William. Population. New York, N.Y.: The Macmillan Co., 1961. Peterson, William. "Migration; social aspects." Ip_1nternational Encyclopedia of the Social Sciences 10:286-291. New York, N.Y.: The Macmillan Co. and The Free Press, 1968. Phelps, Harold A., and David Henderson. Population in its human aspects. New York, N.Y.: Appleton-Century Crofts, Inc., 1958. Pierce, Miriam 0. Bibliography of studies on the economic develop- ment of Pennsylvania and its regions (with supplements I through 1V). Center for Research of the College of Business Administration, University Park, Pa.: The Pennsylvania State University, 1962. Pillsbury, Richard. Images of Appalachia. The Geographical Bulletin 3:2-16, 1971. Pinchemel, Philippe. "Population." l§_Trollope, Christine and Arthur J. Hunt (translators), France: a geographical survey. London, England: C. Bell and Sons, Ltd., pp. 113-118, 1969. Pirtle, T. R. History of the dairy industry. Chicago, 111.: Mojonnier Bros. Co., 1926. Pressat, Roland. Population. Baltimore, Md.: Penguin Books, Inc., 1971. Rauchenstein, Emil, and F. P. Weaver. Types of farming in Pennsylvania. State College, Pa.: Agriculture and Experiment Station Bulletin 305, 1934. 342 Ravenstein, E. G. "The laws of migration." Journal of the Royal Statistical Society 48:167-235, 1885. Ravenstein, E. G. "The laws of migration." Journal of the Royal Statistical Society 52:241-305, 1889. Reinsel, Robert D., and Ronald D. Krenz. Capitalization of farm program benefits into land values. ERS-SO6. Washington, D.C.: U.S. Department of Agriculture, 1972. Reuss, Carl Frederick. "A qualitative study of depopulation in a remote rural district: 1900-1930. Rural Sociology 2:66-75, 1937. Rivers, W. H. R. (ed.). Essays on the depopulation of Melanesia. Cambridge, England: Cambridge University Press, 1922. Robson, B. T. Urban analysis. Cambridge, England: Cambridge University Press, 1969. Rogers, Everett M. Social change in rural society. New York, N.Y.: Appleton-Century-Crofts, Inc., 1960. Rogers, Everett M. Diffusion of innovations. New York, N.Y.: The Free Press, 1962. Rogers, Everett M., and Rabel J. Burdge. Social change in rural societies. Second edition. New York, N.Y.: Appleton- Century-Crofts, 1972. Rohrer, Wayne C., and Louis H. Douglas. The agrarian transition in America; dualism and change. Indianapolis, Ind.: The Bobbs-Merrill Co., Inc., 1969. Romsa, Gerald H. et al. "An example of the factor analytic-regression model in geographic research." The Professional Geographer 21:344-346, September, 1969. Rossi, Peter H. Why families move: a study in the social psychology of urban residential mobility. Glencoe, 111.: The Free Press, 1955. Rothblatt, Donald N. Regional Planning: The Appalachian experience. Lexington, Mass.: Heath Lexington Books, D.C. Heath and Co., 1971. Roxby, Percy M. "Rural depopulation in England during the nineteenth century." Nineteenth Century and After 71:174-190, 1912. 343 Royen, W. Van, and S. Moryadas. ”The economic geographic basis of Appalachia's problems." Tijdschrift Voor Econ. En Soc. Geografie 57:185-193, 1966. Rundblad, Bengt G. "Problems of a depopulated rural community." Lund Studies in GeOgraphy, Series B 13:184-191, 1957. Salter, Leonard A., Jr., and Larry F. Diehl. "Part-time farming research.” Journal of Farm Economics 22:581-600, 1940. Saville, John. Rural depopulation in England and Wales, 1851-1951. London, England: Routledge and Kegan Paul, 1957. Schultz, T. W. "Reflections on investment in man." Journal of Political Economics 70:1-8, Part 11, Supplement, October, 1962. Sestini, Aldo. "Regressive phases in the development of the cultural landscape." 