. . . . . 224575. 13...: v w.” “nun“... ”HESlS lllllllllllllllllllllll 31293 00881 2673 This is to certify that the thesis entitled Economic Impacts of Employment Shifts in Michigan's Metro and Nonmetro Areas, 1978 to 1987 presented by Betty Laverne King Nordeng has been accepted towards fulfillment of the requirements for M;A. degree in Geography Major professor Date 7 ' 20" 93 0.7639 MS U is an Affirmative Action/Equal Opportunity Institution __-—_..——_. ,*4 LIBRARY Mlchlgan State University IN RETURN BOX to remove this checkout from your record. PLACE or before date due. TO AVOID FINES return on J J Ll MSU Is An Affirmative ActioNEquel Opportunity institution chnG-p Economic Impacts of Employment Shifts in Michigan's Metro and NOnnetro Areas, 1978 to 1987 BY Betty Laverne King Nordeng A Thesis Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Geography 1993 ABSTRACT Economic Impacts of Employment Shifts in Michigan's Metro and Nonmetro Areas: 1978-1987 BY Betty Laverne King Nordeng Trends in industrial transformation in Michigan’s metropolitan and nonmetropolitan counties between 1978 and 1987 are examined. The spatially and temporally varying associations of employment specialization in manufacturing, wholesale and retail trade, and FIRE with real income per capita is also analyzed. Results indicate that a sectoral shift from manufacturing to services is occurring, but Michigan still remains more dependendent on manufacturing than the nation as a whole. The industrial structure varied between metropolitan and nonmetropolitan counties. Counties specialized in manufacturing employment were predominantly metropolitan or nonmetropolitan adjacent, while counties specialized in wholesale and retail trade employment were predominantly nonmetropolitan. One half of the metropolitan counties were specialized in FIRE employment, while only a quarter of the nonmetropolitan counties were. Regression analysis indicated that specialization in manufacturing and FIRE employment associated with higher income per capita, while specialization in wholesale and retail trade employment associated with lower income per capita. Hi to my parents, who taught me that anything wOrth having was worth working for iv ACKNOWLEDGEMENTS I would like to thank my advisor, Dr. Sharmistha Bagchi-Sen, for her constructive criticism, guidance and support throughout my graduate studies at Michigan State University. I'd also like to thank my committee members, Dr. David Campbell and Dr. Gary Manson, for their comments and insights; and Mike Lipsey and Ellen White for their computer assistance. I'd like to thank my husband for his patience and support, and my children for their constant distractions and enthusiasm for life. Without them, life would not be near as meaningful, nor as fun. TABLE OF CONTENTS List Of TableSOOCOOOOOOOOOOOOO00.000.000.00...O ..... 0 Vi List Of FigureSOCOOOOOCOOOOOOOO ........ OOOOOOOOOOOOOO Viii CHAPTERS I. INTRODUCTION..................................... 1.1 Industrial Restructuring in the United States 1.2 Gaps in the Literature.................... 1.3 Purpose................................... 1.4 DataOOOOOOOOO0.0000IOOOOOOOOOOO...00...... \OWO‘KJH II. LITERATURE REVIEW................................ 13 2.1 Industrial Shifts in the United States.... 14 2.2 Socio-Economic Consequences of Industrial Restructuring............................. 18 III. ECONOMIC TRANSFORMATION IN MICHIGAN.............. 23 3.1 Industrial Structure in Michigan.......... 27 3.2 Industrial Diversification................ 39 3.3 Employment Specialization................. 42 3.4 Summary................................... 50 IV. ANALYSISOOOOIOOO00.000.000.00.0000000000000000000 53 4.1 Spatial Variation......................... 54 4.2 Temporal Variation........................ 55 4 O 3 Resu1ts. . O O O O O O I O O O O O I O O O O O O O O O O O O O O I O O O O O 56 v. CONCLUSIONS. 0 O O O O O O ..... O O O O I O O O O O I O O O O O O O O O O O O O O 61 APPENDICES...‘OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 66 BIBLIOGRAPHYOOOOOO0.0.0.000...OOOOOOOOOOOOOOOOOOOOOOOO 76 LIST OF TABLES Metropolitan-Nonmetropolitan Variation in Per Capita Income, 1978 and 1987.................. Change in Real Per Capita Income by County, 1978 to 1987......OOOOOOOOOOOOOOOOOOOOOOOIOOOOOOOOO Pearson Correlation Coefficients for Income and Earnings per Capita in Michigan, 1978......... Pearson Correlation Coefficients for Income and Earnings per Capita in Michigan, 1987......... Pearson Correlation Coefficients for Income and Earnings per Capita in Nonmetropolitan Counties in MiChigan' 1978.0....OOOOOOOIOOOOOOOOOOOOOOO Pearson Correlation Coefficients for IncOme and Earnings per Capita in Nonmetropolitan Counties in MiChigan, 1987..0....OOOOOOOOOOOOOOOOOOOOOO Percent Distribution of Employment and Estab- lishments for the United States and Michigan, 1978 and 1987.00.00...OOOOOOOOOIOOOOOOOOOOOOOO Percent Distribution of Employment by Industry Type, 1978 and1987.00..OOOOOOOOOOOOOOOOOOOOOO Percent Distribution of Establishments by Industry Type, 1978 and 1987.................. Change in Employment by County Type .......... Change in Establishment by County Type........ Change in High Tech Employment and the Number of High Tech Establishments in Michigan Between 1978 and 19870....OOOOOOOOOOOOOOOOOOOOO.COO... High Tech Manufacturing as a Percentage of All Manufacturing, 1978 and 1987.................. Employment Change in Selected SIC Categories Change in Relative Entropy Between............ 23 24 26 26 26 27 28 30 31 34 35 36 36 38 41 Geographic Distribution of Counties with Location Quotients Greater than One in 1978... Geographic Distribution of Counties with Location Quotients Greater than One in 1987... Change in the Average Location Quotients...... Metropolitan-Nonmetropolitan Variation in Employment Specialization, 1978 and 1987...... Pearson Correlation Matrix for 1978 Equations Pearson Correlation Matrix for 1987 Equations Metro-Nonmetro Variation in the Association Between Employment Specialization and Income per capitaOOOO000......OOOOOOOOOOOOOOOOOOOOOOO Temporal Variation in the Association Between Employment Specialization and Real Income per CapitaOOOOOOIOOOCOO...OOOOOOOOOOOOOOOOOOOOOOOO Appendix A Change in Employment Entropy by County Appendix B Change in Establishment Entropy by County Appendix C (a) Change in Location Quotients for Manufacturing by County............ Appendix C (b) Change in Location Quotients for Wholesale and Retail Trade by County Appendix C (c) Change in Location Quotients for FIRE by County.......................... 