mousmmuzmou 0F RURAL . . ’ AREAS: MICHIGAN Thesis for the. Degree of M. S. MICHIGAN STATE UNWERSITY RlCHARD JOSEPH NANNEMAN 1.968 {THESIS Nanneman, Richard J Industrialization of rural areas: Michigan. M.S. Thesis 1968 i c:.! i MSU LIBRARIES .-:,‘—. RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. ABSTRACT INDUSTRIALIZATION OF RURAL AREAS: MICHIGAN by Richard Joseph Nanneman In recent years there has been widespread concern about declining employment opportunities in many rural areas. National, state, county, and local groups have been actively engaged in 'promoting' industrial develop- ment for disadvantaged areas. The purpose of the study was to investigate (1) factors of location which have been important to firms locating in rural areas, (2) on the basis of experienced employment growth, which industries are the most likely to be attracted to rural areas in Michigan, and (3) the relationships between comparative employment growth and levels of public and private services in a rural area in Michigan. Recent literature dealing with surveys on factors influencing the location of industry in rural areas was reviewed for manager rankings of locational factors. Markets, raw materials, labor, and transportation were of primary importance for location decisions. Richard J. Nanneman Utilities, local taxes and legislation, and community atti- tudes and inducements to firms were of much less importance. There was evidence that communities attached too much importance to the role that inducements play in at- tracting firms. Community inducements often representing a substantial cost were often accepted by firms but disre- garded while making the location decision. The shift-share technique, based on census employ- ment data, was used to identify those export industries for which employment shifted comparatively into rural areas ‘in Michigan from 1950 to 1960. The manufacturing industries electrical and other machinery, transportation other than motor vehicles, and food and kindred products, had the greatest relative shifts into rural counties in upper Michigan. Miscellaneous man- ufacturing and chemicals and allied products had large re- lative shifts into lower Michigan counties which were not part of a standard metr0politan statistical area. These industries deserve to be studied first to see if the trend can be continued. The statistical methods Of regression and correla- tion were used to test whether some of the differences in comparative growth rates, 1950 to 1960, among counties in a rural area in Michigan can be explained by differences in the level of public and private services available. The dependent variables consisting of share growth rates Richard J. Nanneman of several industry aggregations were generated with the shift-share technique. Share growth rates were used to measure how well a particular county did in an industry (or group of industries) relative to how well all counties in the United States did in this industry (or group of industries). The six explanatory variables used were (1) per- capita government expenditures for 1957, (2) per-capita education expenditures for 1957, (3) education completed, (4) number of transportation facilities, (5) percent of county population rural farm, and (6) size of largest city. The six explanatory variables accounted for less than forty percent of the total variation in share growth . rate of all industries, about fifty-nine percent of the variation in manufacturing share growth rate, and sixty- five percent of the variation in electrical and other ma- chinery manufacturing share growth rate. Per-capita government expenditures was the most 'important variable in explaining employment share growth rate of manufacturing industries. But it was relatively unimportant for share growth rate of all industries. The variable, percent of population rural farm, was the second most important in accounting for differences in share growth rates of both all industries and manufac- turing industries. However, the relationship was negative for all industries and positive for manufacturing industries. Richard J. Nanneman A very strong negative relationship was found be-_ tween number of transportation facilities and share growth rate of all industries. The relationships were not all as expected. Re- sults indicated that some public and private services are industry specific and the relationships can be understood only in terms of the process of change in this area. INDUSTRIALIZATION OF RURAL AREAS: MICHIGAN by Richard Joseph Nanneman A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Agricultural Economics 1968 ACKNOWLEDGMENTS The author wishes to express his sincere apprecia- tion to all those who assisted him in the preparation of this thesis. The author is eSpecially indebted to Dr. A. Allan Schmid who served as major professor and guided the research. Thanks are also due to Drs. Milton H. Steinmueller and Larry Connor who served on the examining committee and helped give direction to the study. The author also wants to express his appreciation to Dr. L. L. Boger and the Department of Agricultural Eco- nomics who provided the finances to make this study possible. To my wife, Kathy, a very special note of thanks is due. Without her support, encouragement, and under- standing, none of this would have been possible. ii TABLE OF CONTENTS ACKNOWLEDGMENTS . . . .4. . . . . . . . . LIST OF TABLES . . . . . . . . . . . . . LIST OF ILLUSTRATIONS . . . . . . . . . . LIST OF APPENDICES O O O O O O O O C C 0 Chapter I. II. III. THE STUDY 0 O O O O O O O O O O 0 Introduction . . . . . . . . . Objectives . . . . . . . . . . Procedures . . . . . . . . . . PROFILE OF MICHIGAN . . . . . . Introduction . . . . . . . . . State Population, Income, and Employment . . . . . . . . . . Northern Michigan and Southern Michigan . . . . . . . . . . . LOCATION FACTORS . . . . . . . . Introduction . . . . . . . . Major Location Factors . . . Analysis of Basic Factors . . Analysis of Secondary Factors Limitations of Surveys . . . Analysis of Location Factors by Industry . . . . . . . . . . . Recent Growth and Present Population Characteristics as Related to Growth Potential 0 O 0 0 O O O I O O 0 Summary . . . . . . . . . . . . iii Page ii vii viii U1U1l-‘ H 00 14 19 19 20 23 31 39 41 47 50 Chapter IV. V. VI. SHIFTS IN INDUSTRIAL EMPLOYMENT . . . Introduction . . . . . . . . . . . Employment and Components of Em- ployment Change in Michigan, 1950 to 1960 . . . . . . . . . . . . . . Employment Redistribution for Non- Standard MetrOpolitan Statistical Areas, 1958-1963 . . . . . . . . . Dispersion of Additional Employment Summary . . . . . . . . . . . . . . IMPACT OF PUBLIC AND PRIVATE SERVICES ON EMPLOYMENT GROWTH O O O O O O O 0 Introduction . . . . . . . . . . . Analysis of Growth as Related to Public and Private Services . . . . Summary and Conclusions . . . . . . SUMMARY AND CONCLUSIONS . . . . . . . APPENDIX A O O O O O O O O O O O O O O I O 0 APPENDIX B O O 0 O O O O O O O O O O O O O O BIBLIOGRAPHY . . . . . . . . . . . . . . . . iv Page 52 52 57 65 68 70 71 71 74 9O 95 107 111 114 Table LIST OF TABLES POpulation of Michigan and the United States, 1900, 1940, 1950, and 1960 . . Per-capita personal income in Michigan and the United States, 1940, 1950, and 1960 O O I O I O O O O O O O O O 0 Employment and percent change in employ- ment, U. S. and Michigan, agriculture and non-agriculture, 1940 to 1950 and 1950 to 1960 . . . . . . . . . . . . . Population of Michigan, Northern and Southern Michigan, 1950 and 1960, total change, percent rural, and net migration . . . . . . . . . . . . EmplOyment in Northern and Southern Michigan, 1950 and 1960, employment change, and employment as a percent of total . . . . . . . . . . . . . . . Plant managers' rankings of location factors (in terms of importance), all manufacturing groups, South Central Louisiana . . . . . . . . . . Plant managers' evaluation of location factors, all industry groups, South Central Louisiana . . . . . . . . . . New plant managers' rankings of nine location factors, all industries, southern Indiana 0 O O O O O O O O O 0 Location factor priority rankings, all industries, Southern Indiana . . . . . Page 10 11 15 16 24 25 27 27 Table ' Page 10. Employment and components of employment change in area 1, 1950-1960, manufac- turing and other commodity producing industries . . . . . . . . . . . . . . . . 58 11. Employment and components of employment change in area 2, 1950-1960, manufac- turing and other commodity producing industries . . . . . . . . . . . . . . . . 62 12. Regional share component, 1950 to 1960, agriculture, electrical & other mach- inery, and chemicals & allied products manufacturing, areas 1, 2, and 3 . . . . . 64 13. Employment in Michigan and in non-standard metropolitan statistical areas, 1958-1963, coefficient of redistribution for non- SOMOSOA. I I O O O O I O O O O O O O O O O 67 14. Number of counties gaining and losing em- ployment by industry, area 1, 1950 to 1960 O O O O O O O O O O O O O O O O O O O 69 15. Share growth rates, all industries, manu- facturing industries, and electrical & other machinery manufacturing, 1950-1960, in Northern Lower Peninsula . . . . . . . 77 16. Regression and correlation coefficients for all industry share growth rate, Northern Lower Peninsula . . . . . . . . . 82 17. Regression and correlation coefficients for manufacturing industries share 'growth rate, Northern Lower Peninsula . . 85 18. Regression and correlation coefficients for electrical & other machinery manu- facturing share growth rate, Northern Lower Peninsula . . . . . . . . . . . . . 87 vi LIST OF ILLUSTRATIONS Figure Page 1. Major cities of standard metropolitan statistical areas in Michigan . . . . . . . l3 2. Areas used for shift-share analysis . . . . . 56 vii LIST OF APPENDICES Page A. Computation of Shift-Share Analysis . . . . . . 107 B. Employment and Components of Employment . Change, 1950-1960, Areas 1, 2, and 3 . . . . lll viii CHAPTER I THE STUDY Introduction In a dynamic and growing economy such as exists in the United States, economic and technological conditions are constantly changing. As shifts occur in economic ac- tivity, some areas become disadvantaged and fall outside the mainstream of activity and growth. Pockets of unem- ployment, underemployment, low incomes, and outmigration develop. In most such areas, the problems are long-term ' and stem from an inability to compete economically. As a result, these areas have drifted into the backwaters of a growing national economy. In recent years, much attention has been focused on the problems of disadvantaged areas. National, state, county, and local groups have formed in a concerted effort to restore such areas to prosperity. Many committees have devoted attention to the problems of obtaining new industry and job Opportunities. On the national level, designation of the Presi- dentfls Appalachian Regional Commission in 1963 was a major cknnonstration of the recognition of depressed conditions in some areas. This commission studied the situation in Appalachia and made recommendations which led to the Appa- lachian Regional DevelOpment Act of 1965. This Act provided for a variety of programs design- ed to stimulate the economic development of the Appalachian Region. Under the Act, public investments are concentrated on the development of physical and transportation facilities and human resources that are expected to generate diversi- fied industry. Subsequently, economic develOpment regions have been designated in such regions as New England, the Ozarks, and in the Northern Great Lakes States, through the assistance of the newly established Office of Regional Economic DevelOpment (ORED) in the Commerce Department. The Public Works and Economic DevelOpment Act of 1965 (established the Economic DevelOpment Administration) provided for help in the economic development of distressed areas. This program required COOperation between public ’and private enterprise and officials at every level of government--local, county, state, and federal.) Loans and grants were made available to designated areas for public works, public services or public facilities related to economic develOpment. To encourage private investment, low-interest, long term loans were made for thepurchase or development of land and facilities for industrial usage. State legislatures, likewise, have been promoting new industries. In 1963 the Michigan Industrial DevelOpment Revenue Bond Act was signed by the Governor. This Act authorizes municipalities to issue revenue bonds, the funds from which are used to acquire industrial facilities for the purpose of leasing them to industrial firms. Perhaps the groups most deeply concerned about em- ployment opportunities have been the local communities. Many communities have formed corporations for the sole purpose of attracting new industry. In 1966, there were close to 180 community industrial develOpment corporations in Michigan. Eighty—three percent of these were non-profit corporations. The President's National Advisory Commission on Rural Poverty has aptly pointed out the consequences of the lack of employment Opportunities and has recommended efforts to increase jobs in rural areas. In the summary of its report to the President, the commission states: Technological progress brought sharp declines in the manpower needs of agriculture, forestry, fisheries, and mining. Other industries have not replaced the jobs lost, and they have supplied too few jobs for the young entries in the labor market. Larger towns and cities have taken over many of the economic and social functions of the villages and small towns. As th communities ra downhill, they offered fewer and fewer Opportunities for anyone to earn a living. The inadequately equipped young people left in search of better Opportunities elsewhere. Those remaining behind have few resources with which to earn incomes adequate for a decent living and for revitalizing their communities. Congress and state legislatures from time to time have enacted many laws and appropriated large sums of money to aid the poverty stricken and to help rural America. Very little of the legislation or money has helped the rural poor. This commission questions the wisdom of massive public efforts to improve the lot of the poor in our central cities without comparable efforts to meet the needs of the poor in rural America. Un- fortunately, as public programs improve the lot of the urban poor, without making similar improve- ments in conditions for the rural poor, they pro- vide fresh incentive for the rural poor to migrate to the central cities. The Commission has endeavored to chart a course to wipe out rural poverty. . . . Improving the op- eration of the private economy in order to provide rural people with better Opportunities for jobs and a decent living is emphasized. The Commission presented recommendations of pol- icies and programs to assist the rural poor, many of which dealt with employment Opportunities. The Commission recommends that the United States adOpt and put into effect immediately a national policy designed to give the resident of rural America equality of opportunity with all other citizens. This must include equal access to jobs. The Commission recommends a thorough overhaul- ing of our manpower policies and programs . . . to deal effectively with rural unemployment and under- employment. To finance develOpment, the Commission recom- mends Federal grants, loans, and industrial devel- Opment subsidies as well as state and local tax reform. Clearly, there is a need to examine the factors and relationships relevant to industrialization of rural 1National Advisory Commission on Rural Poverty, The People Left Behind. A Report by the President's Na- tional Advisory Commission on Rural Poverty (Washington: U. S. Government Printing Office, 1967), pp. x-xi. 2 Ibido ' Xi-Xii. areas. Some of the accepted beliefs on how to procure industrial employment opportunities in rural areas need to be investigated, and, if possible, new leverage points need to be identified. Objectives The purpose of this study is to investigate some of the aspects of procurement of employment Opportunities in rural areas. Identification of the factors important to firm location decisions can furnish persons interested in rural develOpment with increased knowledge about the role local groups might play in obtaining additional in- dustry. Shifts in industrial employment can provide clues as to which industries are most likely to be attracted to rural areas. Specifically, the study will examine: (l) the fac- tors which have been important to firms locating in rural areas, (2) industrial employment shifts in Michigan to de- termine which industries have shown a tendency to locate in rural areas in Michigan, and (3) relationships between the level of public and private services available in rural counties and comparative employment growth rates. Procedures A profile of Michigan is presented in Chapter II. Michigan's pepulation, income, and employment is compared with that of the United States. Also, the highly indus- trialized and heavily populated southern part of Michigan is compared with the northern counties of Michigan. In Chapter III, surveys on locational factors of industry in rural areas are examined. Factors, which come munities are able to manage to some degree or which can be proffered as inducements, are investigated for relative importance in location decisions. The shift-share technique is used in Chapter IV to determine industry shifts between rural and urban areas within Michigan from 1950 to 1960. Manufacturing indus- tries are then analyzed in terms of potential contribution to employment growth in rural areas; based on exhibited employment shifts. The counties in Michigan are aggregated into three groups; (1) those counties in upper Michigan (i.e. the Upper Peninsula and the northern part of the Lower Peninsula) containing no major urban-industrial cen- ters, (2) counties in lower Michigan surrounding the Stand- ard MetrOpolitan Statistical Areas, and (3) counties desig- nated as Standard MetrOpolitan Statistical Areas. The statistical procedures of regression and cor- relation analysis are used in Chapter V to investigate the association between levels of public and private services available in rural counties and relative employment growth rates (1950 - 1960) of the counties. Twenty-seven contig- uous counties, with no major urban-industrial complexes, are used in the analysis. These counties are located in the upper part of the lower peninsula in Michigan. The final chapter contains a summary and some concluding remarks. CHAPTER II A PROFILE OF MICHIGAN Introduction The twentieth century witnessed a rapid transfor- mation of the nation from a rural resource-based economy to an economy based upon manufacturing. In the process of this change Michigan became a leading industrial state. In 1960 only 3.4% of Michigan's employment was in agricul- ture, while 6.4% of U. S. employment was in agriculture. POpulation has increased in Michigan faster than national rates and personal income has characteristically been higher in the state than for the entire United States. Although the state as a whole has been exception- ally prosperous, the same cannot be said for all areas within the state. Most of the growth and prosperity has been concentrated in the southern part of Michigan. While many of the southern counties have prospered, the counties in the northern lOwer peninsula and the counties in the upper peninsula, which constitute a large part of Michigan's land area, have generally declined. The rest of this chapter will deal with prosperity as exemplified by growth in population, income, and industrial employment-growth in the state of Michigan. Major emphasis will be on differential employment growth in the industrialized South and the less prosperous North. State Population, Income, and Employment In 1968 Michigan had a pOpulation of over 7.8 mil- lion. POpulation has increased in Michigan faster than in the nation. Michigan population grew 223% from 1900 to 1960 while that of the U. S. grew 136% (Table 1). Table l. —-POpulation of Michigan and the United States, 1900,1940, 1950, and 1960. Michigan as a percent Year Michigan United States .of U. S. 1900 2,420,982 75,994,575 3.19 1940 5,256,106 131,669,275 3.99 1950 6,371,766 150,697,361 4.23 1960 7,823,194 179,323,175 4.36 Source: Michigan State University, Bureau of Business and Economic Research, Michigan Statistical Abstract (6th ed.; East Lansing, 1966), p. 6. Michigan's prOportion of the U. S. p0pulation grew from 3.19 percent in 1900 to 4.36 percent in 1960. Per _ capita income in Michigan has been above that of the nation since 1938 (Table 2). In 1960, Michigan's per capita in- come was 2,324 as compared to 2,215 for the nation. 