1§_Wagner, Philip L. and Marvin W. Mikesell (eds.), Readings in cultural geography. Chicago, 111.: The University of Chicago Press, 1962. Shapp, Milton J. PRR + NYC spells economic ruin for Pennsylvania. Sworn testimony before the Interstate Commerce Commission. Philadelphia, Pennsylvania, 1963. Sharma, Prakash C. Migration; a selected international research bibliography. Exchange Bibliography 497. Monticello, 111.: Council of Planning Librarians, 1973. Shryock, Henry S., and Jacob S. Siegel and Associates. The methods and materials of demography. U.S. Department of Commerce Publication. Washington, D.C.: Bureau of the Census, 1973. Simkins, Paul D. "Distribution of the aged in Pennsylvania." Proceedings of the Pennsylvania Academy of Science 38(3): 1965. Simkins, Paul 0. "Some implications of Pennsylvania migration." Papers of the Earth and Mineral Sciences 39:52-53. Univer- sity Park, Pa.: The Pennsylvania State University, 1970. Slocum, Warren 0. An areal account of agricultural land abandonment in Pennsylvania, 1945-1959. University Park, Pa.: The Pennsylvania State University. Unpublished M.S. thesis, 1969. Sly, David F. Technology, environment, organization, and migration: an ecological approach to southern Negro migration. Providence, R.I.: Brown University. Unpublished Ph.D. dissertation, 1971. 344 Smith, C. T. "Depopulation of the central Andes in the 16th century." Current Anthropology ll(4-5):453-464, 1970. Smith, J. Russell. Men and resources; a study of North America and its place in world geography. New York, N.Y.: Harcourt, Brace and Co., 1937. Smith, Mervin G., and Carlton F. Christian (ed.). Adjustments in agriculture--a national basebook. Ames, Iowa: Iowa State University Press, 1961. Smith, T. Lynn. "DepOpulation of Louisiana's sugar bowl." Journal of Farm Economics 20:503-509, 1938. Spencer, J. E., and Norman R. Stewart. "The nature of agricultural systems." Annals of the Association of American Geographers 63(4)2529-544, 1973. Spengler, Joseph J. France faces depopulation. Durham, N.C.: Duke University Press, 1938. State of New York. Office of Planning Coordination. Economic viability of farm areas. New York State College of Agriculture, Cornell University, 1969. Stevens, Sylvester K. Pennsylvania--Birthplace of a nation. New York: Random House, 1964. Stevens, Sylvester K., Ralph W. Cordier, and Florence 0. Benjamin. Exploring Pennsylvania--its geography, history, and govern- ment. New York-Chicago: Harcourt, Brace and Company, 1953. Stockwell, Edward G. "Problems related to distribution trends." 13 Population and people. Chicago, 111.: Quadrangle Books, Inc., pp. 268-271, 1968. Stockwell, Edward G. "A methodological consideration for studying the consequences of population decline." Rural Sociology 34:552-555, 1969. Szabo, M. L. "Area-differential productivity of agricultural resources and off-farm migration of the Canadian Prairies." Geo- graphical Bulletin (Ottawa) 7, 4, 7, 2:113-l33, 1965. Szabo, Michael L. "Depopulation of farms in relation to the economic conditions of agriculture on the Canadian prairies." In Gentilcore, R. Louis (ed.), Geographical approaches to_-' Canadian problems; a selection of reading. Scarborough, Ontario: Prentice-Hall of Canada, Ltd., pp. 22-40, 1971. 34S Taeuber, Conrad. "Some current population trends." Talk at the 1972 National Agricultural Outlook Conference. Washington, D.C.: Agricultural Research Service, U.S. Department of Agriculture, 1972. Thoman, Richard 5., Edgar C. Conkling, and Maurice H. Yates. The geography of economic activity. Second Edition. New York, N.Y.: McGraw-Hill Book Co., 1968. Thomas, Edwin N. Maps of residuals from regression. 12 Berry, Brian J. L., and Duane F. Marble. Spatial analysis; a reader in statistical geography. Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1968. Thomlinson, Ralph. P0pulation dynamics; causes and consequences of world demographic change. New York, N.Y.: Random House, Inc. 1965. Thompson, John G. "Urbanization and rural depopulation in France." Journal of Farm Economics 7:145-151, 1925. Thompson, John H. (ed.). Geography of New York State. Syracuse, N.Y.: Syracuse University Press, 1966. Toniolo, A. R. "Studies of depopulation in the mountains of Italy." Geographical Review 27:473-477, 1937. Truesdell, Leon E. Farm population of the United States. Census Monograph 6. Washington, D.C.: U.S. Government Printing Office, 1926. Truesdell, Leon E. Bureau of the Census, Farm P0pulation--1880-1950 with special attention to its relation to number of farms and number of agricultural workers. Technical Paper 3. 1960. Tyson, William E. "Toward a development policy for the Appalachian region of New York State." Paper presented at the annual meeting of the A.A.G. Washington, D.C., August, 1968. Tweeten, Luther G. "The income structure of (United States) farms by economic class.” Journal of Farm Economics 47:207-221, 1965. U.N. Department of Economic and Social Affairs. The determinants and consequences of population trends: new summary of findings on interaction of demographic, economic, and social factors. Vol. 1."New York: U.N. Publications, 1973. U.S. Bureau of Census. County and City Data Book, 1972. 