48 48 49 50 54 54 57 59 66 68 7O 72 74 ‘fifi LIST OF FIGURES Figure Metro-Nonmetro Classification of Michigan Counties Location Quotients: Wholesale and Retail Trade.... Location Quotients: Manufacturing................. Location Quotients: Finance, Insurance, and Real EstateOOIOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO. 10 44 46 47 CHAPTER ONE INTRODUCTION Industrial transformation is a process involving changes in industrial structure and output. In the United States, this transformation is often defined as a reorganization within manufacturing industries and a growth in service industries. Improvements in transportation, communication, production technology, and the development of a global reservoir of employment have allowed manufacturing firms to spatially reorganize modes of production (Bradbury, 1985: Young, 1986). Industrial transformation in the United States has involved four types of spatial changes: the movement of industries from city to suburb; Industrial Heartland to Sunbelt; metropolitan to nonmetropolitan areas: and relocation of industries to foreign countries (Noyelle and Stanback, 1983; Young, 1986: Haynes and Machunda, 1987; Markusen and Carlson, 1989; Esparza, 1990: Moriarty, 1991: Ray, 1992). It was anticipated that regional shifts in manufacturing within the United States would result in a reduction of interregional inequality; however that has not occurred (Glickman and Glasmeier, 1989). Instead, the process of industrial transformation has increased intermetropolitan inequality within and between U.S. census regions and income inequality between U.S. metropolitan and nonmetropolitan areas (Amos, 1989; Barancik, 1990; Angel and Mitchell, 1991: Deavers, 1991). l 2 In the Industrial Heartland the transformation to a service economy and the spatial reorganization of production has resulted in manufacturing employment losses at a much greater scale than anywhere else in the United States (Markusen and Carlson, 1989; Connaughton and Madsen, 1990). Job loss in the Great Lakes region accounted for half of all jobs lost in the nation between 1979 and 1986 (Markusen and Carlson, 1989). Between 1963 and 1986, this region also experienced an overall decline in manufacturing output and a below average annual growth rate in total Gross Regional Product (Connaughton and Madsen, 1990). Service sector output increased between 1963 and 1986 in the Industrial Heartland, but did not offset the loss in manufacturing output especially in Illinois, Indiana, Michigan, Ohio, and Wisconsin (Connaughton and Madsen, 1990; Goe, 1990). In this study, trends in industrial transformation in Michigan's metropolitan and nonmetropolitan counties between 1978 and 1987 are examined; in addition, the spatial and temporal associations of sectoral employment specialization ‘with economic indicators are analyzed. The following sections include a brief review of industrial restructuring and its impact on the labor market in the United States, gaps in the literature, a statement of purpose, and data description. 3 1.1 Industrial Restructuring in the united States In the United States corporate profits began falling in the mid 1960's, as rising taxes and labor costs increased production costs, while increasing global competition decreased corporations' ability to increase prices (Harrison and Bluestone, 1990). Corporations, especially those in manufacturing, introduced a variety of cost-cutting strategies to improve profitability. These strategies included automation, the relocation of production to low- wage areas within the United States and abroad, flexibility such as subcontracting, substituting part-time and temporary labor for fulltime labor, demanding concessions from their workforce, and finally abandoning core businesses such as steel, in favor of newer enterprises and more speculative ventures (Kutscher and Personick, 1986; Harrison and Bluestone, 1990; Moriarty, 1991). Many of these strategies reduced the bargaining and political power of unions, increased work fragmentation and underemployment, and reestablished a dual labor market, whereby two workers could ‘be paid vastly different wages for the same job. New employees in the auto parts industry, for example, could be paid 45% less than the former base wage (Harrison and Bluestone, 1990). As a result, real hourly wage rates declined for every industrial sector except mining (Harrison and Bluestone, 1990). Although wage rates dropped for every industrial sector, wage disparity increased within sectors (Harrison 4 and Bluestone, 1990; Levy and Murnane, 1992). In the financial sector, white-collar professionals received higher wages than semi-skilled and unskilled workers (Harrison and Bluestone, 1990). In manufacturing, semiskilled jobs declined faster than total manufacturing employment, reducing the demand and therefore the wages of individuals with lower education levels (Levy and Murnane, 1992). Average wages in industries such as producer services and high-tech manufacturing increased, while average wages in industries such as consumer services and retail trade decreased (Grubb and Wilson, 1992). Wages increased sharply for college graduates relative to high school graduates, and the wage differential between high school graduates and dropouts also increased. These factors increased inequality not only between, but also within age, race, and sex groups (Harrison and Bluestone, 1990; Bound and Johnson, 1992; Grubb and Wilson, 1992; Levy and Murnane, 1992). Since wages are the principal income component for most of the population, these changes in wages were a key factor in the ’recent rise in income and consumption inequality (Harrison and Bluestone, 1990; Cutler and Katz, 1992), It has been suggested that the dramatic increase in the size of the workforce due to the large number of baby- boomers, women, and minorities entering the work force has contributed to the decrease in wages (Harrison and Bluestone, 1990; Levy and Murnane, 1992). While demographic changes are a factor, the increase in the size of the 5 workforce was offset by a large increase in jobs (Harrison and Bluestone, 1992). Unfortunately, the majority of jobs created were low wage service jobs. It was the proliferation of these types of jobs coupled with a decrease in high wage manufacturing jobs that contributed most to increasing inequality (Harrison and Bluestone, 1990: Levy and Murnane, 1992: Ray, 1992). If the majority of jobs created had been higher wage jobs, the increase in inequality may not have occurred. Job creation alone is not enough; the quality of jobs created is also important. (Harrison and Bluestone, 1990: Ray, 1992). In Michigan, much of the attention has been focused on Detroit, Flint and the auto industry, which have and continue to experience serious employment losses. Other places and industries in Michigan, however, have also suffered from employment losses. Between 1980 and 1986 double digit unemployment occurred in metropolitan and nonmetropolitan counties (Redman and Rowley, 1989). Unemployment was higher in nonmetropolitan counties and the 'income gap between metropolitan and nonmetropolitan areas increased. Earnings per worker, which includes wages and net proprietor income, decreased by 8% for metropolitan counties, and 12% for nonmetropolitan counties between 1979 and 1986 (Redman and Rowley, 19893'Deavers, 1991). 1.2 Gaps in the Literature Much of the literature regarding economic transformation in the United States has examined the structural changes and their impacts at a regional level (Jones and Kodras, 1990: Markusen and Carlson, 1989: O’hUallachain, 1985) and/or focused on metropolitan regions (Goe, 1990; Esparza, 1990; Kellerman, 1985: Mead, 1991; Noyelle and Stanback, 1983; Wheeler, 1990). This regional and metropolitan emphasis can be attributed to the fact that the majority of the U.S. population now lives in metropolitan areas. The statistical data base for Metropolitan Statistical Areas (MSAs) and census regions is also much larger than that available for individual nonmetropolitan counties, increasing the opportunity for more detailed studies of these regions. Literature on nonmetropolitan restructuring has‘ stressed that the nonmetropolitan economy has been structurally transformed to a diverse heterogeneous economy, which is no longer dependent upon agriculture (Bonnen, 1990: 'Castle, 1987; Henry, Drabenstott, and Gibson, 1988; Swanson, 1989; Hady and Ross, 1990; Marsden, Lowe and Whatmore, 1990: McNamara and Gunter, Rainey, 1976; Summers, Horton and Gringeri, 1990). Nevertheless, much of the literature continues to focus on the agricultural industry (Bonnen, Nelson, and Deavers, 1988; Commins, 1990; Fitzsimmons, 1986: Susman, 1989). 