10 Table 2.--Per-capita personal income in Michigan and the United States, 1940, 1950, and 1960. Per Capita Income (Dollars) Year Michigan United States 1940 679 595 1950 1,700 1,496 1960 2,324 2,215 Source: Michigan State University, Bureau of Business and Economic Research, Michigan Statistical Abstract (6th ed.; East Lansing, 1966), p. 93. The United States increased total employment by 26.7 percent from 1940 to 1950 and 15.5 percent from 1950 to 1960 (Table 3). Michigan had corresponding growth, rates of 31.7 percent in the 40's and 14.0 percent in the 50's. During the two decades, agriculture declined as a source of employment in both Michigan and the United States. Michigan's employment in agriculture declined at a faster rate--minus 25.3 percent, in the 40's and minus 42.4 per-- cent in the 60's, as compared to -l7.9 percent and -38.5 percent for the United States. With the decline in agricultural employment there has been heavy migration to urban areas. In 1950, 29.3 percent of Michigan's pOpulation was classified as rural.1 lNew urban census definition. ll «OOAMMO msflpcflbm psoscno>oo .m.: .oeaa new omma .OJAH “ceasefisaoa co msmemo .m.s meflpfio .H-e use H-m .a .AmGAH "COOwCficmmzv oomeEoo mo pcoewpmoom .m.= «m .Ho> «owmfluomma was omqfi-oqofl “mezzoo Sn unmESOHQEm aw mcnmppmm zygote .hnnm¢ .Q HHozoq 50pm Uopstonow pcoEmOHdEo HODOp mo pcoopoa :.m was coma cw mpspHsowpm< 0.4H o.osw.~ o.sos.m A.Hm o.sos.m a.st.H mafiapmgeen Has o.mH H.mse.w H.44N.N m.am H.:am.m o.HHG.H mang20fiama-aoz s.ma- H.Na a.amH m.mm- a.amH m.mHN magpasoflpme cmmwnoflz pcm5m0HQEo prop mo pudenda 4.0 mm3_ome OH onspasownw< m.mH e.msm.ee a.ssq.em ~.GN a.qeq.em m.mkm.mq mmflepmseefi Hae m.mm a.mHH.me m.emm.om m.em N.~mm.om H.mqa.em magpaeoflsmm-eoz m.mm- s.emm.g k.~Ha.e m.~a- ~.~Hm.o s.0m4.m anabH:OHpm¢ mopSm copes Aooaa-ommav Aomaa-oemav mmemeo coma omaa mmemeo omma osmfl pcmonom AwUOMmsostpcOESOHQEM pcmopmm AmpcmmsochDOOESOHQEm acoc paw weapHOOABmm «OOMflEOHE cam .m.: mwmmeommemmmw mmmmmmeCMWpWmecwOMMWWMMMMWMus.m manwe 12 This had declined to 26.6 percent by 1960. The economic development of Michigan during the 20th century has been very dynamic. Production of durable goods, particularly motor vehicles, has to a large degree determined the indus- trial base and structured social and demographic patterns. As if by magnet, the state's population, influenced by the great automobile centers in the south, was drawn into urban clusters throughout Southern MiChigan (Figure l). The num- ber of cities with more than 50,000 persons grew from 3 in 1910 to 10 in 1950 and reached 17 in 1960. Urban growth in the state has taken place largely in the southern part of the lower peninsula. No city with population in excess of 25,000 in 1960 was located north of the city of Midland. Most of these cities were within the Detroit urban complex. During the first half of the century, Detroit grew from 285,704 to 838,517 inhabitants, whose prosperity re- flected that of the automobile industry. Almost one-half of the states pOpulation now resides in the Detroit urban area of Wayne, Oakland, and MaComb counties. While the motor vehicle industry has been the focal point of industrial develOpment, other major industries have contributed to the growth of the urban complexes. Cereal production at Battle Creek, furniture manufacturing at Grand Rapids, paper products at Kalamazoo, and other industries provided impetus to the growth of cities of Southern Michigan. 13 Figure 1.--Major cities of standard metrOpolitan statistical areas in Michigan. x Major Cities .0 .;..... MICHIGAN “If; . i . "1.. L . i . . . «an SdglnéflJ... nexégin . . . on... . orad'di'r'rfii'fifi'i’fi“ "£11 knapflds I .i. .Flin . - ' "Ia-#Ta'm .52.... WW” ‘ .._.. 1._L.Li2§.i. 9 ! ‘muwur—g lamW'-wk§m (Ann X . Kailama CO JI k ! . ‘ ”Fain. as.- ’WJE%§%O'r ~1~10-J.-__FJ { l 14 Northern Michigan and Southern Michigan While Southern Michigan has experienced tremendous growth in urban areas, there are no major urban areas in any county north of Bay County (Figure 1). In 1960, almost 92 percent of the state's 7.8 million people resided in the 41 southern counties (Table 4). Total Michigan population increased more than 1.4 million from 1950 to l960--with 98 percent of this increase accounted for by southern counties. Total increase in Northern Michigan was only 30,048 people, or 4.97 percent of 1950 pOpulation. The slow population growth rate in Northern Mich- igan was caused in large part by heavy outmigration. The 42 counties lost 46,958 people through outmigration. The 1960 census classified 62.8 percent of the pOpulation in Northern Michigan as rural. Only 23.4 percent of the peo- ple residing in Southern Michigan were classified as rural. Incomes are higher in Southern Michigan, and unem- ployment is lower. .Estimated buying income per-capita in the southern counties was $1,698 in 1959, while that in 1 the northern counties was $1,309. The census of 1960 1Michigan State University, Bureau of Business and Economic Research, Michigan Statistical Abstract (6th ed.; East Lansing, 1966), pp. 101-103. Note: The dollar figures are simple averages of counties. A weighted (by pOpulation per county) average would result in a larger difference. 15 Table 4.--P0pu1ation of Michigan, Northern and Southern Michigan, 1950 and 1960, total change, percent rural, and net migration.a Northern Southern Michigan Michigan Total. Number of counties 42 41 83 POpulation 1950 604,406 5,767,360 6,371,766 Population 1960 634,454 7,188,740 7,823,194 Total Change 30,048 1,421,380 1,451,428 Percent Change 4.97 V 24.64 22.78 Net Migration -46,958 209,190 Rural as percent of total pop. 1960b 62.82 23.45 26.64 aCalculated from Michigan State University, Bureau of Business and Economic Research, Michigan Statistical Abstract (6th ed.; East Lansing, 1966), pp. 9-15. bCalculated from U. S. Bureau of the Census, U. S. Census of Population: 1960, Volume 1, Part 24, p. 184. estimated that 9.9 percent of the civilian labor force in Northern Michigan was unemployed and 5.8 percent was unem- ployed in the South. Actually, the unemployment in the South was higher than normal, because the 1960 census co- incided with a recession in the automobile industry. The distribution of employment is similar to that of pOpulation. The 41 southern counties accounted for 92.1 percent in 1950 (Table 5). Total employment in the northern counties grew by 6,560 workers or 3.4 percent, while employment in the southern part of Michigan increased almost 15 percent, 1950-1960. l6 .mm-~ op H-s .aa .Aqua .oowmmo mcflpsfind pzoscno>oo .m. "20pmcflzmw3v mosoEEoo mo pcosppmaom .m.: «m .Ho> aowmauomma new ommH-oamH «kpczoo AL pcoszoadam cw menoppwm npsopo .hmanm< .Q Haozog Eonm pmpmasoamow mo.m m©.m pcoonmmammmmm¢ so.mH 0mm.~em mma.aeq.m mm~.~®o.m messagesrmm-eoz ms.em- ~em.es- owe.ek s4fi.mwfl asepasoname m.ma H.Na Ha.:H maw.amm www.mem.m mmm.NHN.N mmflppmaeafi Hae cmmanowz cumspsom om.~ mH.mH Ham mo pcoouoa mm .m< mm.mH mmmawm wamqmwfl mazgwmfl . waspHOOprmucoz so.mm- mem.0m- omn.4a www.4m mpspasoflbma N.s a.s mq.m oem.e eae.kaa ema.aaa maflnpmseefl afle cmeEOHz cnmzpnoz Aowmav Aommav pcmsmoadso .cowz omcmzo omcwzo owma ommfl Hmpop mo pcoonoa wcoopom pcthOHQEm pcoshoaasm m mm pcoshoadem m.Hmpop mo pcoonma w mm pacemoadso new ammumso pcosxoaaso «ome cam ommH «cwmflQOHz sponpzom ocm cnozpnoz cfi pcoEAOHaemun.m meme 17 The major reason for slow growth of the northern crounties was heavy reliance on declining agriculture, \Nhich dwindled by almost 59 percent in those counties. lAgriculture had accounted for more than 18 percent of Nor- thern workers in 1950. By 1960 this had dropped to 7.3 percent--a loss of 20,372 workers. All non-agricultural industries grew almost as fast in northern as in southern counties (17.2 as compared to 18.0 percent) from 1950 to 1960. Although it is un- likely that the growth rates would be quite as similar if the census had not occurred during an auto-makers recession in 1960, this does illustrate that in Northern Michigan, as in many similar regions, the problem is not so much a lack of growth in non-agricultural industries, but a major decline in agriculture, forestry, and mining. Non-agricul- tural industries need to grow much faster in order to pre- vent decline in the economy. The heavy reliance on declining agriculture and slow growth industries creates many problems of adjustment for rural areas. Low incomes and high unemployment can be relieved only to the extent that outmigration occurs or new employment Opportunities are created in other indus- tries. Outmigration erodes the existing social and cap- ital structures, and maintenance of social overhead, such as educational systems, becomes extremely difficult. 18 Thus, great demand arises for new or expanded in- dustries that will provide employment opportunities. Man- ufacturing industries are particularly desired because of their export commodity producing nature and their attendant retail and service requirements. It should be noted that there are some counties in Southern Michigan suffering a plight similar to those in Northern Michigan. However, there is the important differ- ence that most counties in Southern Michigan, even those with predominantly rural economies, are inextricably tied to counties with large urban-industrial centers. Economic adjustments in these counties are associated with the growth of the urban counties. CHAPTER I I I LOCATION FACTORS Introduction The location of manufacturing establishments de- pends on various factors, some of which are basic to a location decision, and some of which are not major deter- minants, yet influence the selection process. Factors influencing the location of industry are .critical to area and community develOpment. Knowledge of location factors can be particularly valuable for local development groups charged with the responsibility for in- dustry "procurement." Despite the many publications dealing with factors of location, little is known about the relative importance of economic, social, political and institutional factors affecting industrial location in rural areas. Although many rural communities have been heavily involved in ef- forts to attract industry, there is confusion regarding the amount of resources and the relative effectiveness of ‘various activities they might use to influence plant loca- tion decisions. 19 20 The purpose of this chapter is to review what face tors have been influential in attracting industry to rural areas, and what role the community might play. Major Location Factors Much theoretical and empirical work has been done on identification of location factors. Weber, one of the originators of location theory, regarded transportation and labor as the basic forces determining the region in which a plant could minimize costs.1 Agglomerating forces (marketing advantages, proximity to auziliary industries, external economies of scale, etc.) then determined a par- ticular location within an area. This general theory of location is applicable to all economic systems; but, ex- cludes institutional (i.e., interest, taxes, insurance, and management) and other factors. More recently, Greenhut developed a model consist- ing of three broad groups of locating factors: (1) demand, (2) cost, and (3) purely personal considerations. He then outlined an inclusive list of five demand factors, four costs factors with numerous subgroups, and three personal factors.2 lAlfred Weber, Theory of the Location of Indus- tries (Chicago: University of Chicago Press, 1928). 2Melvin L. Greenhut, Plant Location in Theory and Practice (Chapel Hill: University of North Carolina Press, 1956), p. 279. 21 The factors of location treated in this paper are: l). Markets-—proximity to consumer, producer, or expanding market 2). Raw materials--quantity, quality and relative prices 3). Labor--supp1y, costs, and quality 4). Transportation--facilities available, costs and quality of facilities (transportation costs are also reflected in industry orientation) 5). Utilities--avai1abi1ity, rates and quality 6). Taxes and legislation--police and fire protec- tion, health and sanitation codes, and zoning ordinances 7). Personal preferences 8). Community attitudes and economic inducements The first four may be regarded as basic factors in that one or more is of primary economic importance to every location decision. Relative importance of factors The relative importance to a firm of each location factor, in theory, is determined by the proportion of costs or revenues accounted for by that factor. For example, if labor costs are a large part of a firm's total costs, then the firm will locate to it's best advantage with regard to the type, quantity, and cost of the labor desired. 22 However, the firm will not necessarily locate where labor. costs are a minimum, but will pick another location so long as other costs can be reduced more than labor costs are increased. But, the labor intensive firm will have a greater incentive to locate at a minimum labor cost site. Surveys of rural areas Several studies have been conducted to determine the factors that affected plant location in rural areas. Wallace surveyed 72 new industrial plants which located in Southern Indiana from January, 1955, through December, 1958.1 The study covered 33 counties which have declined in pOpulation as employment in agriculture decreased. Polopolus and Williams surveyed the managers of 43 new plants in a five county area in South Central Louisiana.2 This area is a rural area meeting the various criteria of designation by the Area Redevelopment Adminis- tration as a "redevelopment area." 1L. T. Wallace, Factors Affecting Industrial Loca- tion in Southern Indiana, 1955-58, Indiana Ag. Exp. Sta. Res. Bul. #724 (Lafayette: 1961). 2Leo Polopolus and Robert W. Williams, Factors Influencing the Location of Manufacturing Plants in a Rural Area, La. Ag. Exp. Station, D.A.E. Res. Report No. 324 (1963). 23 Another study, not done in a rural area, but useful here is a study on location decisions in Michigan by Eva Mueller and others.1 Analysis of Basic Factors Managers of the 72 new plants in Southern Indiana indicated that the four most important influences on their location decisions were availability of labor, adequacy of transportation, proximity to local and regional producer goods and consumer goods markets, and availability of lo- cal and regional raw materials2 (Tables 8 and 9). Response by plant managers in South Central Louis- iana indicated that proximity to producer goods and consu- mer goods markets and the availability, quality, and price of raw materials were the major factors responsible for location decisions3 (Tables 6 and 7). Michigan manufac- turers mentioned most frequently labor costs, proximity to customers, markets and availability of skilled labor as their most important locational factors.4 Markets-~In both the Indiana and the Louisiana study, proximity to markets was cited as the most lEva Mueller et a1., Location Decisions and Indus- trial Mobility in Michigan, 1961 (Ann Arbor: University of Michigan, 1961). 2Wallace, 1. 3Polopolus and Williams, 1. 4Mueller et a1., 4. 24 .GN .a .nmwaao .qmm .oz psomm: condomom .m.<.Q .COHpmpm pcosflnoqu HBLOOHOOHnmd wcmflmflsoq «mop¢ Hansm m cw mpcmam meansbomOSsz mo cowpmooa one mewocOSHMCH weepomm .memwfiaflz..3.unonom ocm mSHOQOHom ooq "condom I H H IH H Nl—fi r-l HM 410! IN NW N HH—SHmHmI—I MN JNNr-I r—IH MN J own mr—I \O .3 \O'l—l H I H J COHpmHmwm m n m m pom mo N I C30 ICU OH ma Nd) (13.3 noowH m.m0pmapoonm mmflpfififiome ocean manaafimse OH compo cam Hmooq mowpfiHHpD mmxme weapfippm managesoo COAmcmaxo fipflawomm OHanHm>< cowpmpn0dmcwne boson m.pouowmdoum Acopflnv gonad mfiwflnoume 3mm mpmxnmz OH m w w o m a n Go xcmn some ca pogomm mcflomaa mnommcms mo AOQEBZ nopowm spaced sedan seam .mcmwmfisoq Hmnpcmo Epsom .mqsopm wcflpspommocms Haw «AmocODLOQEfi mo menow CHV mMODOOM coameOH mo mMCchmn .mnmmmcme pcmHmun.w canoe 25 Table 7.