346 U.S. Department of Agriculture. Economic and social problems and conditions of the southern Appalachians. Misc. Publication 205. Bureau of Agricultural Economics, Bureau of Home Economics and Forest Service. Washington, D.C.: U.S. Govern- ment Printing Office, 1935. U.S. Department of Agriculture. 1973 Handbook of agricultural charts. Agricultural Handbook No. 455. Washington, D.C.: U.S. Government Printing Office, 1973. U.S. President Appalachian Regional Commission. Appalachia. U.S. Government Printing Office, 1964. Vaughan, Lawrence M. Abandoned farm acres in New York. Ithaca, N.Y.: Agricultural Experiment Station Bulletin 490, 1929. Wagner, Philip L., and Marvin W. Mikesell (ed.). Readings in cultural geography. Chicago, 111.: The University of Chicago Press, 1962. Walmsley, D. J. Systems theory: a framework for human geographical enquiry. Research School of Pacific Studies, Department of Geography Publication HG/7 (1972). Canberra, Australia: Australian National University, 1972. Walters, A. Harry. Ecology, food and civilization. London, England: Charles Knight 6 Co., Ltd., 1973. Warren, William D. "Appalachian railroad towns: a class of declining urban places." Paper presented at The East Lakes Divisional Meeting of the A.A.G. Indiana, Pa.: October 7, 1972. Watson, J. W. "Rural depopulation in southwestern Ontario." Annals of the Association of American Geographers 37:145-154, 1947. West, Quentin M. The changing missions of agricultural economics research. Washington, D.C.: Division of Information. Economic Research Service, U.S. Department of Agriculture, 1973. White, Leslie A. The science of culture; a study of man and civili- zation. New York, N.Y.: Grove Press, Inc., 1949. Wilcox, Walter W. The farmer in the second world war. Ames, Iowa: The Iowa State College Press, 1947. Wilbanks, Thomas J., and Richard Symanski. "What is systems analysis?” The Professional Geographer 20(2):81-85, 1968. Wilson, Harold F. The hill country of northern New England. New York, N.Y.: Columbia University Press, 1936. 347 Wittick, Robert 1. Some general statistics programs used in spatial analysis. Technical Report 7l-l. East Lansing, MI: Michigan State University, Computer Institute for Social Science Research, 1971. Woodham-Smith, Cecil. The great hunger. New York, N.Y.: The New American Library of World Literature, Inc., 1964. Wrigley, P. 1. Types of farming in Pennsylvania. Pennsylvania Experiment Station Bulletin 479, 1946. Young, W. R. Conscription, rural depopulation, and the farmers of Ontario, 1917-19. Canadian Historical Review. 53. 1972. Zelinsky, Wilbur. "Changes in the geographic patterns of rural population in the United States, 1790-1960." Geographical Review 52:492-524, 1962. Zelinsky, Wilbur. A prologue to population geography. Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1966. Zelinsky, Wilbur. Nonmetropolitan Pennsylvania: a demographic revolution in the making? University Park, Pennsylvania: The Pennsylvania State University, College of Earth and Mineral Sciences, 45:1, 1975. APPENDIX A VARIABLES INCLUDED IN THE RESEARCH APPENDIX A VARIABLES INCLUDED IN THE RESEARCH Agricultural and Population Variables in the Correlation Matrices* Independent Variables for 1910 Average value of land and buildings per farm acre Average value of implements and machinery per farm acre Average expenditure for feed per farm reporting Average expenditure for fertilizer per farm reporting Farm persons per 100 acres Value of hired labor per farm reporting Average size farm in acres Average cropland harvested acreage 9. Percentage harvested cropland, hay 10. Corn yields in bushels per acre 11. Land productivity: Agricultural sales per acre 12. Labor productivity: Agricultural sales per agricultural person 13. Ratio farm expenses to value of agricultural products oowomc-MNH Independent Variables for 1930, 1950, 1970 1. Percentage of farms reporting tractors 2. Percentage of farms reporting automobiles 3. Percentage of farms reporting trucks 4. Average dollars spent per farm for gas 5. Average value of land and buildings per farm acre 6. Average value of implements and machinery per farm acre 7. Average expenditure for feed per farm reporting 8. Average expenditure for fertilizer per farm reporting 9. Farm persons per 100 acres 10. Value of hired labor per farm reporting 11. Average size of farm in acres 12. Average cropland acreage per farm reporting *Unless otherwise indicated the raw data represents the census year or the previous year. 348 13. 