7 The dominance of agricultural studies can be attributed to several factors. Historically, agriculture dominated the social and ideological structure of nonmetropolitan U.S., the present day heterogeneity of nonmetropolitan counties has slowed the development of social and political institutions outside of agriculture, agriculture still dominates the physical landscape, and some authors such as Crown (1991) have continued to use the terms interchangeably. All of these factors coupled with agrarian fundamentalism contribute to the perception that the nonmetropolitan United States continues to be an agrarian based economy (Swanson, 1989; Bonnen, 1990). While some nonagricultural studies have been done on nonmetropolitan communities (Roepke and Freudenberg, 1981; Aiken, 1990; Glickman and Glasmeier, 1989; Glasmeier and Borchard, 1989; Glasmeier and Glickman, 1990: Glasmeier and Kays-Teran, 1989; Amos, 1988: Barancik, 1990; Esparza, 1990; Smith, 1990), significant gaps still exist in the literature concerning the impact of economic restructuring in 'nonmetropolitan areas. The differential impact of industrial transformation on metropolitan and nonmetropolitan counties has not been fully examined. A clear understanding of the present economic structure in most nonmetropolitan counties is also lacking. Many authors (Henry, Drabenstott, and Gibson, 1988; Lapping, Daniels, and Keller, 1989; Bonnen, 1990; Hady and Ross, 1990; Marsden, Lowe, and Whatmore, 1990) have stated that nonmetropolitan 8 America has evolved into a diverse, heterogeneous economy, which cannot be analyzed in aggregate, but Hady and Ross (1990) are among the few authors who have attempted to describe the economic structure present in all nonmetropolitan counties in the United States. Their model, which classified counties based on the percentage of labor and proprietor income generated by selected industries, was static, and left many counties unclassified. Some authors (Smith, 1990; Fik, Amey, and Malecki, 1991) have examined the impact of certain industries on nonmetropolitan counties within a particular state, but the body of literature is not large enough to provide a comprehensive picture of the impact of industrial transformation on nonmetropolitan regions. Social and economic data for nonmetropolitan areas is extremely limited (Swanson, 1989); and as a result, little attention has been paid to the economic and social well being of nonmetropolitan residents (Flora, 1990). 1.3 Purpose This study examines trends in industrial transformation in Michigan's metropolitan and nonmetropolitan counties between 1978 and 1987. The spatially and temporally varying associations between industrial specialization and the level of real per capita income in metropolitan and nonmetropolitan counties in Michigan will also be analyzed. The spatial analysis examines the varying association between industrial specialization and the level of real per 9 capita income between metropolitan and nonmetropolitan counties. The temporal analysis examines the varying association between industrial specialization and the level of real per capita income over time. The three major industrial sectors considered are manufacturing: wholesale and retail trade; and finance, insurance and real estate (FIRE). 1.4 Data In general, counties are differentiated as metropolitan, nonmetropolitan urban, and nonmetropolitan rural. In this study, nonmetropolitan urban and nonmetropolitan rural are combined and the analyses differentiates between metropolitan and nonmetropOlitan counties However, in the descriptive analysis, nonmetropolitan counties are divided into nonmetropolitan urban and nonmetropolitan rural. Adjacent and nonadjacent counties which contain an urban population of 20,000 to '49,000 are considered nonmetropolitan urban, so are counties which contain an urban population of 2,500 to 19,999 and are adjacent to a metropolitan county. Nonadjacent counties with an urban population under 19,999 are classified as nonmetropolitan rural (Figure 1.1). Variables representing economic conditions include: personal income and earnings by place of residence. Personal income was chosen as a measure of economic well- 10 Metro-Nonmetro If Classification of Michigan Counties Nonmetro Rural Figure 1.1. Metro-Nonmetro Classification of Michigan Counties 11 being, because it includes both employment and nonemployment income. It is derived from two sources: 1) earnings, which includes labor and proprietor income and 2) nonemployment income, which includes dividends, interest, and rent, and transfer payments. Although nonemployment income is increasing in Michigan, earnings still comprise the bulk of personal income (Bureau of Economic Analysis, 1981, 1990). Income data were obtained from the Bureau of Economic Analysis and U.S. Department of Commerce, (1981 and 1990). The consumer price index was used to adjust 1987 income and earnings to 1978 dollars, compensating for inflation between 1978 and 1987. Industrial employment data are used to understand the patterning of job growth and job loss.across industrial sectors. Changes in the number of establishments per industry category are used to examine business establishment expansion and contraction during the study period. Variables selected to examine changes in industrial structure include the total number of employees and 'establishments in each county in the following categories in 1978 and 1987: agricultural, fishing, and forestry (AFF) services; mining; construction; manufacturing: . transportation, communication and public utilities (TCPU); wholesale trade; retail trade: finance, insurance, and real estate (FIRE); and other services, which includes personal, health, recreation, hotel, and any services not included in the other service categories. Employment and establishment 12 data were obtained from County Business Patterns (U.S. Department of Commerce and Bureau of Census, 1980 and 1989), which provides annual information on the number of employees and number and size of establishments for each industrial category listed. County Business Patterns does not include government employment. It also does not distinguish between full-time and part-time employment. CHAPTER TWO LITERATURE REVIEW Factors used to explain rising income inequality in the United States have included the characteristics of the dominant industries, level of industrial specialization, supply and demand shifts such as the size and age of the workforce, variations in education and skill, rising global competition, monetary policy, and the quantity and quality of jobs. Other explanations such as changes in the relative importance of location factors, technological changes, business cycles, variations in the export potential of regions, backwash and polarization effects, capital and labor flows, and product