--Plant managers' evaluation of location factors, all industry groups, South Central Louisiana. Relative importance upon location decision Important to Plant location very moderately No major Disadvan- factor critical important influence tageous Number of plant managers Markets Closeness to con- sumer market 17 19 S O Closeness to pro- ducer market 5 13 2S 0 Closeness to ex- panding market 8 2b 13 1 Raw materials Accessibility 19 12 9 2 Quality 19 18 S 0 Relative prices 20 16 3 2 Particular require- ments 5 O O 0 Labor Supply: ' Men 8 23 11 O WOmen 1 7 22 0 Costs: Men S 20 17 O WOmen O 7 23 0 Quality: Skilled 9 26 8 O Unskilled 2 13 26 O Dependable S 28 9 O Attitude toward work 7 27 8 0 Transportation Available facilities 9 18 1b 2 Quality of facilities 1 2h 16 2 Costs 10 l9 13 0 Available facilities for expansion Land w . . .2. ' ll ' 29 0 Buildings 1 12 29 0 Transportation facilities 0 Labor supply 0 13 29 Raw materials 5 12 2h OOH 26 Table 7.--(Continued) Relative importance upon location decision Important to Plant location Very moderately No major Disadvan- factor critical important influence tageous Number of plant managers Community attitudes Availability of sites and plants 0 1h 28 0 Availability of fi- nancial assistance 0 7 3S 0 Special inducements O h 37 O Attitude of people 0 2 0 0 Taxes Local tax assessment and rates 1 6 3S - 0 State taxes 1 6 3O 1 State and local tax exemptions 2 9 3O ' 0 Utilities Availability 17 20 6 0 Utility rates 0 9 31 2 Quality of existing facilities and services 1 8 33 0 State and local legislation Sanitation codes 2 2 38 0 Zoning ordinances O 2 80 0 Right to work laws 1 3 38 O Other legislation 0 0 29 0 Source: Leo Polopolus and Robert W. Williams, Factors Influencing the Location of Manufacturing Plants in a Rural Area, Louisiana Agricultural Experiment Station, D-A.E. Research Report No. 32h, (1963), p. 28—29. 27 Table 8.--New Plant managers' rankings of nine location factors, all industries, Southern Indiana. Number of managers placing factor in each rank Factors ranked 1 2 3 h S 6 7 8 9 Markets 28 13 8 h h 2 h 2 5 Raw materials 16 15 8 6 6 S 2 2 9 Transportation 8 13 21 12 3 7 l l 5 Utilities 2 5 8 11 10 h S L: 19 Labor 19 15 16 8 - l - l 8 Taxes - 2 9 12 10 8 b 2 21 State and local legislation 1 1 3 lb 9 b 9 h 23 Community attitude 5 S 3 12 6 5 b 7 21 Other 15 2 h S 3 5 h’ 3 22 Source: L. T. Wallace, Factors Affecting Industrial Location in Southern Indiana, 1955:1958, Indiana Agricultural Experi- ment Station Research Bulletin No. 72h, (Lafayette: Purdue University, 1961), p. 10. Table 9.--Location factor priority rankings, all industries, Southern Indiana. Managers' response totals for the first Factor four factor rankings in Table 3 Labor 58 Transportation . Sh Marke ts 5 3 Raw materials AS Utilities 26 Other 26 Community attitude 25 Taxes 23 State and local legislation l9 Source: Ibid. 28 influential location factor. Proximity to eXpanding mar- kets was more important in Indiana than either producer or consumer goods markets. However, managers in Louisiana placed more importance on being near consumer markets. In both cases, producer markets were nearly as impOrtant as consumer markets. Raw Materials--Manager responses also emphasized the importance of raw materials in both areas. Access- ibility, quality, and relative prices were similarly important. LaboEf-Although markets were ranked first by far the most frequently in Indiana, labor received a higher response total of the first four factor rankings and was ‘second in number of primary rankings. Labor was not near- ly as important in Louisiana. HOwever, reSponses in both studies emphasized the importance of dependable labor with a willing attitude toward the work. Availability of skill- ed 1abor was relatively unimportant to the firms located in these areas. Wage rates are often regarded as a location factor of great importance, especially for labor intensive, labor oriented industries. However, in both Southern Indiana and South Central Louisiana, plant managers placed labor costs below both supply and quality of labor. This indi- cates that wage levels were not very influential in the selection of a site by the new firms located in these two 29 areas. Since both areas were recognized as labor surplus- 1ow wage areas, either the firms locating there were a type of industry for which wage levels were relatively un- important or wage rates, in general, are not essential lo- cation factors. Fuchs contends that on the national-all industry level the data do not support the view that relative wage levels are significantly associated with interstate move- ment of industry.1 However, this does not mean that wage levels may not have been important for particular indus- tries, but the factor was not significant for the general redistribution of manufacturing. Furthermore, he argues, the relative wage level has been the most significant factor in determining the direction of population migration. "The absence of a sig- nificant relation between comparative growth of manufac- turing and wage levels, suggests that the movement of sur- plus labor to industry has probably been as important as the movement of industry to labor in recent decades."2 The relative extent of unionization is often men- tioned as a factor in plant location. Although unioniza- tion may be correlated with high wages the "union issue" is often noted quite apart from wage levels. Almost two 1Victor R. Fuchs, Changes in the Location of Man- ufacturing in the United States since 1929 (New Haven: Yale University Press, 1962), p. 94. 2 Ibid., p. 101. 30 thirds of the managers in Southern Indiana considered unions a detriment.1 In addition to union wage scales some reasons given were: (1) loss of managerial preroga- tives, (2) concern that "outside meddling" would increase costs through strikes and work stoppages--possibly even when their local plant had no grievance, and (3) previous unhappy experiences with unions, and the belief that the workers did not want unions. Managers were often more concerned with local labor relations than with state labor laws, particularly in labor—intensive industries and where a considerable labor force is to be employed in the new plant. In some cases the labor relations history of the community was used as a basis for selecting the specific location.2 Transportation--transportation received relatively few number one rankings, but, at the same time, rated highly in terms of the number of times it was ranked sec- ond, third or fourth in the Louisiana and Indiana studies. This factor was of much greater significance in Indiana than in Louisiana. Firms in Louisiana tended to be more market oriented and, thus, product transportation rates were more important than in Southern Indiana. Availability and cost of truck transpOrtation facilities received more weight than rail, air, and water facilities. lWallace, p. 11. 2Glenn E. McLaughlin, and Stefan Robock, Why Indus- try Moves South (Washington: National Planning Associa- tion, 1949) ' p. 730 31 Analysis of Secondary Factors The previous four basic factors are, for the most part unalterable on a local basis. The following to some extent are manageable by communities and, therefore, of greater interest for this study. The basic factors normally account for a large part of manufacturing costs. Relative differentials in markets, raw materials, labor and transportation often are not substantial within regions or areas. Therefore, the basic factors are important in selection of regions or areas and secondary factors are more effective in deter- mining the site chosen within a specific area than in de- termining either the specific area or the general region.l Utilities--Less significance was attached to util- ities than to any of the factors so far discussed. Utility rates were a relatively minor consideration as compared to utility availability and quality. Managers (in Indiana and Louisiana primarily) were concerned about an adequate and reliable service. This suggests that adequate facil-. ities are a precondition for locating in a community, but utilities will usually not be an important factor in se- lecting one community over others because most communities have the facilities. In Indiana managers placed electricity as most important, followed by water, gas, sewage facilities, and coal. 1Wallace, 1. 32 In Michigan, an adequate water supply was one of the most frequently mentioned requirements, owing, per- haps, to the favorable position which the state enjoys in water resources and the industry located in the state for, that reason.1 Taxes and local legislation--Although state and local government officials view tax rates and assessments as important to industrial location decisions, most stud- ies indicate otherwise.2 In Indiana and Louisiana both state and local taxes were considered equally unimportant. There are three principle reasons why state and local taxes are apt to be unimportant: First, state and local taxes are usually only a small part of total costs of production and a large difference in tax rates would constitute only a small difference in total costs of the firm. Second, many managers view taxes as purchasing a service, higher levels of taxation associated with more and better services. Third, due to the intricate tax structureof various localities, firm A may have a lower tax rate than firm B at the same location. Undoubtedly there are instances where one locality is favored over another because of taxes, but this is likely to be only lMueller et al., 10. 2Fuchs, 88. 33 when the two communities are equally satisfactory in more important location factors. State and local legislation was evaluated by man- agers in Indiana and Louisiana as normally having no major influence upon the location decision. Plant protection (e.g., police and fire) and sanitation codes were attributed the most importance and were followed by zoning ordinances, political environment and water laws. PersonalApreferences--Despite the emphasis placed on the quantifiable cost and demand considerations by man- ufacturers, there is evidence that the process of plant location cannot be fully understood in purely pecuniary terms. Detailed questioning often reveals that many plants were started in their present location because the founder lived there, had business connections, or because suitable plant facilities were readily available.1 This sort of historical accident is far from the normally conceived method of locating by thorough investigations and calcu- lation, but 19 percent of the managers in Louisiana ranked "prOprietor's hometown" as the most influential factor affecting location. Personal considerations and historical accidents are likely to play a much greater role in the location of small firms than in the location of branch or subsidiary lMueller, 4. 34 plants of the large firms. Accounting cost considerations seem to play a greater role in new investment by large firms.1 Community attitudes—~The factor, community atti- tudes, includes an assessment of community spirit and com- munity willingness to cooperate with industry. Community spirit is evidenced in the attitude toward industry of local leaders and chambers of commerce, and especially in community provision of facilities and economic inducement for new firms. Inboth Louisiana and Indiana, community attitudes had no major influence upon the location decision. Avail- ability of sites was by far the most important aspect of this factor. Availability of building facilities and pro- vision for plant expansion was often noted in connection with availability of sites. Managers showed an interest in being able to see specific site alternatives, and to have documentation of land-use planning. Except when there were personal preferences for an area, sites were selected on the basis of quantifiable cost factors. There are factors which many communities can use as bargaining or negotiating factors. Those most commonly lIbid., 17. 35 used to induce new firms include financing, taxation, sites, buildings and research.1 Financing—-The usual inducement of this type is community financing of site, building, or plant with provision for long term repayment and low interest rates. The techniques or plans for fi- nancing most commonly available for communities are bond plans, both property and revenue, state authorities and financing, and capital credit banks. Also community members often play a very - active role in formation of new firms through pro- motion of stock and bond issue. Taxation--Exemption from prOperty tax, special treatment on inventories, use and sales tax, cor- porate income tax and recently tax concessions on reinvested profits have been used as bargaining points. Sites--There is an advantage in having specific site alternatives available to show prospective firms. Such sites may be held as industrial dist- ricts, community owned prOperty, or private 1Paul Brann, Actions to Influence Location Deci- sions of Firms, Papers presentediat the Conference Area DevelOpment (Athens: University of Georgia, 1962), pp. 126-127. 36 property under option. These areas may be planned and develOped or they may be held for development as they are sold to industry. Buildings-~Some communities seek to gain an advan? tage by having buildings under community ownership or available for immediate occupancy. These may be abandoned structures or shell buildings construc- ted for eventual occupancy. Research-—Location decisions are made by managers on the basis of their knowledge about specific alternative locations. Communities can inform managers of an accessible location by providing information useful to the firms decision. Such information might include data on labor supply, transportation facilities, markets, the location of complementary industries, and the general facil- ities available. Although Wallacel concluded that communities af- fected a limited but significant number of Southern Indiana location decisions, there is evidence communities attach too much importance to the role that economic inducements play in attracting firms. Furthermore, economic 1Wallace, 16. 37 inducements extended at a cost to the community may often Ibe unnecessary. Six of the 72 firms interviewed in Southern Indiana chose their sites primarily because of the incentives of- fered, while 15 additional firms accepted the incentives but did not consider the incentives crucial to their loca- tion decision. Twelve had not even considered other spe-' cific sites. Four of the six firms for which the incen- tives were crucial were in one industry, fabricated metal, machinery, and equipment.1 Twelve out of 43 firms in South Central Indiana received inducements, but only one indicated that thein- ducement was very important to the location decision. Eight of the remaining 11 considered the accepted induce- ment as not important and three did not even consider the inducement while making the decision. Some of the firms accepted inducements from more than one source, but a high prOportion were Louisiana 10-year tax exemptions. The other inducements were free land, nominal building rent, low land and building rent, a low interest loan from the Area Redevelopment Administration. Considerable evidence indicates that worth-while industry generally does not favor direct tax exemptions 1Seven firms listed community actions designed to discourage location in the community. But in only one in- stance were these actions successful in preventing location. 38 or other financial subsidies but, rather want tobecome a part of the community by paying their own way.1 There is widespread belief that research and dis- semination of information on local conditions and oppor- tunities can be an effective means of influencing location decisions. "Industrialists seem to have high regard for state and local develOpment agencies which can provide interested new or out-of—state firms with expert advice and information regarding local conditions and opportuni- ties. Personal solicitation by 'industrial ambassadors' 'was stringently advocated by a number of Michigan manufac- turers."2 Various agencies developed and provided information regarding communities in Southern Indiana that was impor- tant to plant managers. Local chambers of commerce, local industrial committees and local governments along with utilities were ranked by managers providing the greatest amount of help in providing information. There is some difference in Opinion over whether communities can more effectively influence location deci- sions of small firms as compared to large firms. In Louisiana it was found that little variation existed lHarold T. Smith, A Suggested Program for Financing Area Development in Michigan, a report of the W. E. Upjohn Institute for Employment Research (Kalamazoo: 1960), p. 51. 2 Mueller et al., 9. 39 regarding the importance of inducements by size of firm.l, In contrast, the larger the plant, particularly a branch plant or subsidiary, the less influence a community had on location decisions in Southern Indiana.2 Michigan manufac- turers also supported the view that financial concessions have some value in attracting smaller firms.3 Limitations of Surveys Although labor, markets, transportation and raw materials are the most important, it is apparent that the relative importance of each of the basic locational fac- tors varies with the region studied. Also the aspect of each factor emphasized varied with the region. For exam- ple Southern Indiana manufacturers stressed availability of labor while Michigan manufacturers noted cost and skill of labor as important to their location decision. In reality this diversity is to be expected be- cause of the wide variations in the needs of different types of plants. The factor that is mentioned most often in Southern Indiana is not the same as in Michigan or in South Central Louisiana because the new firms have located in the respective regions because of the characteristics lPolopolus, p. 42. 2Wallace, p. 18. 3Mueller et al., 9. 