14. 15. 16. 17. 18. 19. 20. 21. 349 Percentage cropland harvested Percentage harvested cropland, hay Pastureland-cropland ratio Percentage of total farms that are dairy Corn yields in bushels per acre Land productivity: Agricultural sales per acre Labor productivity: Agricultural sales per agricultural person Ratio of farm expenditures to value of agricultural products Percentage farm Operators working 100 days off the farm APPENDIX B SOURCES OF DATA Variable Number 1 1- 2. 3- 4- 7- 9- 1o, 11, 12, 13, 15, 17, 18, 19, 20, 21 1' 2) 3’ 4! 6! 7D 8, 9, 10, 11, 12, 13, 14, 15, 17, 18, 19, 20 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20* a b 4, 21 5, 6, 7, 8, 9, 1o, 11, 12,* 14, 17, 18,* 19,* 20* S,* 21 APPENDIX B SOURCES OF DATA Source U.S. Bureau of the Census. U.S. Census of Agriculture: 1925. Part 1 (The Northern States). U.S. Bureau of the Census. U.S. Census of Agriculture: 1950. Vol. 1. Counties and State Economic Areas. Part 2 (Middle Atlantic States). U.S. Bureau of the Census. U.S. Census of Agriculture: 1969. Vol. 1. Area Reports. Part 7 (New York) and Part 9 (Pennsylvania). U.S. Bureau of the Census. U.S. Census: 1930. Agriculture Vol. 2. Part 1. The Northern States. U.S. Bureau of the Census. U.S. Census: 1940. Agriculture Vol. 1. Part 1. Statistics for Counties. U.S. Bureau of the Census. U.S. Census: 1910. Vol. 7. Agriculture Reports by States with Statistics by Counties U.S. Bureau of the Census. County and City Data Book. 1972 and 1973. U.S. Bureau of the Census. U.S. Census of Agriculture: 1945. Vol. 1. Part 2 (Middle Atlantic States). 350 Variable Number 8,d 16 16 18, 19 Notes: 351 Source U.S. Bureau of the Census. U.S. Census of Agriculture: 1954. Vol. 1. Counties and State Economic Areas. Part 2 (Middle Atlantic States). U.S. Bureau of the Census. U.S. Census of Agriculture: 1964. Statistics for the States and Counties. U.S. Bureau Of the Census. U.S. Census: 1930. Agriculture. Vol. 3. Type of Farm. Part 1 (The Northern States). Census volumes are published by the United States Government Printing Office. The dependent variable information is available in Chapter 111. Appendix B needs to be used with Appendix A. Although numbered sequentially each variable for 1910 is given the same number as for 1930, 1950, and 1970. Most variables are expressed in relative measures calculated from absolutes by the author. 31940 data. bTotal operators working off farms. C194s data. d1954 data. 81964 data. *Special tabulations and aggregative calculations. 7. APPENDIX C CORRELATION COEFFICIENTS MATRIX (1909-1910) 0000.. 00m..1 0000.1 ..0..1 000.. mvhm.- m0v0. 5.00. .mv.. mvmm. whee. n00.. 000.. 0. 00m..- 0000.. 0.vs. ommv. cumm- vvve. 000m. 000m. vmvv.u 0000.1 mums. wao. mmm0.- N. 0000.1 0.vn. 0000.. nmom. ween. vvvn. oNNN.n m.wv. 0.0.. 000.. 0.00. w0vm. mv0n. .. ..0..- ommv. mmon. 0000.. v~0v. vo...- 5000. hm.N. nvm..u mm0v.- ¢~.m. 0m0..- vmmm. 0. 000.. 0500. movm. v~0v. 0000.. when-1 NMNN. mmon. 0.vm.u wmvm.- >000. oovm.u .vo..- 0 nvnm.- cvvv. vvvm. #0...- mnom.1 0000.. 0mm.. 0N0..- wmo~.- no.0. n0...- 0mmm. .00..- m n0v0. mwom. ONNN.- mono. NmNN. 0m~.. 0000.. 0.~0. 05.0.1 00mm.- 00.~. .mmo.- vmnv.- n 5.00. 0m0n. n.wv. nm.~. mmon. 0mm..- 0.N0. 0000.. enmo. mnvv. ~m0o. nvvv. 00.0. o .mv.. vnvv.- 0.n.. nQ~..u 0.vn.u mne~.- 05.0.1 QmN0. 0000.. 00.v. 000..- @000. mmnm. m nvwm. mm00.- 000.. mmov.- mmvm.- 00.0. 00mm.- nave. 00.v. 0000.. mvm.. .nvo. wmnm. v mnov. mums. v.00. v~.m. n0vo. 50...- 00.N. Nm0o. 0m0..- mvm.. 0000.. vom.. .v.~. m moo.. wao. w0vm. 0m0..- oov~.- 0mmm. .wmo.- nvvv. coco. .nvo. vom.. 0000.. n.m0. N 000.. mm~0.- wv0m. vmmm. .vo..- .00..- vmnv.- 0m.m. mmnm. nmnm. .v.~. m.mo. 0000.. . n. N. .. 0. 0 w n o m v m N . .0.0.-000.. x.mh