cycle stages have also been used (Walker and Storper, 1981; Lonsdale, 1982; Massey, 1984: Bradbury, 1985: Blackley, 1986; Booth, 1986; Amos, 1988: Kale, 1989; Harrison and Bluestone, 1990: Moriarty, 1991: Bound and Johnson, 1992; Grubb and Wilson, 1992: Levy and Murnane, 1992; Ray, 1992). This study examines the association between industrial specialization and income levels. The following sections include a discussion of industrial shifts in the United States, and socio-economic impacts of industrial shifts. l3 14 2.1. Industrial Shifts in the United States In the first half of the 20th Century the United States moved from an agrarian based economy to a manufacturing based economy. Between 1939 and 1953, the farm population declined by 35 percent as manufacturing industries in metropolitan areas in the North East and North Central- regions expanded to meet the demands of the military, and then retooled in the post war years to fill consumer demand for durable goods (Cochrane, 1976: Deavers, 1991: Hatton and Williamson, 1992). In the late 1970s, these regions began declining as sectoral shifts from manufacturing to services took off (Markusen and Carlson, 1989: Moriarty, 1991). The highest manufacturing employment losses occurred in large, highly unionized cities (O'hUallachain, 1990), which had previously attracted rural migrants (Cochrane, 1976: Hatton and Williamson, 1992). The net gainers for manufacturing employment during this time period included regions formerly ‘in decline and newly industrializing regions: parts of New England, the West North Central, Mountain, Pacific and the South (Markusen and Carlson, 1989). These shifts in the location of production were accompanied by changes in the mode of production and changes in the relative importance of industrial location factors. Classic location models had predicted that producers would largely locate facilities at a point which minimizes the 15 cost of distance (Young, 1986: Kale, 1989). Technological changes in transportation and industrial production, however, have reduced the relative significance of distance as a location factor (Lonsdale, 1982: Young, 1986). Access to markets is still an important locational consideration (Fik, Amey, and Malecki, 1991: Patton and Markusen, 1991), but it is time and network linkages now, rather than actual distance, that determine transportation costs (Dubin, 1991: Drabenstott, 1991). Changes in the mode of production have resulted in a spatial reorganization of the production process, especially among large manufacturing firms. Large corporations have maintained their top-level management offices in large central cities (Wheeler, 1991: Ray, 1992), but moved middle management offices to the suburbs, and production units to more remote, low cost locations (Ray, 1992). Since the majority of manufacturing employment is in production, decentralization of production has led to manufacturing employment declines in larger cities, and manufacturing employment growth in smaller cities '(O’hUallachain, 1990). This does not mean that smaller cities have reaped all of the benefit at the expense of the larger urban centers. In the process of restructuring, a social and spatial division of labor has occurred. While the number of high wage, semiskilled jobs has declined in metropolitan counties, the number of professional and managerial jobs has increased. Manufacturing employment in metropolitan 16 counties has become more capital intensive, involving more high wage nonproduction related activities, while nonmetropolitan manufacturing activity has become more low wage production oriented (Blackley, 1986: Harrison and Bluestone, 1990: Deavers, 1991: Phillips and Miller, 1991: McGranahan, 1992). In the service sector, the fastest growing sector in the 1980's has been producer services and retailing. Producer services have preferred metropolitan locations (Kirn, Conway, and Beyers, 1990: Wheeler, 1990), while retailing and other trade related industries such as wholesaling show relatively more dispersion to nonmetropolitan areas. As service and high-tech manufacturing industries began to increase in the United States, it was anticipated that these industries would encourage nonmetropolitan growth (Kirn, Conway, and Beyers, 1990). Many of these are often considered to be footloose industries. Factor price equalization and product cycle models predicted these industries would decentralize, and ‘improvements in telecommunications gave nonmetropolitan areas better access to information and nonlocal markets (Kirn, Conway, and Beyers, 1990: Dillman, 1992). However, few nonmetropolitan counties have been able to attract high- tech manufacturing and higher order service industries, such as producer services. Nonmetropolitan counties which have been able to attract these industries are usually adjacent 17 to metropolitan counties, or in remote counties with large populations (Barkley and Keith, 1991). Many nonmetropolitan counties have failed to attract ~ high-tech manufacturing and higher order service industries, because they lack infrastructure such as constant voltage electricity and private telephone lines (Glasmeier and Borchard 1989: Glasmeier and Kays-Teran, 1989). Other counties which have the necessary infrastructure have not been able to attract these industries, because service industries have preferred to remain in metropolitan counties where they can access agglomeration economies, have greater opportunities for face to face contacts (Henry, Drabenstott, and Gibson, 1988: Goe, 1990: Deavers, 1991), and have ready access to a high skilled labor force (Kirn, Conway, and Beyers, 1990). Improvements in telecommunications have also allowed these companies to serve nonmetropolitan areas without relocating away from metropolitan regions (Kirn, Conway, and Beyers, 1990; Dillman, 1992) As a result, large metropolitan corporations can compete directly with small *local businesses. Producer services which do locate in nonmetropolitan counties are usually branch offices of multilocational organizations. These offices mainly perform back office functions such as data processing (Glasmeier and Kays-Teran, 1989). The wage rate for back office jobs is less than high order white collar jobs (Kirn, Conway, and Beyers, 1990). Their jobs are also less stable, as multilocational 18 companies can and do relocate facilities to other locations in search of lower labor costs (Kirn, Conway, and Beyers, 1990: Ray, 1992). Multilocational firms in manufacturing and services also tend to bring their own managers and skilled workers with them. As a result, many of the jobs they create are low-wage jobs (Smith and Barkley, 1989: Kirn, Conway, and Beyers, 1990: Phillips and Miller, 1991: Ray, 1992). These companies tend to have network linkages established with suppliers outside of the local area, and therefore do not generate the same multiplier effect as a company which purchase parts and supplies locally (Glasmeier and Borchard, 1989: Kirn, Conway, and Beyers, 1990; Phillips and Miller, 1991). Locally owned producer services, in contrast, have strong links to the local economy and provide a higher proportion of high skill, high wage jobs (Kirn, Conway, and Beyers, 1990). ’2.2 Socio-econonic Consequences of Industrial Restructuring As the United States shifts from a manufacturing based economy to a service based economy, metropolitan