40 of that region. For instance, Southern Indiana was gen- erally accepted as a labor surplus area and new firms loe cated there because of the availability of labor. Thus, in a survey of plant managers in Southern Indiana, avail- ability of labor would be stressed. Likewise Michigan (i.e. Lower Michigan) is an area with high wage rates. Thus, any new firms which have located there would do so for skilled labor and a survey Would reflect this. In this way, local surveys do not indicate accurately the correct ranking of location factors for all industries, or even necessarily for a very large part of the industries. Rather, a specific type of new firm is heavily represented in each of the regions and a factor cited as important in one is not necessarily regarded as important in another region. It has been the tendency to identify locational factors by a survey of new plant managers in a local area and then to attribute these same factors the same relative importance for future develOpment of that local area. How— ever, it is unlikely that those factors stressed in local areas will exert like influence for continuous growth in the future or even that if similar factor influence did persist it may not be a basis for substantial growth in rural areas. The surveys of local rural areas identified prox- imity to markets as the most influential location 41 factor.1 (Here it should be remembered that markets were most influential for the typg_of firm attracted to rural areas.) Fuchs states, "the evidence suggests that demand has not been a major determinant of locational change, 1929—54. We infer this, first, because market oriented industries are only a small part of total manufacturing, and, second, because a significant portion of the changes in demand which did occur should properly be considered as the result of the redistribution of manufacturing, rather than its cause."2 Thus, two points are relevant: (l) iden- tification of the relative importance of location factors in one rural area is not necessarily applicable to another local area, nor to rural areas in general, and (2) the fac— tors, identified in a local rural survey as of primary im- portance, may be comparatively inconsequential for substan- tial long-term growth of underdeveloped rural areas. Analysis of Location Factors by Industry A ranking of the importance of major factors by manufacturing firms will vary from firm to firm, depending on the particular needs of that firm. Individual manufac- turing firms in the same industry often have similar needs and thus will attach similar importance to the various lWallace, p. 10 and Polopolus, p. 25. 2Fuchs, p. 162-63. 42 location factors. Thus, in many instances, it makes sense to use the industry as the unit of analysis. Identification of important locational factors for each industry can be more fruitful than ranking factors by relative importance for all industries. If the important factors for each industry are recognized, develOpment agen— cies can match an industry(s) with the attributes of the particular industry(s) for which the community can satisfy the most location requirements. For some of the industrial groupings dealt with, the fact will be mentioned that this industry has been growing or declining in the East North Central Region (including Michigan) relative to the rate of growth of the industry in all the regions. This is impor- tant because, except for industries strongly oriented to natural resources, rural areas tend to suffer greater com- parative losses than do metrOpolitan areas as the industry shifts away from a region. Food and kindred products Because the food processing group is such a broad category, certain industries within the category make lo- cation decisions on different criteria. Many food indus- tries are market oriented, that is, tend to locate close to local selling points. Such processors include ice cream, bakery, and beverage products. Major locational shifts of these industries have been related to geographical 43 shifts of income and population. In South Central Louis- iana the remaining significant factors were transportation, utilities, available plant facilities, community attitudes and taxes, in that order. Another group of food processors are primarily influenced by the location of raw materials. These pro- cessors normally use bulky, perishable, and/or heavy raw materials that lose weight in the process. Examples in- clude vegetable canneries, poultry processing plants, and cheese factOries.1 For this group in South Central Louis- iana markets, labor, transportation, available facilities to expand, utilities, and prOprietor's hometown were the next important factors. When the source of raw material changes, these processors change location.2 Textile, clothing and leather products Labor is usually identified as the most critical location factor by managers in this group. The desire to avoid unions and to utilize less expensive labor has been important in national employment shifts in this group. Michigan and the East North Central States have suffered comparative losses of employment in this group. lFuchs, 250. 21bid. 44 Lumber, wood andgpapergproducts There is no clear indication of the most important factors in this group. PrOprietors hometown, markets and raw materials each received number one rankings in South Central Louisiana. Logging, sawmills, and millwork, which account for a large part of the employment in this group, are greatly influenced by the pull of natural resources. Likewise paper products are oriented to natural resources. . The East North Central region has been a comparative loser of lumber and wood manufacturing because of Western sources of raw material.1 Furniture The East North Central region has also experienced comparative losses in furniture manufacturing during 1929 to 1950. Expensive labor and lack of raw materials appear to be the primary reasons.2 Printing andgpublishing Proximity to markets and labor are important to printing and publishing plants. Three of four firms in South Central Louisiana located their plants primarily to lFuchs, 252. 21bid., 253. 45 utilize the typesetting skills available. Other factors, in order of importance, were markets, available facilities, and community attitudes. Chemicals and allied products The chemicals industry, apparently, is very sensi- tive to transportation rates and costs. Plants tend to locate either at the raw material source or within a local market. Plants were located in Southern Indiana and in South Central Louisiana, primarily to serve the local mar- ket. Some managers realized transportation rates on final products placed their manufactured goods at a disadvantage in wider consumer markets and at the same time were con- cerned with transportation rates of raw materials from other states. Petroleum and coal Natural resources has been the dominant factor in selection of a specific site--followed in South Central Louisiana by transportation, available facilities, and markets. Stone, clay and glassgproducts Manufacturers of stone, clay and glass products .primarily seek locations either near raw material sources or near consumers depending on the particular product of 46 the individual firm. Also in many cases the raw material is available at various locations. Then closeness to a desired market is the crucial factor. Concrete products are a large portion of this group in rural areas. In South Central Louisiana concrete plant managers considered closeness to consumer market as the most critical factor. This was followed by proprietor's hometown and community attitudes. Primary metal products The attraction of industrial or producer goods markets has been the major influence on primary metal pro- ducers in Southern Indiana;also of importance was Southern Indiana'scentral location for the widespread input sources for these plants and a supply of trainable labor. Fabricated metal and machinery products In Southern Indiana and South Central Louisiana metal fabricators primarily desired proximity to consumer markets. A trainable or skilled labor supply was also im- portant. In addition, facilities available for expansion and community attitudes were mentioned. In Southern Ind— iana eight of twenty-eight firms in this group were influ- enced by community inducements. The inducements were most- ly financial aid in obtaining building and land. 47 Motor vehicles and equipment Location of motor vehicle manufacturing plants is known to be influenced by the economies and other advan- tages of locating in metropolitan areas. Other transpor- tation manufacturing appears to be more amenable to locat- ing in non-metrOpolitan areas. Recent Growth and Present Production Characteristics as Related to Growth Potential In the short run, at least, the chances of attain- ing additional jobs in a small community is greater for those communities which had a recent favorable movement1 of industry.' The Michigan Department of Commerce, Office of Economic Expansion, tabulates shifts in plant locations as they appear in published sources.2 In their reports of Movements of Industry, for each calendar year, the com- munity is identified in which the favorable movement took place. lFavorable movement includes: move-ins from other states, local expansions (new buildings constructed at the same site or within the same county), in-state eXpansions (companies establishing operations in a community different from their existing location), new starts (by Michigan peo- ple with Michigan capital), and re—Openings of Operations. 2Published information about plant eXpansions and industrial movements is subject to limitations. News re- leases are usually uncertain as to date when the movement will actually occur. Also many small industry movements are not newsworthy and so are never reported. 48 In the 1960 census of population, there were 419 communities in Michigan with pOpulations of 5,000 or less.1 Of these 419 communities, 188 did not have a reported fav- orable movement of industry during the years 1960 through 1966. A total of 146 communities had at least one re- ported favorable movement of industry during the period 1960 through 1962. During the remaining four years, 1963 thrOugh 1966, a total of 189 communities had at least one favorable movement reported. Of the 146 communities having a favorable movement the first period (60-62), 104 communities, or 71 percent, also had a favorable reported movement of industry during the second period (63-66). Of the 273 communities which did not have a favor- able movement reported in the first period, only 85, or 31 percent, had a favorable movement of industry during the second period. For small communities seeking industry, the first step or the first increase in industrial employment is very important. After the initial growth, a community has a much greater chance of additional growth in employment. 1Includes all incorporated places of 5,000 or less population (except incorporated places less than 2,500 lo- cated in an urbanized area) and all unincorporated places of 1,000 to 5,000. 49 Another aspect of industrial movement is the source or point of origin of the firm providing the additional employment. In Michigan, during the decade 1957 through 1966, sixty-seven percent of all reported industrial move- ments were local expansions. New starts accounted for 16 percent. In-state expansions (companies establishing Op- erations in another community in addition to their existing operations) represented 9 percent, and move-ins from other states accounted for 6 percent. Thus, at any point in time, the potential for em- ployment growth within a community is highly related to the existing industrial complex. There is a resistance to establishing plants at places other than the existing fa- cilities. Moves or expansions into new communities by firms already established in another community, either in- state or from another state, accounted for only 15 percent of all reported movement of industry in Michigan, 1957-1966. Community programs designed to increase local em- ployment should not overlook possibilities for expanding existing industry. Programs facilitating expansion of local industry may be less costly than attempting to over- come the resistance of non-local firms to establishing facilities at other than their present location. The potential for attracting new firms would also .be greatest for those industries which tie-in (either on in- put or output side) with existing and eXpanding industries. 50 Summary While the managers of all firms placed great em- phasis on several or all four of the basic location fac- tors--markets, raw materials, labor and transportation, the relative rankings depended on the industry represented and on the individual firm. Also, the overall importance of any one of the four was different for each of the areas surveyed. This follows from the differing characteristics and spatial lo- cation of the areas surveyed. Firms will locate in an area for the relative advantage that can be obtained in that area. Manager rankings of the factors reflect this advan- tage, which will vary for different areas. Secondary factors were of much less importance to firms. However, they have received much attention because communities can bargain with firms over these factors. In general, community offered inducements were thought to be ineffective. Evidence also suggests that often inducements are extended at a cost to the community, and firms take advantage of the offer when it was in no way crucial to their decision. Managers often thought highly of "industrial am- bassadors" who could provide useful information about prospective locations. 51 Recent industrial growth and present production patterns within a community are essential starting points for understanding the growth potential of that community. CHAPTER IV SHIFTS IN INDUSTRIAL EMPLOYMENT Introduction Changing economic conditions often cause dramatic relocation of industrial activity from one region to anoth- er. But perhaps even more important is the continuous re- adjusting of individual industries among states and even among areas within a state. This shifting may be in response to growing or de- clining markets, new transportation routes, utilization of different resources, variations in labor availability or any number of locational factors. Any change in the impor- tant locational factors can make one area relatively more attractive than other areas. Theoretically, when an area gains an advantage its prOportion of industrial activity will increase. Quite often this increase in activity will not be an easily detected upturn in the area's entire eco- nomy, but rather, only marginal movements by one or two specific industries. Even though these changing conditions may be quite obscure at any point in time they can be very significant over a period of time. Thus, it is important 52 53 for anyone interested in an area's development to be aware of the industries for which the area has an advantage and could be most easily attracted. In this chapter, the question to be answered is; what industries in the state of Michigan have been shifting into rural areas? On the premise that the to be identified industries found it advantageous to shift to rural areas,' some indication can be obtained as to which industries are most likely to locate in rural areas in Michigan. Devel- Opment agencies can then concentrate efforts on those in- dustries for which the rural community has a relative advantage. Technique used The technique to be used here for identifying the industries in which an area is advantaged is shift-share analysis. Shift analysis1 is a simple analytical approach based on a study of three main components of area growth: that part attributable to state growth, that part attribu- table to rates of grOwth of all industries, and that part due to differences between rates of growth of industries within a single area and rates of growth of the same indus- tries in other areas. It is a means of comparing growth 1An example of calculation is given in the Appendix 54 of the areas of the state with the growth of each other by using the entire state as a base. This method identifies and quantifies three compo- nents of area industrial employment growth. State growth, the first component, is the amount by which employment in industry i would have grown in area j, if it had grown at the same rate as the state all-industry average growth rate. The second component, industry mix, is the amount employment in area j would have grown in industry i more than (less than) state all-industry growth becauSe industry i was a fast (slow) growth industry in the state. Area share, the third component, is the amount of additional (less) employment in area j in industry i be- cause industry i grew faster (slower) in area j than it did in the state. Area share determines whether an area maintained, gained, or lost in its 'proportion' of the state's employment in industry i. By summing the second component across all indus- tries, it can be determined whether an area within the 'state gained or lost employment relative to the rest of the state because of it's mix of industries. An area with a high porportion of fast growth industries will gain re- lative to the remaining areas. Likewise, by summing the third component across all industries, it can be determined whether an area 55 maintained, gained, or lost in its proportion of employment in total industry.1 But, more important for purposes of this paper, this technique can be used to identify those industries which grew comparatively faster in rural areas. Delineation of areas I The 83 counties in Michigan have been delineated into three groups or areas for shift analysis (Figure 2). Area 1 includes all 42 counties in the Upper Peninsula and in Northern Lower Michigan. Area 3 includes 14 counties in Lower Michigan which are Standard Metropolitan Statis- tical Areas (SMSA). Area 2 comprises 27 non-SMSA counties in Lower Michigan. Three areas were designated in order that the rural counties could be compared with the highly industrialized counties and, at the same time, counties which are near large employment centers could be analyzed in terms of the influence of urban centers, even though some of these counties may be quite rural in orientation. Of primary interest here is area 1, which is es- sentially rural with no major pOpulation and industrial concentrations. lThe shift-share technique was used and the data published by the U. S. Department of Commerce. The Depart- ment of Commerce data is similar, but on the national basis instead of a state basis. Refer to footnote 1 in Appendix A, p. 109. 56 Figure 2.--Areas used for shift-share analysis. Area 1 [:1 Area 2 Area 3 E:::] ‘1 . MICHIGAN . ngafi ! Icmm'th f’ L_._. .1 “In .! ‘HACI'INAT L.-. Inn-x . J . 'fi/‘N of Mag if; . liangia- -ii----+' gift/18%: ! ' mcmrba-o—ifinmma. MISS-51:83:09 ' focthmAvco-Tmoom ,maau 'Igloqma'q ‘ £0-.r a”.;.: I".;"'. :. 3... : - ° ' usual 3 11. . a . a . o '.. "'JISKI “ ' :.——‘ .0 50* 1 — - . 0. “cl“ z U s ICUIVYON 1111742669.. 1. 0.. .r. '0 d. .0 o. . '..c.':'; o. o :|:.o ’. - ';°." -1 f air-‘3 an”; we?!) .: .- . arr v imam" uwuom 45'. J - - -2 « 1 ;°. JACKSON ‘msurrmw (warms. / O o s ‘0.." . . 57 Area 2 includes the counties in Southern Michigan which do not contain major employment centers. Some of these counties are rural by all standards, and yet can be expected to be dissimilar from the counties in area 1 be- cause of proximity to the SMSA's. Proximity to large in- idustrial centers affects their economies through selective industrial movements in addition to providing a general stimulus to the county economy.’ Employment and Components of Employment Changegin Michigan, 1950 to 1960 The rate of growth of all employment in Michigan was 14 percent from 1950 to 1960. During the same period, employment in all industries increased: 3.4 percent in area 1, 16.5 percent in area 2, and 14.5 percent in area 3. Table A in the Appendix B contains employment and the components of employment change for the three areas in Michigan from 1950 to 1960. Thirty-two industries were used for each of the three areas. Employment and components of employment change in upper Michigan Table 10 presents employment and components of em- ployment change for area 1 during 1950 to 1960. Only man- ufacturing and other commodity producing industries are shown because they are generally considered to be basic or income generating industries through production for exPort. 58 .UOAMHooam poc cam .umcpo one .mposoonm pocpmofl ccm panacea «mposcopm omfiHHm one noose amponponm mmme pew mmao.aocopm «Hopes pepeOanmm .Hmpoe Redefine "moosaocfi meanspowmscms msoocmfiamomws cam heaped .HHH .o .e oaeoe .xeoooooeo Nem.0m- emo.m- mem.ma- emm.o omm.ea Nae.em ooopaooaome eme.a - eom.m- amo.m - Hmw.H aee.m Hmm.mH onopaoeoe .ooos .oooeoa mae.a - emo.a- HHH.H - om: mwo.H maw.m .eaooo e ooHoHooe oopo: eam.m cap - mae.a ome.a 4em.mfi o~o.oa .mee .ooae e oooeo H amm NNN - as: 4mm mea.m emm.m .oooo ooaaao e oaooaeooo mm; - 4e - poo - mad ama. HN4.H moaeoeoae e ooaoeoooom flea so - NH - Ham mme.fl Hem.a .mme Hoooeoe mom - m4 - mmw.H - one.a aee.flfl eee.aa space: an - m we - mm and sea oooeoooo Haas oaaexoe mam aOH mom pom aem.m Hm4.H meeooeaooo e moeeoaee 95; H8 mm; awn pram SEN oeoooooo posses e oooe 8.3 no «mo S moos Rm .mee .266 .32... oofio mom.m mmm.H mom - Hem.a www.ma mem.aa ooaoooeeoooo poooeooo ooa.m omm.m mew men Hem.m Hee.m .mme .ooee ooeoo e .ooam owoooo oeoom .xmz neeooo coma omaa ooaeeoeooH 8.8 more US$65 83m mmcmco pcmshoamem mo mucocomeoo ca pcoemoaasm .mmfinpmsccfi mcfloaponm huwcossoo gonna one weenspowmscms noomfluomma .H cons cw omcmco pOOESOHQEm mo monocomsoo 6cm pOOEhOHasma-.OH canoe 59 The industries are ranked by size of the regional share component. Of the manufacturing industries, area 1 obtained the largest increase in its share of electrical and other machinery manufacturing. Also, area 1 gained in its share of contract construction, other transportation equipment manufacturing, food and kindred products manufac- turing, printing and publishing manufacturing, and textile mill products. Textile mill prOducts, however, were unim- portant and actually had a negative total change in employ— ment. The implications of this ranking are apparent. If past trends continue, the odds are against any significant growth of motor vehicle or lumber, wood, and furniture manufacturing in area 1. On the other hand, industries such as electrical and other machinery manufacturing, and food and_kindred products have exhibited tendencies to shift into this area. The ranking in table 10 in no way indicates the most desirable industry to be sought, but only indicates which industries showed a tendency to locate in this area. The industry (as well as the particular firm) easily at- tracted may very well not be the most desirable. Desir- ability of an industry is the result of several factors, one of which is growth rate. It is important to a commu- nity that a firm locating there will continue to increase employment. 60 Those industries which increased state employment at a rate faster than Michigan all-industry rate of growth have a positive figure in table 1 in the industrial mix column. The positive figure is the amount of additional employment in area 1 due to that industry's rate of growth above the Michigan all-industry rate of growth. From this standpoint, other and miscellaneous manufacturing, elec- trical and other machinery manufacturing, other transpor- tation equipment manufacturing, printing and publishing manufacturing, and food and kindred products manufacturing are the most desirable, in that order. On the basis of employment growth due to industrial mix and regional share, four industries are apparently the best prospects for locating in area 1. The four industries are electrical and other machinery manufacturing, contract construction, other transportation equipment, and food and kindred products manufacturing.1 Although contract const- ruction was a slow growth industry, its total growth was the second largest. 1The recreation industry is often considered to be making great strides in upper Michigan which has favorable natural resources for that industry. This was not evident during the decade 1950 to 1960 (see entertainment, recrea- tion services industry in Table A, Appendix B, p. 111). Area 1 suffered a decline in employment in entertainment and recreation services. However, the industry classifica- tion includes many activities, such as motion pictures, theatres, bowling, pool, baseball, football, and golf, which would decline with a pOpulation decrease. The in- dustry also includes bathing beaches, swimming, parks, 61 Employment and components of employment change in lower Michigan, non-SMSArs, 1950—1960 Table 11 presents employment and components of em- ployment change for counties in Lower Michigan (area 2) which are not part of a Standard Metropolitan Statistical Area. The effect on industrial location in rural counties resulting from proximity to large urban-industrial complex- es can be illustrated by comparing table 11 with table 10. The number of employees in each industry per county is much higher in area 2, and the effect of proximity to .large centers is evident in the particular industries which shifted into area 2. Motor vehicles and equipment manufacturing had the largest shift into this area. However, this may not be an established trend because during the census year 1960 the automobile industry was in the low part of a short-term cycle. The dramatic decreases in automobile employment in Detroit for that year caused a comparative shift to area 2, but could be completely reversed in a short time. Also, the total change was negative. picnic grounds, and rental of boats, etc. The industry classification may cover up a trend toward these latter resource based activities. Also, it is not possible to know how much employ- ment in retail sales, lodging, etc., was related to these activities. (The data on components of employment growth obtained with the shift-share technique could be used in a valuable study of inter-industry linkages. For instance, the association between recreation and retail trade might be examined.) 62 .NHH .o .e oases .xaoooeeeo oaN.N Hae.m- Nae.m acm.~ mmo.mH Nae.mfl oooeoooo possess e oooe mma.H emu - eem.a oeo mom.o mem.a meaaoaaooo e woaooaee mOH - me . amm - sou mme.a wmm.a moans: mmH - ma - mefl - mm 4mm use ooaooeoae e_seeooeom omm - ma mam - omH 4mm ewm.a oooeooeo Haas oaaoxoe mao.ma Ho aom.m am~.4 mmo.m4 wo:.om Spocanoos ponpo e .omHm own oaa 0H - mad mee.a efle.a .mms Hoeoooe mmm.4 mam.a eme.m mam emo.e awe.~ .oaooo nonpooeoeoooep conga oom.mm- aee.a 0mm.ee- mee.HH mmH.om mmm.mm ooooaooeowa oeH.H aom.~ mo~.~ - moa.a oao.a oma.e oooeaoeom e oooe..oooeeg eeo.m oom.m eae.a - mao.m mm~.mm Hem.H~ eoeooepeoeoo eooeeooo mma.e mme.~ Hoe.a mwo.H emm.ma Nam.» oeoeoooo ooaaao e oaooasoeo Heo.ea moa.e cam.o omN.o oam.ao aae.ee .mee .ooee e_ooooo Hem - emH.m oem.m - Nem.m maa.em oee.e~ .oaeoo e ooaoaooe pope: owned 22m .3: 6.6.5 83 omafl moan» was H38 3.2 US$85 33m mmcmzo newshoamem mo mpcmcomeoo 2H pcthOHaEm .moanpmsocfi wcfionconm hufioossoo gonna use mcfluopommscoe .owmanomma am mono OH omcoso pOOEAOHmso mo masocomsoo cam pcmshOHQEMau.HH wanes 63 The next largest increases in the share of area 2 were the industries other and miscellaneous manufacturing, chemicals and allied products, contract construction, lum- ber, wood and furniture manufacturing, and other transpor- tation equipment manufacturing. Several'additional indus- tries had a small positive regional share increase in area 2. Of those industries exhibiting a tendency to shift to area 2, the five accounting for the largest individual increases in employment due to fast growth were other and miscellaneous (primarily metal industries), chemicals and allied products, contract construction, other transporta- tion equipment, and electrical and other machinery manufacturing. In general (i.e. area 1 agd_area 2), electrical and other machinery, and transportation equipment, other than motor vehicles, were the most important industries resisting an overall trend toward location in metropolitan areas. In addition, other and miscellaneous manufacturing, and chemicals and allied products gained competitively in counties in lower Michigan, other than SMSA's. In upper Michigan, additional industries shifting into the area were food and kindred products, and printing and publish- ing manufacturing. 64 Several examples are presented in table 12 that illustrate the relative shifts that have taken place among the three areas. Table 12.--Regiona1 share component, 1950 to 1960, agri- culture, electrical & other machinery, and chemicals & allied products manufacturing, areas 1, 2, and 3.a Electrical and Chemicals and. * * * Agriculture other Mach. mfg. Allied products Area 1 -5,656b -2,380 -222 Area 2 1,779 61 2,438 Area 3 3,884 -2,444 -2,216 "gAppendix, Table A. The columns should sum to zero, but do not in some cases due to rounding of growth rates. Agriculture was the source of a major part of em- ployment loss in counties distant from large urban centers. All three areas lost employment in agriculture due to the state decline in agriculture. However, area 1 lost 5,656 additional employees because agricultural employment de- ' clined faster in area 1 than in the entire state. In large part, this was due to the relatively poor agricul- tural land in Northern Michigan. Areas 2 and 3 gained what area 1 lost, with the major part going to area 3. This partly exemplifies the fact that agriculture in Mich- igan is becoming urban-oriented as large scale farms in- crease on the urban fringes. 65 Electrical and other machinery grew comparatively in area 1, with the major loss in area 3. Chemicals and products gained in area 2. By summing the regional share component across all industries it is possible to see how competitive an area was with the rest of the state in total employment. This is shown in table A in the Appendix. Employment Redistribution for Non-Standard Metropolitan Statistical Areas, 1958-1963 The coefficient of redistribution is essentially another way of looking at the phenomena discussed above. It can point out those industries which are shifting by comparing an industry's percent distribution of employment among areas at two different points in time. For example, a coefficient of redistribution can be calculated for food products manufacturing which cOm- pares; the percentage food manufacturing in non—SMSA coun- ties is of food manufacturing in Michigan for 1950, with the proportion in 1960. By simply subtracting the 1950 percentage from the 1960 percentage, the absolute increase or decrease in an area's prOportion of an industry is ob- tained. If the result is divided by 100 the value of the coefficient ranges between -1 and +1. A positive value for an industry in area 1 indicates that the industry 66 shifted into the area; and the larger the value, the greater was the positive shift.1 Table 13 presents the coefficient of redistribu- tion for non-SMSA counties in Michigan for the period 1958-1963. The coefficient presents a picture similar to that of the change in regional share presented in tables 1 and 2. Except here, area 1 and area 2 are com- bined as simply non-SMSA counties. Also the industrial classification is somewhat more detailed and includes more recent data. An example of interpretation from table 13 would be; counties, which were not SMSA's in Michigan, had an additional 6.5 percent of total electrical machinery man- ufacturing employment in 1963 which they did not have in 1958. In general, the ranking in table 13 indicates the importance of the industrial shifts into non-SMSA counties. Electrical machinery and equipment was second only to leather products, which was minor in terms of total em- ployment. Miscellaneous manufacturing, rubber products and ' plastics, primary metal industries, and products of petro- leum and coal are ranked high in terms of favorable 1For a more complete formulation refer to Walter Isard, Methods of Regional Analysis: An Introduction to Regional Science (Cambridge: M.I.T. Press, 1960), p. 254. 67 moms megawaw one .msmcmo one ca co>Hm mGOHpmowmwmmmHo ouam mo ommuobm one mafia: an cmpmasoflmo .mmmfl cam mmmH .mnonspommscm: no msmcoo .oopoesoo mo .pmon «.m.=m oma.- amH.SN oHH.aH ome.m moo.m omoaw e.seHo .oooom mso.- eme.aw moo.om emo.e Heo.m ooeeoxam e oeopaepoe amo.- meo.ma Hm4.a emo.fl 4mm.H ooeeaoe poo oeeoeopooom mmo.- mmo.mm Nae.e~ Ham.HH eme.HH oooeoooo_ooAHHo oeo ooooa moo.- oam.em mom.me Sem.oa mmfl.afi opoeoooo Hoeoe_oopooaeooe eooo.- mm4.Hm~ oom.mmm mea.mm sea.o~ .oaeoo ooaeeoeoooooee Hoo. mme.ma eao.wa sem.m mmm.~ oaeooe oooeo one Hormone moo. Nea.om see.am Hum.m moe.m .ooe.ooaaao e .Hoe .moapoaoe woo. ema.om mee.mm mom.HH Neo.ma oeoeooeo ooaflae one oflooaeooo moo. Hao.ea Hmm.mH Hmm.oa mmm.a oooeoooo oooe_ooa tenses ado. omo.mm omm.ao mme.aa ema.am oeoeooeo oooooex one oooe 6H0. aoo.~ mee.~ amp amp Hooo e esoaoopoo mo oeoooooe moo. Nao.~ SNH.N ems 64m oooeoooe Haas oaaexoa mmo. mo4.me Noe.ae 0mm.ma Hem.4fl ooeeooeooa Hoeo5.meoeeum mmo. www.44H aoe.~ma Hmm.mm 04a.4~ A.ooao .oxov eeooaoooz ego. amm.ea som.ma mma.e aaw.m ooaeooao poo oooeoono eoooom mac. mmH.oH omo.o~ mma.m em~.m .wme ooooooaaoooaz moo. emo.om Hom.mm oao.ea mae.ma .oeooo e .oooe Hooeepooam oea. mom.m oma.w Hoo.m mee.H oeoeooee poeeooa e pooeoog .<.m.z.m-coe mom mwma mmmfl mesa mmafi 8333 uses co eooeoameooo 1 oeomaeoaz as o.e.m.z.m.ooo pOOEAOHQsm Happy OH pcosmoaasm .¢.m.z.mucoc pom soapsnanpmfloon mo pcoaoammooo .mwmaummma .mmmnm Hwoapmfipmpm cwpfiaoaonpos.onmpcwpmucoc ca one cmmw£OHz cw newshoamsm:u.ma mange 68 redistribution toward non-SMSA counties. Previous analysis (i.e. tables 10 and 11) indicated that this movement was primarily into counties close to the counties containing urban-industrial complexes. Other industries, such as food and kindred products and printing and publishing, which table 10 indicated as shifting to area 1, did not have a high coefficient be- cause area 2 contains a large part of the total employment in non-SMSA's. In fact, the ranking in table 13 may be viewed as a combination of tables 10 and 11, with only 30 percent of the weight on area 1. The shifts exhibited in the shift analysis for the period 1950 to 1960 persisted into the period 1958 to 1963. Dispersion of Additional Employment The dispersion of additional employment created by an industry among the counties in the area also affects the chances of a particular county for procuring that in- dustry. For instance, of those counties in area 1, which obtained additional employment in motor vehicle manufac- turing during 1950 to 1960, two counties accounted for more than 50 percent of the new employment; eleven of the 42 counties gained while twenty-nine lost. The chance of a particular county in area 1 gaining employment in the motor vehicle industry is small based on past performance. 69 Manufacturing industries are ranked in table 14 by the number of counties in area 1 that experienced an actual employment increase in that industry. Table l4.--Number of counties gaining and losing employment by industry, area 1, 1950 to 1960a .Number of Counties Gained Accounting for Lost Employ- at least 50% Employ- ment of Gains ment Elec. & other mach. 37 5 4 Food and kindred 32 7 10 Printing & pbl. 28 6 10 Other 5 Misc. 28 6 13 Other trans. equip. 23 2 12 Lumber, wood, furniture 20 6 21 Chemicals 5 allied 16 2 18 Apparel mfg. 15 2 15 Motor vehicles & . . . ll 2 29 Textile mill products 9 1 23 aCalculated from: Lowell D. Ashby, Growth Patterns in Em- ployment by County, 1940-1950 and 1950-1962, Vol. 3, U. S. Department of Commerce (Washington: U. S. Government Printing Office, 1965), pp. 7-1 to 7-28. Additional employment in electrical and other ma- chinery manufacturing was widely dispersed amOng the coun- ties. Thirty-seven of the 42 counties in area 1 gained additional employment during the decade. 70 Food and kindred products manufacturing was second in terms of the number of counties which experienced gains in that industry. This was followed by printing and pub- lishing manufacturing and other and miscellaneous manufac- turing. Summary Shift analysis was used in this chapter to identify those industries which exhibited a relative tendency to locate in rural counties in Michigan during 1950 to 1960. Several manufacturing industries shifted relatively into rural areas. These industries deserve to be studied first to see if the trends can be continued. Electrical and other machinery manufacturing, tran- sportation equipment other than motor vehicles, and food and kindred products manufacturing had the greatest relative shift into counties distant from large urban-industrial centers. Miscellaneous manufacturing and chemicals and allied products had large relative shifts into counties which were not standard metropolitan statistical areas, . but were in Lower Michigan. The dispersion among the rural counties of the -additional employment provided by the various industries was usually the greatest for the industries which had the largest shifts into rural counties. CHAPTER V IMPACT OF PUBLIC AND PRIVATE SERVICES ON EMPLOYMENT GROWTH Introduction The attraction of industry is a competitive matter. As communities, areas, and regions vie for the employment created by industry, they are confronted with the task of improving their locational advantage. . Not only must an area's locational advantage be improved to attract industry, but it must be made favor- able relative to other areas. The question of relative costs and returns is crucial; a favorable Opportunity at a given place may not be exploited because of an existing greater advantage in another location. An essential start- ing point for understanding a region's or an area's poten- tial for growth is a realistic appraisal of relative ad- vantage with regard to input-output access; that is, con- sideration of a location's relative advantage in prOcure- ment of inputs and distribution of output. 