and nonmetropolitan counties are experiencing different changes and rates of growth in employment. ‘While low wage consumer service jobs have increased in metropolitan and nonmetropolitan counties, growth in high wage producer 19 service jobs has occurred primarily in metropolitan counties. Metropolitan growth in these industries has also been faster. Between 1979 and 1987, nonmetropolitan employment for the nation increased 8%, while metropolitan employment increased by 20% (Mazie and Killian, 1991). In nonmetropolitan counties most of the employment growth came from consumer services (Deavers, 1991: Mazie and Killian, 1991). Jobs in this sector tend to be low-wage across occupational categories, and are often part-time (Mazie and Killian, 1991). Slow economic growth dominated by consumer service growth has resulted in higher unemployment and underemployment for individuals remaining in these counties. Nonmetropolitan economic growth could be increased if more attention was directed toward non-traditional goods and services of nonmetropolitan areas for which demand may be growing (Castle, 1987). High amenity areas, for example, have been able to attract migrants and some industries, such as computer software firms and upscale food manufacturing andclothing marketers that other areas have not (Dillman, '1992). Tourism has also become more important to nonmetropolitan communities. While tourism can create jobs and provide additional income, most of the jobs it creates are low-wage and seasonal. Tourism also increases land-use conflicts in nonmetropolitan areas (Clark, 1992). As a result of industrial shifts, nonmetropolitan counties and central city counties were left behind in the 19808, because both areas lack the capital and human 20 resources necessary to promote development (Swanson, 1989). According to Mead (1991), metropolitan poverty results not from the lack of opportunities, but rather from the inability to take advantage of the opportunities present. Individuals below the poverty level in central city counties often lack the education and training necessary to find and maintain a well paying job (Wilson, 1987: Mead, 1991). Furthermore, adequate, affordable childcare is often not available for these individuals (Mead, 1991): and public transportation has not kept pace with firm decentralization, restricting the mobility of job-seeking individuals (Dubin, 1991). As a result, many of these individuals only have access to low wage jobs (Mead, 1991). Nonmetropolitan poverty results from low wages, unemployment, underemployment, depression in the primary industries, and state welfare rules that exclude significant numbers of individuals poor by national standards (Molnar and Traxler, 1991). Structural reasons for nonmetropolitan. poverty include several factors. Nonmetropolitan counties 'have a product mix which includes a significant proportion of unpriced public goods. Many counties are dependent on the extraction of raw materials, which are shipped through monopsonistic commodity markets to other locations for value added processing. Processing plants and distribution systems in nonmetropolitan counties are usually owned by entities located in metropolitan counties or other nations, who have no vested interest in the local community. 21 Significant transfer costs are imposed by economic, political, and physical distance from markets: and globally, industrial goods are being substituted for natural resources (Apedaile, 1991: Molnar and Traxler, 1991), decreasing the demand for traditional nonmetropolitan exports. Nonmetropolitan poverty is also exacerbated by high levels of migration, because the better educated and trained individuals are the ones most likely to migrate (Tweeten and Brinkman, 1976: Molnar and Traxler, 1991: McGranahan, 1992), and the least likely to return (DaVanzo, 1983). Twenty- seven percent of individuals migrating out of nonmetropolitan areas in 1986-1987 had 4 years of college: and outmigrants were twice as likely to have a college degree than those left behind (Molnar and Traxler, 1991). This loss of human capital reflects the fact that the more highly skilled jobs are disproportionally concentrated in metropolitan areas (Tweeten and Brinkman, 1976: Mazie and Killian, 1991: Molnar and Traxler, 1991), and metropolitan workers get paid more than nonmetropolitan workers for the ‘same job (Mazie and Killian, 1991). This disparity increases with the education level and is increasing over time. By the end of the 1980s, college graduates in metropolitan areas were being paid 30% more for the same job as their nonmetropolitan counterparts (Deaver, 1991: McGranahan, 1992). The earnings gap for high school graduates was only 12% (Deavers, 1991). 22 This continued migration of educated, skilled workers to metropolitan areas decreases the human capital necessary for development, and increases the perception that nonmetropolitan workforces are composed primarily of low educated, unskilled workers. As a result, nonmetropolitan counties find it difficult to attract high tech manufacturing industries or high order service industries which could provide better paying jobs. Since 1978, interstate and intrastate economic, social, and political inequality has increased within metropolitan areas and between metropolitan and nonmetropolitan areas (Edwards, 1976: Blackley, 1986: Alter and Long, 1988: Amos, 1988: Amos, 1989: Swanson, 1989: Aiken, 1990: Bonnen, 1990: Jones and Kodras, 1990: Summers, Horton, and Gingeri, 1990: Angel and Mitchell, 1991: Mead, 1991: Moriarty, 1991: McGranahan, 1992). In 1987 unemployment in nonmetropolitan areas in the U.S. was 31 percent higher than it was in metropolitan areas. Nonmetropolitan per capita income declined from 77 percent of metropolitan per capita income ‘in 1979 to 73 percent of metropolitan per capita income in 1987 (Jones, 1988: Barancik, 1990). Nonmetropolitan incomes have not only lagged behind metropolitan incomes, but they have also been less stable over time (Henry, Drabenstott, and Gibson, 1988). CHAPTER THREE ECONOMIC TRANSFORMATION IN HICHIGAH In Michigan, nonmetropolitan per capita income increased from 73% of metropolitan income in 1978 to 77% of metropolitan per capita income in 1987. Even though the gap narrowed, differences in income increased in Michigan during this time period (Table 3.1). While no significant difference existed between per capita income in metropolitan counties and nonmetropolitan counties in 1978, a significant difference did exist between per capita income in metropolitan counties and nonmetropolitan counties in 1987 (Table 3.1). Table 3.1. Metro-Nonmetropolitan Variation in Per Capita Income, 1978 and 1987. Nonmetro Metro F’ Prob > F' INC 1978 6055.31 8208.45 1.16 0.6384 INC 1987 7654.54 9970.31 2.91 0.0013 Realper capita income increased in all counties, except St. Clair between 1978 and 1987. However, not all counties experienced the same level of increase, thereby contributing to an increase in the difference in income between metropolitan and nonmetropolitan counties (Table 3.2). Earnings comprise the bulk of personal income in Michigan, but the proportion of personal income derived from earnings is decreasing, especially in nonmetropolitan counties, where nonemployment income rose from 36.8% of 23 24 Table 3.2 Change in Real Per Capita Income by County, 1978-1987 Real Change in Real Inc