71 72 Perloffl presents a scheme of types of regions that can exhibit different growth potentials. He class- ifies regions by (1) good or poor access to basic and in- termediary inputs from external and national sources, (2) good or poor access to basic and intermediary inputs in yhome region, (3) good or poor access to external or na- tional markets, and (4) good or poor access to markets in home region. It is apparent that a location which can be class- ified as good by all four standards is preferable to a place classified as poor. Various regions would be pre- ferred according to the market served and the range of input procurement. However, spatial relationships and such factors as past growth and climate are factors not subject to control by communities. To the extent that a region's location in space and resource endowment determine potential for grow- th, a community will not be able to affect its own growth. However, it is believed that regions and areas can influence their own potential for growth--one way is through the level of public and private services made available. "There are many things that a region can do to enhance its locational advantages, particularly with regard 1Harvey S. Perloff, How a Region Grows: Area De— velopment in the U. 5. Economy, Supplementary Paper No. I7 TNew York: Committee for Economic DevelOpment, 1963), pp. 29-34 0 73 to facilities, as in improving transportation, and major 1 services, such as better education." "More and improved highways, airports, and communications are necessary to develop . . . the region."2 The existence and size of urban-industrial concen- trations is often taken as an indication of the level of private services available. The probability of future growth in industries that ship products to other regions is greatest at points which already have some industrial concen- tration. . . . Manufacturing requires a set of available community facilities and business ser- vices. Where these are well developed, the ser- vices are cheaper when used by several firms rather than a single firm. The industry location process is quite compli- cated, and is different for each industry. Avail- ability of complementary business services is only one of many factors . . . but availability of these services in a place incrgases the probability for industrial growth there. In addition to the industrial framework existing in urban complexes, the concentration of peOple can pro- vide a home market for new industry. Size of city, thus, is often viewed as an indication of potential for growth. The purpose of this chapter is to test; can some of the relative differences in employment growth among lIbid., 29. 2Max F. Jordan and Lloyd D. Bender, An Economic Survey of the Ozark Region, U.S.D.A. Agricultural Economic Report No. 97 (Washington: U. 8. Government Printing 0f- fice, 1966), p. 69. 31bid., 67. 74 counties in a rural area be explained by differences in the level of public and private services available? Analysis of Growth as Related to Public and Private Services The statistical methods of regression and correla- tion are used to analyze the relative importance of certain variables in explaining comparative employment growth, 1950-1960, in some rural areas. The area.--The area studied is the upper portion (all counties above Bay) of the lower peninsula of Mich- igan. Twenty-seven contiguous counties are included in the area, which is predominantly rural. The 1950 census classified the pOpulation in 18 of these counties as entirely rural and the remaining coun- ties were all above 40 percent rural. Only three had cities with pOpulations over 10,000, with the largest being 16,000. Within the context of Perloff's model, this area has relatively poor access to both inputs and markets from external and national sources. The variables.--The dependent variables chosen to be explained were (1) share growth rate of all industries, (2) share growth rate of manufacturing industries, and (3) share growth rate of electrical and other machinery man-- ufacturing. 75 Share growth rate is the rate of growth in employ- . ment due to differences between rates of growth of indus- tries within a single county and rates of growth of the same industries in all counties in the United States. It is simply the regional share component (or competitive ef- fect), for the period 1950 to 1960, divided by the initial employment in 1950. I Computations for the regional share component are illustrated in Appendix A. Table A in Appendix B presents an example of the regional share components. However, there is this difference; the Appendix illustrates the regional share component for regions within Michigan, and, in this chapter, counties (rather than regions) were used. .Also, the entire United States was used as a base, rather than the state of Michigan. The components of employment change by county, for the United States, were published by ‘the U. S. Department of Commerce in Growth Patterns in Employment by County, 1940-1950 and 1950-1960. The question may be asked; did this county's share (If this industry grow faster than did the average of all (J. S. counties' shares? If it did grow faster, its share grtnvth rate will be positive. If its share grew slower, then its share growth rate will be negative. The magni- tude of the share growth rate, then indicates how much better (if positive) or how much worse (if negative) the county performed than did all U. S. counties. 76 The share growth rate for the industry electrical and other machinery manufacturing was calculated for each county (for 1950 to 1960) and used as a dependent variable in the statistical analysis (Table 15). The same calculations can be applied to groups of industries, such as all manufacturing industries. The share growth rate of all manufacturing was calculated for each county from 1950 to 1960 and also used as a dependent variable. I Likewise, the share growth rate may be calculated for all industries for each county. This may be obtained by simply summing the regional share component of the indi- vidual industries and dividing that total by the total all industry employment at the beginning date. The all indus- try1 share growth rate, 1950 to 1960, was also used as a dependent variable. The six explanatory variables used were; (1) per- capita government expenditures, (2) per-capita education expenditures, (3) education completed, (4) tranSportation facilities, (5) percent of population rural farm, and (6) size of largest city. The independent variable, per-capita government expenditures, comprises total general expenditures for all local government in 1957. Not only the county 1Same industry classification as in Table A, Appendix B. 77 Table 15.--Share growth rates, all industries, manufactur- ing industries, and electrical & other machinery manufacturing, 1950—1960, twenty-seven counties in Northern Lower Peninsula.a Share Growth Rates (percent) bAll Electrical and Counties Industries Manufacturing Other Mach. Mfg. Alcona 7.37 189.90 2,433.33 Alpena 19.69 46.37 82.10 Antrim 0.09 142.44 1,110.00 Arenac 5.27 159.12 285.00 Benzie - 6.61 34.85 276.47 Charlevoix 1.30 40.54 1,345.45 Cheboygan - 4.38 32.79 38.23 Crawford 22.28 89.76 -28.57 Emmet -16.76 -40.80 -15.38 Gladwin 12.69 102.56 200.00 Grand Traverse - 5.67 10.84 17.30 Iosco 44.68 0.14 158.82 Kalkaska 1.38 186.71 fi 500.00 Lake 0.54 103.84 163.63 Leelanau 8.60 77.90 170.83 Manistee -12.94 8.35 27.61 .Mason - 4.92 8.40 -95.65 Missaukee - 9.28 40.81 233.33 Montmorency 10.89 63.39 220.00 Ogemaw - 3.46 123.79 23.07 Osceola - 3.76 105.89 304.54 Oscoda 10.48 49.09 133.33 Otsego 7.70 53.78 36.36 Presque Isle 12.64 65.69 454.54 Roscommon 2.18 34.30 45.00 aCalculated from regional share components as presented in Lowell D. Ashby, Growth Patterns in Employment by County 1940- 9 0 nd 19 - 960, Volume 3, U. S. Department of Com- merce (Washington: U. S. Government Printing Office, 1965), pp. 7-1 to 7-28. All industries includes all employment in all thirty-two industries listed in Table A in Appendix. 78 government, but also any municipality, township, school district, and special district expenditures were included. Only utility systems and locally administered employee retirement systems were omitted. Categories of expendi- tures were education, highways, health and hospitals, pub- lic welfare, police, fire protection, natural resources, sewers and sewage disposal, other sanitation, parks and recreation, and other general expenditures.1 Per-capita expenditures for education consisted of direct local government provision and Operation of elemen- tary and high schools. A11 expenditures were included which were administered by local agencies.2 The measure used for education completed was the percent of the county population consisting of males, 25 years and older, with at least four years of high school completed in 1950.3 Transportation facilities included the total num- ber of state and U. S. routes, airports, buslines, common carriers, and railroad lines.4 This variable was included 1U. S. Bureau of the Census, Local Government Fi- nances and Employment in Relation to Pcmulation: 1957, State and Local Government Special Studies No. 45, pp. 127-136. 21bid. 3U. S. Bureau of the Census, U. S. Census of POp- ulation: 1950. 4Michigan Economic Development Department, Research Division, County Economic Data Sheets, 1961. 79 as a simple total because many writers have suggested that, after a firm has chosen an area in which to locate, then available transportation facilities are important in the selection of a specific site. Size of largest city was simply the total pOpula- tion of the largest city or town in the county in 1950.1 The final explanatory variable used was, the per- cent of the county population Classified as rural farm by (the 1950 census. This variable was included to take into account the differing degrees of urbanization which would not be reflected in the size of the single largest city. Regression and correlation analysis For each of the dependent variables, the following statistics were calculated; R2, F-statistic for the overall regression, regression coefficients, beta weights, F- statistics for individual variables, partial correlation coefficients, and R2 deletes. A review of the literature suggests that the fol- lowing relationships should be found: (1) Per-capita government expenditures are expect- eui to be positively related to share growth rates. New and expanding industries require police and fire protec- tirna, access roads, sewerage, water, and other services. 1U. 5. Bureau of the Census, U. 8. Census of Pop- ulation: 1950. 80 The absence of these facilities may seriously deter in- dustry from locating in an area. Thus, areas providing a high level of public services should experience high growth rates relative to those areas with less provision for these services. Although the measure, per-capita expenditures, may tend to obscure some significant differences in adequacy ’ and efficiency (i.e. per—unit costs tend to be higher in A smaller communities) of public services among counties, the measure does provide an indicator which is applicable to all services, and it gauges the level of services avail- able relative to the pOpulation base to be served. (2) An equally straightforward relationship may not exist between employment share growth rates and, ei- ' ther, per-capita expenditures for education, or education completed. To the extent that "good" school systems and availability of an "educated" labor force are desired by industrial firms and influence location decisions, the relationship will be positive. However, since better educated workers are likely to be more mobile, outmigration may be higher in counties which spend more for education and have higher educational levels. Outmigration may sap the most productive workers and.potentia1 entrepeneurs from an area. To the extent 'that a greater amount of education contributes to emigra- ‘tion of the labor force, a negative relationship with em- ployment share growth rates can be expected. 81 (3) Transportation facilities should be positively associated with share growth rates. Counties containing more facilities for transporting raw materials and proces- sed goods can be expected to be preferred by new industry to areas with fewer facilities. In many instances, the presence or absence of a particular transport facility, such as an airport or a railroad terminal, can be the deciding factor in choosing a specific area in which to locate after the general region has been chosen. (4) Size of largest city should be positively asso— ciated with share growth rates, and for percent rural farm the relationship should be negative. An industry does not exist alone, but requires services and markets provided by other firms and industries. Small communities do not have as many tie-in industries as larger communities and cannot be expected to generate as much new and eXpanding industry. Also community facilities and business services are often cheaper when used by many firms rather than a few. Larger communities can be expected to exhibit higher share growth rates because of the existing complex of com- munity facilities and business services. All industry share growth rate Table 16 presents the statistics calculated with the share growth rate of all industries as the dependent variable. 82 Table 16.--Regression and correlation coefficients for all industry share growth rate, Northern Lower Peninsula. Overall regression R2 = .398 Significance level = .086 Partial 2 Independent Regr. Beta Sign. Corr. R Variables Coef. Weights Level Coef. Delete Transportation facilities -1.44 -r.582 .030 -.462 .23 Percent rural farm - .53 -.572 .030 -.462 .23 Per-capita educa- . tion expenditures .27 .539 .145 .321 .33 Per-capita govern- ment expenditures - .18 -.488 .208 -.279 .35 Education completed -2.86 -.285 .244 -.259 .35 Size of largest city - .41 -.140 .526 -.143 .38 The six variables, as a whole, are not very success- ful in explaining differences in share growth rates. The overall R2 is .398; that is, only 39.8 percent of the total Ivariation in share growth rates among counties is explained by the six variables. Furthermore, the significance level is only .086. However, two of the individual variables, transpor- tation facilities and percent rural farm, are of interest because the individual coefficients are significant. 83 The transportation facilities variable is slightly the more important in accounting for variation in share growth rates. This variable does account for some of the variation (partial correlation coefficient = -.462) in share growth rates over that accounted for by the remainder of the independent variables and the mean of the dependent variable. However, the relationship is negative; that is, high growth rates are associated with low number of trans- portation facilities, when the effect of the other variab- les is removed. This relationship is not as expected and it is somewhat difficult to explain. Perhaps the variable (num- ber of transportation facilities) is not an appropriate variable since it does not incorporate relative costs.1 Nevertheless, this relationship does suggest that trans- portation is not important for growth in all industries (as opposed to only manufacturing industries). Growth in employment in activities such Outdoor recreation may not require numerous highways, railroads, and airports. In fact, the reverse may be true--the more primitive the area, the greater the attraction for sightseers and outdoorsmen. The variable, percent rural farm, is the next most important in accounting for variation. The coefficients are negative as eXpected. 1Also, this variable did not include port facilities which may have had some influence in the counties bordering the lakes. 84 The variable, size of largest city, is of partic- ular interest because it is often the basis for designa- tion of growth centers in rural areas. In the correlation analysis, this variable is the least important (i.e. lowest beta weight) in accounting for variation in share growth rate for all industries. Although not significant, the coefficient is negative. The implication is that the ex- isting largest cities with this rural area are not the points at which comparative gains in employment are being ' made in all industries. However, this does not necessarily implicate revision of criteria for designation of growth icenters, since the largest cities may have a greater con- centration of fast growth industries even though they are not increasing their prOportion of them relative to the entire nation. Manufacturinggshare growth rate The same six independent variables were used to analyze share growth rate of manufacturing industries. The six variables accounted for almost 60 percent of the variation, and the overall regression is significant at .004 (Table 17). Only two of the variables are individually signif- icant. They are per-capita government eXpenditures, which was the most important, and percent rural farm. Both re- lationships are positive. This indicates that manufacturing 85 Table l7.--Regression and correlation coefficients for manufacturing industries share growth rate, Northern Lower Peninsula. 2 Overall regression R = .589 Significance level = .004 Partial 2~ Independent Regr. Beta Sign. Corr. R Variables Coef. Weights Level Coef. Delete Per-capita govern- ment expenditures 1.19 .71 .032 .457 .48 Percent rural farm 1.84 ‘ .45 .044 .433 .49 Per-capita educa- tion expenditures -1.11 -.48 .115 .346 .53 Size of largest city -3.08 -.22 .219 -.273 .56 Education completed -6.92 -.15 .452 -.l69 .57 Transportation facilities 1.56 .14 .513 .147 .57 growth is related to expenditures for government. At the same time manufacturing firms increased employment rela- tively more in areas away from the regions larger cities. The coefficient for transportation facilities, although positive, is not significant at any reasonable level. Two eXplanations are possible; first, transporta- tion facilities were not important to the manufacturing firms locating in the area, or, second, all counties had a minimum number of facilities which were adequate for 86 the needs of the firms. The second explanation seems more likely. The manufacturing firms may find relatively un- used facilities in less urbanized counties, preferable to more, but congested, facilities in urbanized areas. Electrical and other machinery manufacturing share growth rate The six variables account for 65 percent of the variation in share growth rate for electrical and other machinery manufacturing, and the overall regression is significant at .001 (Table 18). Per-capita government expenditures is the most important variable in accounting for variation, and the association is positive. The second most important variable, transportation facilities, is also positively associated with share growth rate of electrical and other machinery manufacturing“ When the effect of all other variables is removed, coun- ties with more transportation facilities had higher share growth rates. None of the other individual variables have coef- ficients significantly different from zero at even the .10 level. 87 Table 18.--Regression and correlation coefficients for electrical and other machinery manufacturing share growth rate, Northern Lower Peninsula. 2 Overall regression R = .655 Significance level = .001 Partial 2 Independent Regr. Beta Sign. . Corr. R Variables Coef. Weights Level Coef. Delete Per-capita govern- ment eXpenditures 11.35 .752 .015 .509 .53 Transportation facilities 41.65 .404 .045 .432 .57 Percent rural farm 7.59 .199 .298 .232 .63 Education completed -66.97 -.l6l .382 -.196 .64 Size of largest city -10.88 -.089 .595 -.120 .65 Per-capita educa- tion expenditures 1.33 .064 .814 .053 .65 Importance of the private and public service variables The measures of private and public services were more successful in explaining employment share growth rates for manufacturing industries than for all industries. Per- capita government expenditures was the most important var- iable in explaining differences in share growth rate of manufacturing industries. The strength of the relationship suggests that expenditures for public facilities did 88 contribute greatly to the comparative growth of counties in manufacturing industries. But not for all industries; the relationship with all industry share growth rate was negative and the test that it accounted for any variation above that accounted for by the other variables was not very significant. Per-capita expenditures for education was the third most important variable in accounting for variation in share growth rate in both all industries and manufactur- ing industries. However, in neither case was the reduction in R2 resulting from deletion of this variable very great. The association with all industries was positive, but with manufacturing industries the association was negative. The measure of education completed was in all cases negatively associated with share growth rates, when all other variables were held constant. However, since the significance levels were very low, it is possible that 'education completed did not affect share growth at all. The association of transportation facilities with share growth rates was different for the various industrial groupings. For all industries the association was nega- tive; that is, counties with a large number of transporta— tion facilities had small share growth rates. And, there is 97 percent probability that this observed relationship was not due to chance alone. 89 For manufacturing industries, the association was positive, but there is only 50 percent probability that transportation facilities accounted for any_differences in share growth rates above that accounted for by the other variables. For the single manufacturing industry, electrical and other machinery manufacturing, the association with transportation facilities was pOsitive, and there is 96 percent probability that this variable did account for variation in addition to that of the other variables. The variable, percent of population rural farm, was the second most important in accounting for differ- ences in share growth rates of both all industries and manufacturing industries. However, the relationship was negative for all industries and positive for manufacturing industries. In both cases, there is greater than 95 per-‘ cent probability that this variable accounted for varia- tion above that accounted for by the other variables. The relationship of percent rural farm with elec- trical and other machinery manufacturing share growth rate was also positive. But the probability that it accounted for any of the variation is very low. Size of largest city was, on the whole, the least important of the six variables. In all cases, the degree of association was very low, even though the relationship was always negative. 90 Summary and Conclusions The degree to which the variables were correlated, in most instances, was quite low. The six variables ac- counted for less than 40 percent of the variation in share growth rate of all industry, about 59 percent of the varia- tion in manufacturing share growth rate, and 65 percent of the variation in electrical and other machinery manufac- turing share growth rate. In no instance did a single variable account for more than 26 percent of the variation in share growth rate above that accounted for by the other variables and the mean of the dependent variable. The relationships between the public and private service variables and share growth rates, 1950-1960, were ' not all as expected. For instance, the coefficients for size of largest city were always negative, even though at very low significance levels. Similarly, counties with a high prOportion of rural farm population grew comparatively faster in manufacturing employment, but not in all industry employment. Transportation facilities, while very impor-' tant in the single manufacturing industry of electrical and other machinery manufacturing, were insignificantly related to share growth rate of all manufacturing. Per- capita education expenditures and education completed were not significant and usually inversely related to share growth rates. 91 The results indicate that some public and private service levels are industry specific; that is, affect var- ious industries differently. Counties with high per-capita government expenditures were likely to have had high share growth rates in manufacturing industries, but not in all industries. Any conclusions from the analysis cannot necessarily be applied to development programs for other rural areas. The analysis must be understood in terms of (l) the total picture, or process of change taking place in this rural area, and (2) the spatial location of this area in rela- tion to other areas. These 27 counties were adjusting to a declining dependence on agriculture. Workers leaving agriculture had to find other employment or migrate out of the area. Much of the non-agricultural employment opportunities that existed were provided by manufacturing. The area, in a very real sense, is the hinterland of heavily urbanized Southern Michigan. Location in South- ern Michigan would be advantageous to those industries heavily dependent on: proximity to large home and external markets, economies of large industrial complexes, and ex- ternal sources of inputs. The counties with manufacturing firms and industries, which grew more rapidly than the national average, then, were not those with the largest cities, but those with the relative advantages to be found 92 in the more rural counties. A large part of the addition- al employment in these counties was in food and related products manufacturing, lumber, wood and furniture manufac- turing, electrical and other machinery manufacturing, and miscellaneous manufacturing. Also, educational levels and expenditures were not a drawing point for these manufacturing industries. In fact, the analysis indicated that, at best, higher levels of education did not hinder comparative growth. The level of government services, however, was associated with relative growth. Transportation facilities were similarly important for specific manufacturing indus- tries. The degree of association between public and pri- vate services and share growth rate of all industry was much lower. This was probably due to the various indus- tries reacting differently to the economic stimuli produced 'by changes in the employment base of the area. In total the effects may have tended to cancel. Whether or not the studied associations are strong enough or weak enough to call for changes in rural devel- Opment programs cannot be answered from an understanding ‘of the relationships alone. Something of the costs and expected benefits from public and private services has to 93 be known in order to form good policies and programs for develOpment of rural areas.1 Implications for designation of rural growth centers Growth centers are often designated on the basis of shOpping patterns, commuting patterns, etc., with the re- sult that the largest city is usually regarded as the growth center of the area. However, in the areas studied, comparative gains in employment were not registered in the largest cities, primarily because of the changing economic base of the area. A superior approach to designation of growth cen- ters might involve: (1) Determining what is likely to be the future role, in terms of economic activity, of the entire rural area in relation to external markets and sources of inputs. 1A major step in identifying the costs and benefits associated with new industry in a rural community has been made in several studies, such as; H. A. Wadsworth and J. M. Conrad, Impact of New Industrygon a Rural Community, Agri- cultural EXperiment Station Research Bulletin No. 811'(Laf- ayette, Indiana: Purdue University, 1966), and Max F. Jordan, Rural Industrialization in the Ozarks: Case Study of a New Shirt Plant at Gassville, Ark., U.S.D.A., Agricul- tural Economic Report No. 123 (Washington: U. S. Govern- ment Printing Office, 1967). In the main, these studies concentrated on income from increased direct, business- linked, and consumer-linked employment. More comprehensive and detailed analyses are needed concerning long run spill- over effects on the environment, surrounding communities, and other sectors (especially the public sector) of the community's economy. 94 (2) Identifying the particular subareas and loca- tions within the rural area which are most likely to best serve those emerging economic activities that are likely to provide the economic base of future employment. This would involve identifying locational factors for specific industries and factor availability within subareas of the region. This does not necessarily mean that the largest cities within the area are inappropriate as growth centers. There is the possibility that, if the largest cities in the area were large enough, then these cities might exert sufficient influence on the surrounding area to take up the slack in employment. Programs to create large cities out of the existing cities in the area may be a superior investment policy if those cities will then be able to spearhead sustained growth for the region. More information and experience, concerning the long run effects of alternative programs for growth, is needed in order to select programs with the greatest po- tential for develOpment of the region. CHAPTER VI SUMMARY AND CONCLUSIONS In recent years, there has been widespread concern over declining employment Opportunities in many areas. National, state, county, and local groups have been ac- tively engaged in 'promoting' industrial development for disadvantaged areas. The Public Works and Economic Devel- 0pment Act of 1965 provided for help in the economic de- velOpment of distressed areas. The Economic DevelOpment Administration provided for low-interest long-term loans for the purchase or develOpment of land and facilities for industrial usage. The Michigan Industrial Development Revenue Bond Act of 1963 authorized municipalities to issue revenue bonds in order to acquire industrial facilities for the purpose of leasing to industrial firms. Many local communities have formed industrial development corporations for the sole purpose of contacting and procuring new industry. The President's National Advisory Commission on Rural Poverty has pointed out the consequences of the lack of employment Opportunities in rural areas and has recom- mended efforts to increase jobs in rural areas. 95 96 The purpose of this study has been to investigate three major questions. First, what locational factors have been important to firms locating in rural areas? Second, on the basis of past growth, which industries are the most likely to be attracted to rural'areas in Michigan? Third, what relationships have existed between comparative employment growth and levels of public and private services? Several surveys on locational factors of industry in rural areas were examined in order to gain some idea of the relative importance of various locational factors, with particular emphasis on those factors which can be used as inducements. The shift-share technique was used to point out those industries which demonstrated relative employment shifts favorable to rural areas in Michigan from 1950 to 1960. Statistical procedures of regression and correlation analysis were used to investigate the asso— ciation between levels of public and private services available and comparative employment growth rates in a 27 county rural area in Michigan from 1950 to 1960. Locational Factors Recent literature on factors influencing the loca- tion of industry in rural areas was reviewed for manager rankings of locational factors. Surveys of rural areas in Southern Indiana and in South Central Louisiana were relied on heavily with supplementary references to a survey in Michigan and a national study. 97 Four locational factors were revealed to be most important in making location decisions. These were mar~ kets, raw materials, labor, and transportation. The rela- tive ranking of these four factors depended on the industry represented and on the individual firm. The remaining factors, utilities, taxes and legis- lation, personal preferences, and community attitudes and inducements were of much less importance. However, they are to some degree manageable by local communities. Util- ities and taxes and local legislation were the most impor- tant of these factors, but often played only a minor role in the selection of a specific site because many communities were similar with respect to these two factors. Much evidence was found that quite often plant sites have been selected on the basis of prOprietor's homef town, previous business connection, or readily available plant facilities. This was particularly true of small firms. Community attitudes and offered inducements to new industry had very little influence on location decisions. In most instances, even actions within the community to discourage plant location there were disregarded. Community offered inducements often used were: community financing arrangements, tax exemptions or con- cessions, community owned or held plant sites and build- ings, and research concerning information on labor supply and general facilities available. 98 There was evidence that communities attached too much importance to the role that inducements play in at- tracting firms. Economic inducements were often extended at a cost to the community and accepted by the firm when the firm had not even considered the inducement while (making the location decision. However, many managers did think highly of indus- trial ambassadors who could show them specific site alter- natives, provide documentation of land-use planning, and make available information on labor supply, transportation facilities, complimentary business establishments, and other general information. The overall importance of a particular factor and the aspect of the factor emphasized was often different for each of the areas surveyed. This follows from the differing characteristics and spatial location of the areas surveyed. Firms located in an area for the relative ad- vantage that was to be obtained in that area, and manager rankings of factors reflected this. A study of the movement of industry in Michigan indicated that in the short run, at least, the chances of attaining additional jobs in a small community is greater for those communities which had a recent favorable move- ment of industry. Also, there is a resistance to estab- lishing plants at places other than the existing facilities. 99 Recent industrial growth and present production patterns within a community are essential starting points for un- derstanding the growth potential of that community. Shifts in Industrial Employment The shift-share technique was used to identify those industries for which employment shifted comparative- ly into rural areas in Michigan from 1950 to 1960. Those industries for which rural areas increased their share relative to the rest of the state were considered to have found an advantage in rural areas relative to the rest of. the state. If this relative advantage persists in rural areas for the industries identified as having shifted com- paratively into rural areas, then these industries are ones mostly likely to be attracted by rural community efforts. The manufacturing industries electrical and other machinery, transportation equipment other than motor ve- hicles, and food and kindred products had the greatest relative shifts into rural counties in upper Michigan. Agriculture had the largest shift out of this area. Miscellaneous manufacturing and chemicals and allied products had large relative shifts into lower Mich- igan counties which were not standard metropolitan statis- tical areas. 