Per Inc Per Income Per Capita Cap 1978 Cap 1987 Abs. Percent Metropolitan Counties Bay 7,574 8,899 1,325 17.50 Berrien 7,596 8,897 1,301 17.13 Calhoun 8,254 8,801 547 6.62 Clinton 7,523 8,680 1,157 15.38 Eaton 7,843 9,450 1,607 20.49 Genesee 9,015 9,711 696 7472 Ingham 8,593 10,060 1,467 17.08 Jackson 7,672 9,014 1,342 17.49 Kalamazoo 8,346 10,328 1,982 23.75 Kent 7,967 10,421 2,454 30.80 Lapeer 7,302 9,060 1,758 24.08 Livingston 7,497 11,326 3,829 51.08 Macomb 9,414 11,576 2,162 22.97 Midland 8,682 10,842 2,160 24.87 Monroe 7,748 9,386 1,638 21.14 Muskegon 6,920 8,319 1,399 20.21 Oakland 10,975 15,024 4,049 36.89 Ottawa 7,601 9,919 2,318 30.49 Saginaw 8,359 9,326 967 11.57 St Clair 7,218 7,182 -36 -0.50 Washtenaw 9,266 13,223 3,957 42.71 Wayne 9,221 9,903 682 7.39 Metro Ave 8,208 9,970 1,762 21.46 Nonmetropolitan Counties Alcona 5,118 7,152 2,034 39.73 Alger 4,883 6,590 1,707 34.96 Allegan 6,687 8,663 1,976 29.55 Alpena 6,498 7,754 1,256 19.33 Antrim 6,158 8,080 1,922 31.22 Arenac 5,921 7,226 1,305 22.05 Baraga 5,762 6,852 1,090 18.92 Barry 6,382 8,969 2,587 40.54 Benzie 6,351 7,802 1,451 22.85 Branch 7,428 7,731 303 4.09 Cass 7,479 8,869 1,390 18.59 Charlevoix 6,504 8,190 1,686 25.92 Cheboygan 5,785 7,257 1,472 25.44 Chippewa 5,110 6,703 1,593 31.17 Clare 5,303 6,474 1,171 22.08 Crawford 5,478 6,424 946‘ 17.27 Delta 5,893 7,802 1,909 32.40 Dickinson 7,433 9,165 1,732 23.31 Emmet 6,966 9,416 2,450 35.17 Gladwin 5,461 6,635 1,174 21.50 Gogebic 5,949 7,124 1,175 19.75 Table 3.2 (cont’d) Grand Traverse Gratiot Hillsdale Houghton Huron Ionia Iosco Iron Isabella Kalkaska Keweenaw Lake Leelanau Lenawee Luce Mackinac Manistee Marquette Mason Mecosta Menominee Missaukee Montcalm Montmorency Newaygo Oceana Ogemaw Ontonagon Osceola Oscoda Otsego Presque Isle Roscommon Sanilac Schoolcraft Shiawassee St Joseph Tuscola Van Buren Wexford Nonmetro Ave Michigan 8,048 7,152 6,654 5,516 6,902 6,253 5,615 5,775 5,662 4,766 5,217 4,925 7,155 7,571 6,549 5,541 6,347 6,720 6,047 4,644 5,903 4,859 6,348 5,509 5,790 5,889 4,820 5,025 5,084 4,108 6,574 5,511 5,262 6,279 5,578 7,592 7,297 7,115 6,764 6,459 6,055 6,626 25 9,801 8,017 8,055 6,948 9,288 7,179 7,290 7,508 7,835 6,384 6,262 5,892 9,898 9,253 9,281 7,557 7,782 7,776 7,561 6,451 7,218 6,402 7,635 7,012 7,393 7,649 6,496 7,273 6,306 6,364 8,188 6,995 7,197 9,556 8,712 8,049 8,730 7,996 7,807 7,053 7,655 8,268 1,753 865 1,401 1,432 2,386 926 1,675 1,733 2,173 1,618 1,045 967 2,743 1,682 2,732 2,016 1,435 1,056 1,514 1,807 1,315 1,543 1,287 1,503 1,603 1,760 1,676 2,248 1,222 2,256 1,614 1,484 1,935 3,277 3,134 457 1,433 881 1,043 594 1,599 , 1,642 21.78 12.10 21.05 25.96 34.57 14.82 29.83 30.00 38.39 33.96 20.03 19.64 38.34 22.21 41.71 36.39 22.61 15.71 25.03 38.90 22.28 31.76 20.27 27.28 27.68 29.88 34.76 44.74 24.04 54.91 24.56 26.92 36.78 52.19 56.18 6.02 19.64 12.38 15.42 9.20 26.41 24.79 26 total personal income in 1978 to 46% in 1987 (Bureau of Economic Analysis, 1981, 1990). Although earnings are becoming a smaller percentage of income, the correlation between income and earnings remains strong (Tables 3.3 and 3.4). Table 3.3. Table 3.4. Pearson Correlation Coefficients for Income and Earnings per Capita in Michigan, 1978. INCOME EARNINGS INCOME 1.000 0.966 EARNINGS 0.966 1.000 Pearson Correlation Coefficients for Income and Earnings per Capita in Michigan, 1987. INCOME EARNINGS INCOME 1.000 0.918 EARNINGS 0.918 1.000 Even in nonmetropolitan counties, where the proportion of income derived from earnings is smaller than the state average, the correlation between income and earnings per capita is high (Tables 3.5 and 3.6). Table 3.5. Pearson Correlation Coefficients for Income and Earnings per Capita in Nonmetropolitan Counties in Michigan, 1978. .INCOME EARNINGS INCOME 1.000 0.918 EARNINGS 0.918 1.000 27 Table 3.6. Pearson Correlation Coefficients for Income and Earnings per Capita in Nonmetropolitan Counties in Michigan, 1987. INCOME EARNINGS INCOME 1.000 0.810 EARNINGS 0.810 1.000 The following sections examine changes in Michigan's industrial structure between 1978 and 1987. Percent distribution of employment and establishments is used to examine patterns of job growth, job loss, and business establishment expansion and contraction across industrial sectors. Entropy is used as a measure of industrial concentration/diversification: and location quotients are used to examine the level of employment specialization in manufacturing, wholesale and retail trade, and finance, insurance, and real estate. Industrial Structure in Michigan As the United States shifts from a manufacturing based economy to a service based economy, traditional manufacturing states such as Michigan are experiencing large declines in manufacturing output and employment. In spite of this decline, Michigan still remains more dependent on manufacturing than the country as a whole (Table 3.7). Michigan has experienced employment and establishment growth in the service sector, but its service sector in 1987 still comprised a smaller proportion of its employment and establishments than the national aggregate (Table 3.7). 28 .006mmo ms6uc66m ucoacuo>ou .m.D 6.0.0 .coums6nwez .msuouuem mmoswmsm hucsoo .6666 .6666 .msmcoo no seousm use 00606500 no ucmauuemoo .m.D Bonn eueo 6:66: ooueasoaeo “mousom .muouuo 0:60:306 on one 6006 on an ace uos mes maeuoa 6 .23026 m06uovoueo mo6>uom Honuo mnu s6 neonaos6 uoc mm6uumsec6 0069606 nonuo ace use .6090: .cowueouomu .nuaeos .Hecomumm 600560c6 mmo6>60m umnuo 6 .mueumm 6e0m o:e..00:e6=mc6 .mocec6m 06 mmHm 6 .6069666ub 066nsm one .mco6ueo6ssasoo .co6neuuommseua 66 Dave 6 .wmow>60m mc6sm6m one uuummuom .6e65u6906604 one moofi>60m and 6 6.666 6.666 6.666 6.666 6.666 6.666 6.666 6.666 666066 6.66 6.66 6.66 6.66 6.66 6.66 6.66 6.66 66606>666 06666 6. 6.6 6.6 6.6 6.6 6.6 6.6 6.6 66666 6.66 6.66 6.66 6.66 6.66 6.66 6.66 6.66 06666 666066 6.6 6.6 6.6 6.6 6.6 6.6 6.6 6.6 06669 666666663 6.6 6.6 6.6 6.6 6.6 6.6 6.6 6.6 66606 6.6 6.6 6.6 6.6 6.66 6.66 6.66 6.66 666060066666: .6.6 6.66 6.6 6.66 6.6 6.6 6.6 6.6 666005606660 6. 6. 6. . 6. 6. 6. 6. 6.6 66666: 6.6 6. 6.6 6.6 6. 6. 6. 6. 66606>006 666 6666 6666 6666 6666 6666 6666 6666 6666 660006 :e66n06z mmueum 00u6c9 cev6zowz mmueum omu6ca HewuumsocH 66262666666966 6262666626 .6666 use 6666 .cevwno6z use wmueum omufico may you musoanmfineumm one newshoamam mo cowusn66um6o unmoumm .6.6 wanes 29 Tables 3.8 and 3.9 show the percent distribution of employees and establishments in nine industrial sectors for different county types in 1978 and 1987. These tables indicate that the industrial structure for employment is considerably different than the industrial structure for establishments. The three largest employment categories for the state and all county types are manufacturing, retail trade and other services: while the three largest establishment categories are construction, retail trade and other services. This is due, in part, to differences in firm size, which varies across sectors. As a result, equal increases in the number of establishments in different sectors will generate different levels of employment. For example, construction firms, which on average employed 8 workers per establishment in 1978, comprised 10.2% of the establishments in Michigan, but only 4.1% of the total employment for that year, while manufacturing firms averaged 82 employees per establishment in 1978 and 62 employees per establishment in 1987. Manufacturing establishments comprised approximately 9% of the establishments for the state and each county type in 1978 and 8% in 1987, but accounted for approximately 40% of the total employment in 1978 and 30% of the total employment in 1987. Retail trade and other services combined accounted for approximately 60% of the establishments, but only 40% of the total employment. Establishments in the other services sector on average employed 13 persons per establishment in 1978 and 12 persons 30 .006000 0n6un660 un0enn0>oo .0.D 6.0.0 .noumn6n0e3 .mnuouuem 000:6000 >unnoo .6006 .0006 .msmn0o 00 ne0unm one 00605600 no un05unem0o .0.0 3060 eueo 0n6ms o0ue650600 ”006500 .mnonno 0n6onson ou 09o «006 on a: ooe uon aes 06euoa 0 .nsonm 0066000ueo 0o6>n00 n0nuo 0n» :6 o0o060n6 uon 006numson6 0o6>60m n0nuo >06 one .60uon .0069006006 .nu6e0n .60n00600 00o060n6 00o6>n00 n0nuo 0 .006000 6600 one .0onensmn6 .0onen60 06 0060 6 .0060666un 066000 one .0n06ueo6nsaaoo .no6ueunomwnens 06 0009 6 .0006>n00 0n6ne60 one 06000600 .6e6006006600 0ne 00o6>u00 000 6 .666660 6660 0666666060 000690650600 :6 o00: 0a 6663 065060 uena .000006050 00 n0nasn Heouoe on» 600 00:06 on» no unwomo6s 0n» o0u9u6umnsm 0>en H 00oneuwn6 000:» 06 .n0>60 06 000>06060 00 6006:: Hesuoe 0n» nenu u0nuen 000e6 e .o006566 06 nouo00 0no >00 :6 06660 00 60055: 0nu 06 .0060 .006unsoo 000:» 600 060066e>e uon n0uuo 06 n0mono 0609000 6e6numson6 00» 30600 eueo un0ann0>00 .060006 e 04 .0366 >0e>660 £063 >60600 ou o0000nansw 06 009000 oou6n0 0n» n6 006unsoo neu66omonu0finon 600 eueo 0500 6 0.006 0.006 0.006 0.006 H.006 6.006 0.006 0.006 0060908 0.00 0.0a 6.00 0.06 6.00 0.66 0.6N 0.0m 00006>600 nonuo v.6 0.0 0.v 0.6 0.6 0.6 0.0 0.0 vmmnm 0.60 6.00 0.60 6.00 0.0m 0.00 0.60 0.06 0oeufi Haeuom n.v 0.0 0.v 6.6 v.v 6.6 0.0 0.0 0oena 06000HO£3 6.6 0.6 6.6 0.6 0.6 0.6 0.6 6.0 0009 0.00 «.06 0.00 0.60 v.00 6.00 0.60 0.06 Onwunuoemnnez 6.0 0.0 6.0 6.0 6.6 «.6 0.0 6.6 noduonuumnoo 6.6 6.0 0.H 6.0 0.6 6.0 a. H. mnanwz 6. 6. 0. 6. 6. 6. 6. 6. ~00o6>n00 000 6006 060a 6006 0606 600d 0606 6006 060d Houoom nenna Hennm 6onuofinoz neu66ononu0z 6e6numnon6 0096aonouuoanoz 66¢ . 0009 aunnoo 6006 one 0606 .0008 huumnonu >2 un0fi>oamsm mo no6usnwnumwo un0ou0m .0.6 06069 31 .006000 on6un6um un0Bn60>00 .0.D 6.0.0 .n0u0n6nmes .mn60uuem 000n6000 aunnoo .6006 .0006 .msmn0u 00 0e0600 one 00605600 00 un0fiuue00o .0.D 5060 eueo mn6mn o0ue6no6eo “006000 .060660 mn6on006 ou 00o 0006 on an ooe uon 606 06eu09 .naonm 0066000uem 006>600 60:90 on» n6 o0on60n6 uon 0066umnon6 006>60m 60:90 6ne one .60uon .no69006006 .nu6e0n .6en0060a 00on6on6 0006>600 60:60 6 0.006 0.006 0.006 6.006 6.006 6.006 6.006 0.00 6066609 0.66 6.06 0.06 0.06 6.66 6.06 6.06 0.06 60006>600 60:60 6.6 6.6 0.0 0.0 0.0 6.6 0.6 0.0 6066 6.66 0.66 0.66 6.66 0.66 6.66 0.06 6.66 00660 6606064 6.0 0.6 0.0 0.0 6.0 0.6 0.6 6.6 00669 066006063 6.6 6.6 6.6 6.6 0.6 6.6 6.6 0.6 0000 6.6 6.0 6.6 6.0 6.6 6.0 6.6 6.0 006606006006: 6.0 0.06 0.66 0.66 6.06 0.66 6.6 6.0 006600660000 0. 6. 0. 0. 0. 6. 6. 6. , 00606: 6.6 0. 6.6 0. 6.6 0. 6.6 0. 00o6>600 664 6606 6606 6606 6606 6606 6606 6606 6606 606000 nenub 6eunm 06u0Enoz nmu66oaouu0z 6e6uumaon6 006660006605002 660 . 6606 0:6 6606 .0066 66600006 66 06005006666600 60 006606666060 6000606 .0.6 06669 32 per establishment in 1987. Retail establishments averaged 13 workers per establishment for both years. A comparison of metropolitan and nonmetropolitan counties shows that while the majority of jobs for both county types were in the manufacturing sector for 1978 and 1987, manufacturing's share of employment is declining, particularly in metropolitan counties. In 1987, . manufacturing employment comprised a larger percentage of the total employment in nonmetropolitan counties than it did in metropolitan counties. Manufacturing employment was particularly concentrated in the nonmetropolitan urban counties, where it comprised a larger percentage of the total employment than in metropolitan or nonmetropolitan rural counties for both years. In metropolitan counties, the second largest employment category was "other services", which includes personal, health, recreation, hotel, and any other service not included in the other service categories listed in Tables 3.7-3.11. Employment in other services grew from 20% of metropolitan employment in 1978 to 27% in 1987. Retail trade was the second largest employment category in nonmetropolitan counties, especially in nonmetropolitan rural counties. In 1987, retail jobs comprised 27.2% of all ' nonmetropolitan rural jobs, just .8% less than the percentage of employment in the manufacturing sector for these counties. 33 Tables 3.10 and 3.11 show the change in employment and the number of establishments by industry between 1978 and, 1987. Absolute change is used to show the change in the number of employees and establishments, and percentage change is used to show the relative importance of that change. Although the number of establishments increased in all sectors for all county types (Table 3.11), employment decreased in some sectors for some county types (Table 3.10). Mining employment decreased in nonmetropolitan counties, while increasing in metropolitan counties. Construction employment increased in metropolitan and nonmetropolitan counties. Increased construction employment in nonmetropolitan counties occurred primarily in nonmetropolitan rural counties, nonmetropolitan urban counties saw a decrease in construction employment. Employment in transportation, communication, and public utilities decreased in metropolitan counties, but increased in nonmetropolitan counties. Manufacturing employment declined in metropolitan and nonmetropolitan counties, with the highest job losses occurring in metropolitan counties.' Within the manufacturing sector, the number of high tech establishments increased, while the number of employees decreased. High tech industries are industries which employ a higher percentage of engineers, engineering technicians, computer scientist, mathematicians, and life scientists-than the national average, and apply science and engineering . .006600 0:66:660 6:05:60>00 .0.0 6.0.0 .noumn6nwe3 .mn60uuem 000:6000 66:000 .6006 .0006 .momn00 60 0e0600 one 00605500 60 6:0566e000 .0.0 5066 eueo 0:600 oeue6oo6eo ”006000 .:3o:w 0066om0ue0 006>60m 60:60 0:» :6 o0on6on6 60: 0066umoon6 006>600 60:60 6:0 one .6060: .no6ue06006 .nu6e0: .6enow600 00on6on6 0006>600 60:60 6 .m6m0nu 06:6 60000006nu 0no6ue6no6eo :6 o0wn 0Q 6663 060066 uen9 .000606050 60 6005:: 6enuoe 0:0 600 00ne6 0:6 00 un6o0o65 0:6 o0unu6umnnm 0>en 6 m00:eumn6 000:» :6 .:0>60 06 000606050 60 60050: 6enuoe 0:6 :enu 60nue6 00:06 e .o066566 06 606000 0:0 6ne :6 05666 60 60050: 0:» 66 .0306 600>660 :663 660500 on o0mm0600om m6 moueum o0u6cb 0:6 :6 0060:000 :0066000660fino: 600 eueo 0:6 00 :00: 6 606.66 600.06 066.00 000.066 066.606 0mneno 60609 600.66 600.06 666.60 000.06 666.00 600.66 660.66 660.666 606.66 600.666 600o6>600 60060 «66.66 .000.6 000.66 600.6 «66.66 000.6 066.66 060.06 006.66 066.66 0060 060.66 600.66 «06.66 666.0 006.66 060.66 000.06 060.60 006.66 600.666 00069 66e60m 006.66 600.6 «00.6 000 000.06 600.6 000.06 660.06 060.06 006.06 0oe69 060006033 000.06 660.6 006.66 600 «66.06 060.6 060.6! 000.6! 060.0! 606.6! D009 000.6! 060.6! 006.0! 066.6! 066.6! 006.0! 060.06! 000.006! 066.66! 066.666! 0:66nu0eunnez 066.66 060.6 «66.6! 606! 060.0 066.6 000.6 666.6 066.6 606.0 :o6uon6um:oo 000.66! 