100 The dispersion among the rural counties of the additional employment provided by the various industries was usually the greatest for the industries which had the largest shifts into rural areas. These industries should be examined first by rural communities desiring additional industrial employment. Public and Private Services as Related to Growth The statistical methods of regression and correla- tion were used to test whether some of the differences in - comparative growth rates among some rural counties in Mich- igan can be explained by differences in the level of public and private services available. The dependent variables were generated with the shift—share technique of analysis. They were (1) share growth rate of all industries for each county (1950-1960), (2) share growth rate of manufacturing industries (1950-1960), and (3) share growth rate of elec- trical and other machinery manufacturing (1950-1960). The variable share growth rate was used to measure how well a particular county did in an industry (or group of indus- tries) relative to how well all counties in this area did in this industry (or group of industries). The six explanatory variables used were, (1) per- capita government expenditures for 1957, (2) per-capita education expenditures for 1957, (3) education completed, 101 (4) transportation facilities, (5) percent of county pOp- ulation rural farm, and (6) size of largest city. The counties studied were 27 contiguous rural coun- ties in the upper portion of Michigan's Lower Peninsula. Only three had cities with a pOpulation over 10,000 with the largest being 16,000. The degree to which the variables were correlated, in most instances, was quite low. The six eXplanatory variables accounted for less than 40 percent of the varia— tion in share growth rate of all industries, about 59 per- cent of the variation in manufacturing share growth rate, and 65 percent of the variation in electrical and other machinery manufacturing share growth rate. In no instance did a single variable account for more than 26 percent of the variation above that accounted for by the other varia- bles and the mean of the dependent variable. Of the six explanatory variables, per-capita gov- ernment expenditures was the most important variable in explaining employment share growth rate of manufacturing industries. But it was relatively unimportant for share growth rate of all industries. Per-capita expenditures for education was the third most important variable in accounting for variation in share growth rate in both all industries and manufacturing industries. However, in neither case did it eXplain much 102 variation, and the association with manufacturing was neg- ative. In all cases the coefficients for the measure of education completed were negative with very low signif- icance levels. A very strong negative relationship was found be- tween number of transportation facilities and share growth rate of all industries. This unexpected result may have been a consequence of the fact that number of transporta- tion facilities does not include quality and relative costs. However, the association with electrical and other machinery manufacturing was positive. Size of largest city was, on the whole, the least important of the six variables studied. The variable, percent of pOpulation rural farm, was the second most important in accounting for differences in share growth rates of both all industries and manufac- turing industries. However, the relationship was negative for all industries and positive for manufacturing industries. The relationships were not all as would be expected and the results indicated that some public and private service levels are industry specific, that is affect var- ious industries differently. The relationships found can be understood only in terms of the total picture or process of change taking place in this rural area, and in terms of the spatial lo- cation of this area in relation to other areas. A large 103 part of the 1950 employment in these 27 counties was depen- dent on agricultural industries. The tremendous decline in agricultural employment during the decade precipitated both migration out of the area and occupational changes to non-agricultural industries. Much of the additional non- agricultural employment was found in manufacturing indus- tries--primari1y food and related products, lumber, wood and furniture, electrical and other machinery, and miscel- laneous manufacturing. The high share growth rates of these industries tended to be registered in the more rural counties. Education of the labor force was probably not im- portant to these industries or, more likely, was a trade- off for the lower wages associated with less education. Availability of transportation facilities was not important as an explanation for all manufacturing share growth rates, but was important for one specific manufac- turing industry. The level of government services was important for the manufacturing industries. Less success was attained in explaining the share growth rate of all industry. This was probably due to the various industries reacting differently to the changing employment base of the area, and individual effects cancelling. 104 This study has dealt with several aspects of how ‘ communities may induce industries to locate in a rural area. Further consideration needs to be given to actual costs and returns to a community of the additional indus- try. Questions of vital concern to communities making concessions to firms are; will the industry pay its way, and what are the expected returns from the location in- ducement offered. More detailed studies need to be done on the total costs and returns associated with industrial development in rural communities. All direct and indirect revenue-generating and expenditure-generating effects of the project, especially those for the public services sector, must be ascertained. Similarly, continued attention at the regional level should be focused on determining apprOpriate policies and programs for promoting industrialization of declining rural areas. For instance, the effectiveness of "growth centers" as a policy instrument and the methods of desig- nating growth centers in declining rural areas warrant further research. Some evidence was found in this study that the largest cities, which would normally be designated as the growth centers, may not have the potential for in- dustrial employment growth in a declining rural area un- less, perhaps, they can be brought to a possible threshold size. 105 These are only a few of the many unknowns concern- ing industrialization of declining rural areas. HOpefully, with more work and experience the problems can be dealt with effectively. APPENDIX A COMPUTATION OF SHIFT-SHARE ANALYSIS APPENDIX A COMPUTATION OF SHIFT-SHARE ANALYSIS1 Shift analysis is a simple analytical approach based on a study of three main components of area growth: that part attributable to state growth, that part attri- butable to rates of growth of the mix of industries in the area that are greater or less than the state rates of growth of all industries, and that part due to differences be- tween rates of growth of industries within a single area and rates of growth of the same industries in other areas. Symbolically this may be expressed as: = gi. + k.. + c.. d.. 1] J - 13 13 where: gi. J, the state growth element for industry i in area j. kij the industry-mix effect for industry i in area j. 1Lowell D. Ashby, Regional Change in a National Setting, U. S. Dept. Com. Off. Bus. Econ. Staff Working - Paper No. 7, 1964; and Growth Patterns in Employment by County, U. S. Dept. of Com., Off. Bus. Econ., 1965, p. xii. 107 108 cij = the area competitive effect (sometimes called regional share component) for industry 1 in area j. dij = the absolute change in employment between two points in time for industry i in area j (that is, E1960 - E1950). For example, considering electrical and other ma- _ chinery manufacturing (industry i) in upper Michigan (area j) between 1950 and 1960, the following applies: Year Employment in upper Michigan 1950 . . . . . 2,471 1960 . . . . . 5,871 absolute change 3,400 Now, the state overall rate of growth between 1950 and 1960 was 0.1399. For electrical and other machinery manufactur- ing it was 0.4130. And for electrical and other machinery manufacturing in upper Michigan it was 1.3759. Letting: r the state overall rate of growth (0.1399) ri = the state rate of growth in indus- try 1 (0.4130) rij = the rate of growth of electrical and other machinery manufacturing (industry i) in upper Michigan (1.3759) 109 with: E1950 being employment in 1950 (2,471) E1960 being employment in 1960 (3,400) Then: gij = E1950 x r = 2,471 x .1399 = 345 ij E1950 X (ri‘r) = 2,471 x (.413o-.1399) = 675 cij E1950 ' ij = 2,471 x (1.3759-.4130) = 2,380 dij E1950 ‘ E1960 5,871 - 2,471 = 3,400 Finally, we have: d.. = g.. + k.. + c.. 13 ij ij 1] 3,400 = 345 + 675 + 2,380 (These figures are presented in Table A, p. 111) This example was worked out using the state rate of growth as a base. The same procedure can be used to analyze shifts among various aggregations of space; for example, regions within the United States, using the national rate 1 as a base. 1The U. S. Department of Commerce published the data generated for all regions, states, and counties in the U. S. using the national growth rate as a base. Lowell D. Ashby, Growth Patterns in Employment By County, 1940- 1950 and 1950—1960, U. S. Department of Commerce TWashifig- ton: U. S. Government Printing Office, 1965). APPENDIX B EMPLOYMENT AND COMPONENTS OF EMPLOYMENT CHANGE, 1950-1960, AREAS 1, 2, and 3 Table A.--Employment and components of employment change, 1950 to 1960, 111 Area l.a Employment in State Industrial Regional Total 1950 1960 Growth Mix Share Change Agriculture 3h,722 lb,350 h,857 -19,573 -5,656 ~20,372 Forestry & fisheries l,h2l 939 198 - 606 - 7h - h82 Mining 113979 11,777 1,676 " 1:833 " 145 " 202 Contract construction 11,373 13,882 1,591 - 905 1,823 2,509 Food & kindred pro- dUCtS mfg. 2:563 3,976 359 h53 601 1:813 Textile mill products 176 137 25 - 72 8 - 39 Apparel mfg. 1,581 1,722 221 - 12 - 68 1L1 Lumber, wood.& furn. 13,231 8,h9h 1,851 — 3,68h -2,90h - h,737 Printing & publishing l,h81 2,299 207 502 109 818 Chemicals & allied 2,386 2,975 33h h77 - 222 589 Electrical & other machinery mfg. 2,h71 5,871 3h5 675 2,380 3,h00 Motor vehicles & equipment mfg. 3,213 1,h68 h50 - 1,111 -1,08h - 1,7h5 Other transportation equipment mfg. 598 2,005 8b 652 671 1,h07 Other & misc. mfg. 10,620 12,86h l,h86 l,h98 - 7h0 2,2bh Railroads & railway express h,550 3,62h 636 - 1,660 98 - 926 Trucking & warehousing 1,610 1,983 225 333 - 185 373 Other transportation 3,309 2,h86 h62 - 1,110 175 - 823 Communications 2,189 2,3h1 306 - 12 1&2 152 Utilities & sanitary Service 2,515 2,7hh 352 - 238 115 229 Wholesale trade h,196 h,859 587 322 - 2h6 663 Food.& dairy products stores 6,79h 6,171 951 - 931 - 6&3 - 623 Eating & drinking places 6,05h 7,027 8h7 - 162 288 973 Other retail trade 16,h03 19,763 2,295 815 250 3,360 Finance, insurance & real estate 2,880 h,301 h03 720 298 l,h21 Hotels & other personal services 6,h55 6,771 903 - 670 83 316 Private households 3,302 5,222 h62 936 522 1,920 Business & repair ' services h,325 3,317 605 197 -1,810 - 1,008 Entertainment, recrea- tion services 1,512 1,219 212 - 381 - 12h — 293 Medical & other pro- fessional services 16,307 2b,087 2,281 9,551 -u,052 7,780 Public administration 7,069 8,817 989 1,026 - 267 1,7h8 Armed forces 8&6 5,932 118 13h h,83h 5,086 Industryynot reported 3,006 h,27h h21 5,075 -h,228 1,268 Total 191,137 197,597 25:739 - 9,598 -10,585 5,560 Table A.--(Continued) Area 2. 112 \ Employment in State Industrial Regional Total 1950 1960 Growth Mix Share Change_ Agriculture 83,959 50,153 11,7h5 -h7,330 1,779 -33,806 Forestry & fisheries h17 28b 58 - 178 - 13 - 133 Mining 1,888 1,785 268 - 289 - 78 - 103 Contract construction 21,571 25,238 3,018 - 1,717 2,366 3,667 Food & kindred pro- ducts mfg. 15,792 18,082 2,209 2,792 -2,711 2,290 Textile mill products 1,28h 95b 180 - 525 15 - 330 Apparel mfg. 1,h17 1,785 198 - 10 lhO 328 Lumber, wood & furn. 7,920 9,090 1,108 - 2,205 2,267 1,170 Printing a publishing h,S75 6,508 6&0 1,5h9 - 256 1,933 Chemicals 8 allied 7,312 12,238 1,023 1,u61 2,h38 8,922 Electrical & other machinery mfg. 30,h06 h3,025 h,25h 8,30h 61 12,619 Motor vehicles & equipment mfg. 27,676 27,115 3,872 - 9,570 5,137 - 561 Other transportation 0 equipment mfg. 2,h69 7,027 3&5 2,69h 1,519 h,558 Other & misc. mfg. hh,7l9 61,390 6,256 6,310 h,105 16,671 Railroads & railway express 7,802 5,816 1,035 - 2,700 - 321 - 1,986 Trucking & warehousing 5,3hl 7,501 7h? 1,103 310 2,160 Other transportation 2,b88 2,883 3b8 - 835 882 - 5 Communications h,360 h,5h2 610 - 23 - b05 182 Utilities & sanitary service 5,137 6,0h2 719 - h85 671 905 Wholesale trade 9,615 11,55h 1,3h5 738 - 1th 1,939 Food & dairy products ' stores 12,037 13,116 1,68h - 1,6h9 1,0hh 1,079 Eating &-drinking places 10,161 13,b3h 1,h22 - 272 2,125 3,273 Other retail trade 35,h55 h5,565 h,960 1,761 3,389 10,110 Finance, insurance & real estate 7,880 12,002 1,087 1,869 1,606 8,522 Hotels & other personal services 9,900 11,308 1,386 - 1,028- 1,050 l,h08 Private households 8,h8h 13,273 1,187 2,h05 1,197 h,789 Business & repair services 9,123 8,213 1,276 h15 -2,601 - 901 Entertainment, recrea- tion services 2,960 2,539 hlh — 7&5 '- 9O - h21 Medical & other pro- fessional services 32,126 56,303 h,h9h 18,816 867 2h,177 Public administration 10,118 13,576 1,h16 1,h67 575 3,h58 Armed forces h,377 l,h17 612 692 -h,26h - 2,960 Industry not reported, 7,093 13,918, .992 11,977 -6.188 6,821 Total b35,062 506,828 60,86h - 5,210 16,112 71,766 Table A.--(Continued) Area 3 113 Employment in State Industrial Regional Total 1950 1960 Growth Mix Share Change Agriculture b1,188 27,617 5,763 -23,218 3,888 -13,571 Forestry & fisheries 32h 319 LS - 138 88 5 Mining 1,676 1,778 238 - 255 123 102 Contract construction 85,h7h 86,bb2 11,958 - 6,803 —h,187 968 Food & kindred pro- ducts mfg. 29,022 h0,32h h,060 5,132 2,110 11,302 Textile mill products 3,b03 2,862 L76 - 1,393 - 2h 9L1 Apparel mfg. 8,580 5,115 681 - 3b - 72 S35 Lumber, wood & furn. 18,810 16,8bh 2,632 - 5,237 639 - 1,966 Printing a publishing 28,88b 36,935 3,h81 8,b23 187 12,051 Chemicals & allied 20,838 25,700 2,915 h,163 -2,216 b,862 Electrical & other machinery mfg. 93,631 129,858 13,099 25,572 -2,bbb 36,227 Motor vehicles & equipment mfg. bbh,07h 3L8,580 62,126 -153,561 -b,059 -95,b9b Other transportatiOn equipment'mfg. 5,576 10,2b9 780 6,083 —2,190 b,673 Other 8 misc. mfg. 151,60h 190,8bb 21,211 21,393 —3,36b 39,280 Railroads & railway express 21,517 16,902 3,010 — 7,8b9 22b - b,615 Trucking & warehousing 22,3h0 29,956 3,125 h,615 - 12b 7,616 Other transportation 20,680 16,327 2,893 - 6,980 - 306 - 8,353 Communications 21,682 25,1h9 3,033 - 115 5b9 3,867 Utilities & sanitary service 27,626 28,092 3,865 - 2,612 - 787 L66 Wholesale trade 52,h98 6h,262 7,3hh h,032 388 11,76h Food & dairy products stores 55,010 Sb,765’ 7,696 - 7,53h - b0? 2&5 Eating & drinking places 50,600 53,908 7,079 - 1,356 -2,b15 3,308 Other retail trade 162,h2h 189,598 22,722 8,071 —3,623 27,170 Finance, insurance & real estate 5h,076 73,253 7,565 13,51h -1,902 19,177 Hotels & other * personal services 51,522 52,2h7 7,208 ,- 5,389 —1,13h 725 Private households 32,917 b5,137 h,60h 9,331 -1,715 12,220 Business & repair ' ‘ services 38,236 b9,738 5,389 1,780 b,b13 11,502 Entertainment, recrea- tion services 15,b3h 13,923 2,160 - 3,885 21b - 1,511 Medical 8 other pro- fessional services lb2,762 289,530 19,971 83,61b 3,183 106,768 Public administration 56,615 72,bhb 7,920 8,215 - 306 15,829 Armed forces 5,225 6,212 731 826 - 570 987 Industry not reported 21,575 71,397 3,018 36,h28 10,376 b9,822 Total 1,777,8231LQEL900 258:715 1h,87l -5,509 258,077 aCalculated from Lowell D. Ashby, Growth Patterns in Employment by County, 19bO-1950 and 1950-1960, V01. 3, U.S. Department of Commerce (Washington: U.S. Government Printing Office, 1965), pp. 7-1 to 7-28. BIBLIOGRAPHY Ashby, Lowell D. Growth Patterns in Employment by County, ‘ 1940-1950 and 1950-1960, Vol. 3. U. S. Department of Commerce. Washington: U. S. Government Print- ing Office, 1965. Brann, Paul. Actions to Influence Location Decision of Firms. Papers presented at the Conference Area Development. Athens: University of Georgia, 1962. Fuchs, Victor R. Changes in the Location of Manufacturing in the United States Since 1929. Ann Arbor: Yale University Press, 1962. Greenhut, Melvin L. Plant Location in Theory and Prac- tice. Chapel Hill: University of North Carolina Press, 1956. Jordan, Max F. and Bender, Lloyd D. An Economic Survey of the Ozark Region. U.S.D.A. Agricultural Economic Report No. 97. Washington: U. S. Government Printing Office, 1966. McLaughlin, Glenn E. and Robock, Stefan. Why Industgy Moves South. Washington: National Planning Association, 1949. Michigan Economic Development Department. Research Divi- sion. County_Economic Data Sheet, 1961. Michigan State University. Bureau of Business and Eco- nomic Research. Michigan Statistical Abstract. 6th ed. East Lansing, 1966. Mueller, Eva. Location Decisions and Industrial Mobility in Michigan, 1961. Ann Arbor: UniVersity of Michigan, 1961. National Advisory Commission on Rural Poverty. The Pepple Left Behind. A Report by the President's National AdviSory Commission on Rural Poverty. Washington: U. 8. Government Printing Office, 1967. 114 115 Perloff, Harvey S. How a Region Grows: Area Develppment in the U. S. Economy. Supplement Paper No. 17. New York: Committee for Economic Development, 1963. PolOpolus, Leo and Williams, Robert W. Factors Influencing the Location of Manufacturing Plants in a Rural Area. Louisiana Agricultural Experiment Station, D.A.E. Research Report No. 324, 1963. Smith, Harold F. A Suggested Prpgram for Financing Area Development in Michigan. A Report of the W. E. Upjohn Institute for Employment Research. Kala- mazoo, 1960. U. 8. Bureau of the Census. Local Government Finances and Employment in Relation to the Population, 1957. State and Local Government Special Studies No. 45. U. 8. Bureau of the Census. U. S. Census of Population, 1950. Wallace, L. T. Factors Affecting Industrial Location in Southern Indiana, 1955-1958. Indiana Agricultural EXperiment Station Research Bulletin No. 724. Laf- ayette, 1961. ~ _Weber, Alfred. Theory of Location of Industries. Chicago: University of Chicago Press, 1928.'