000.6! 066.66! 000.6! 000.66! 666.0! 060.0 66 000.06! 066.6! 0:6:62 066.00 600 066.00 .006 060.00 000 000.66 000.6 066.06 006.6 0006>600 00¢ 0:00 0000 0:00 0:00 0:00 0:00 0:00 0:00 0:00 0:00 606000 a .000 0 .064 0 .000 0 .000 0 .020 6e66umnon6 6e6nm :en6o 606605noz :06660006605noz 66¢ neu66o006u02 :006006: 0069 66:000 0069 66:500 an 6:05606050 :6 00:0:0 .06.6 06666 35 .006000 00660660 6005060>00 .0.0 6.0.0 .0060060003 .00606600 00006000 660000 .6006 .0006 .000000 00 000600 000 00605500 60 6005660000 .0.0 5060 0600 00600 0060600600 ”006000 .03000 0066000600 006>600 60060 006 06 00006006 600 0066600006 006>600 60060 600 000 .60600 .0066006006 .066000 .60000600 00006006 0006>600 60060 6 006.0 006.6 606.0 006.66 660.66 005000 60609 066.06 660.6 066.06 600.6 066.06 006.6 006.66 666.06 000.66 666.66 60006>600 60060 060.06 606 000.06 066 060.06 060 000.06 006.6 066.66 006.6 0060 060.06 606.6 060.66 600 060.66 606.6 060.06 066.0 006.06 066.0 00069 660600 060.66 666 060.0 60 066.66 006 006.66 066.6 066.06 666.6 00069 060006003 006.06 006 066.06 006 000.66 666 006.06 666.6 000.06 000.6 0009 066.66 006 060.0 666 006.66 006 000.66 066.6 006.66 060.6 006606000000: 006.66 606 000.0 60 060.0 666 000.66 000.6 060.66 660.6 506600660000 000.66 00 006.6 6 060.66 60 060.0 06 006.66 00 006562 066.66 066 060.06 66 060.66 606 066.06 666 066.66 060 0006>600 06¢ 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 606000 6 .000 0 .000 0 .006 6 .00< 6 .00< 6066600006 60600. 00060 06605002 006660006605002 . 66¢ 006660006602 0006006: 0069 660000 0069 660000 60 06005006600600 06 000000 .66.6 06009 36 principles in product and process development (Glasmeier and Kays-Teran, 1989; Smith and Barkley, 1991). Metropolitan counties attracted the largest number of new high tech establishments during the study period. However, the largest percentage increase in high tech establishments occurred in the nonmetropolitan counties, particularly nonmetropolitan rural counties. Nonmetropolitan rural counties also suffered the lowest decrease in high tech employment. As a result, nonmetropolitan rural counties were the only county type in_ which high tech manufacturing employment increased as a percentage of total manufacturing employment (Tables 3.12 and 3.13). Table 3.12. Changes in High Tech Employment and Number of High Tech Establishments in Michigan Between 1978 and 1987. EMPLOYMENT ESTABLISHMENTS Absolute Percent Absolute Percent Change Change Change Change Michigan -124,651 -17.34 851 11.37 Metropolitan -118,479 -18.25 637 9.72 Nonmetropolitan -6,172 -7.33 214 24.23 Nonmetro Rural -468 -1.84 134 33.25 Nonmetro Urban -5,704 -12.82 . 80 15.21 Table 3.13. High Tech Manufacturing as a Percentage of All Manufacturing, 1978 and 1987. Percent High Percent High Tech Tech Employment Establishments 1978 1987 1978 1987 Metropolitan 62.80 61.16 57.20 55.80 Nonmetropolitan 50.42 48.63 33.80 36.71 Nonmetro Rural 45.30 46.13 25.62 29.31 Nonmetro Urban 55.54 51.12 41.98 44.10 37 While both metropolitan and nonmetropolitan counties. lost employment in high tech manufacturing between 1978 and 1987, the largest employment losses occurred in metropolitan counties, which lost employment in every SIC code used, except Instruments and Related Products (Table 3.14). Nonmetropolitan urban counties gained employment in the Chemical and Added Products, and Electrical and Electrical Equipment categories, but lost employment in Fabricated Metals, Machinery (except Electrical), Transportation Equipment, and Instruments and Related Product categories. Nonmetropolitan rural counties gained employment in Fabricated Metals, Electrical and Electrical Equipment, and Instruments and Related Products categories, but lost employment in Chemical and Added Product, Machinery (except Electrical), and Transportation Equipment categories. The largest employment gains in Michigan occurred in the service sectors, especially in the retail and other services sectors. Nonmetropolitan counties experienced.the largest percentage increase in both of these sectors. The largest percentage increase in retail employment occurred in nonmetropolitan rural counties, while the largest percentage increase in other services occurred in nonmetropolitan urban counties. The number of employees and establishments in the FIRE sector increased for both metropolitan and nonmetropolitan counties. The majority of employees and establishments in the FIRE sector are located in 38 pcmssuo>oo .m.D ”.0.o .couvcdnmms .mcuwuumm mmmcwmsm mucsou .moflumo mcflucwum .smma .omma .msmcoo no :mwuzm can mouofifioo mo pawsuummwo .m.D Bonn sumo opens noustono "wousom .mucsoo umnu you Hapomm awn» CH momaoanfim mo hogan: stuom on» you mossy ecu mo unwoaofifi on» omusuflumnsm m>mc H .mmosmumCH omen» CH MNH.H Hmm.dl mmH hHh.HI hmH.N mum! Hanna onumasoz can: mac wms.au msm.nu os~.~ mm¢.~ mmm.mu .mom.mn msm.~n swan Hum emu CGQHD OMHOECOZ ouumacoz HH< onus mussoo me~.~ mos.wmu msa.nu om~.emu mmm.¢au Hmo.mau Houumz mmfiuovmumv on omuomamm :« mvcmno ucofihonsm .so>«m ma mwmhoHQEm mo Hones: Hmsuom on» can» Manama mossy m .omuwafla ma aucsou >cm no“ uouomw oco was s« mfihfiu no Hones: on» «H Hmsofi>wocfl Hon oHanHm>m mamaaw no: ma Hw>0H was» us sumo HmwuumaocH .m0HUC500 H now.~ muosoouq owumamu a mucossuumcH mm~.om| mango uuoamcmua can: fiasco oflcouuomam a vauuomam www.mmn Hmowuuowao umwoxw .aumanomz wem.¢al macaw: noumofiunwm mHH.~Hu muosooum 00664 a manoesmso savanna: #H.m THDMB 39 metropolitan counties. These counties also saw the largest numerical increase in FIRE employment and establishments.. The rate of increase was highest in the nonmetropolitan counties, especially the nonmetropolitan rural counties, which have a smaller number of employees and establishments in the FIRE sector. In summary, between 1978 and 1987, the number of establishments increased in all county types, but this positive change in the number of establishments did not translate into a positive change in employment for all sectors in all county types. The highest job losses occurred in manufacturing, with metropolitan counties experiencing the highest losses. Service sector employment increased, but the service sector in Michigan is still smaller than the national aggregate. Most of Michigan's high tech manufacturing and FIRE employment was located in metropolitan counties, while retail trade employment comprised a larger percentage of nonmetropolitan employment. 3.2 Industrial Diversification The nine employment categories shown in Tables 3.7-3.11 were used to calculate relative entropy as a measure of industrial diversification/concentration for each county for both years. Absolute entropy (H) is calculated as follows: H=épi 1n pi 40 where Pi is the proportion of establishments or employees in the ith industrial category, and 1n denotes natural logarithms. Absolute entropy (H) is normalized as follows to facilitate comparison R = (H/ln k)100 where k is the number of industrial sectors. Relative entropy (R) measures the distribution of employees and establishments within the k industrial sectors. Its values range between 0 and 100. A relative entropy of 0 equals total concentration. A relative entropy of 100 implies that the employees and establishments are distributed equally between all sectors (Garrison and Paulson, 1973; Clarke, 1985; Wheeler, 1990). Concentration in a particular industry can give a region a comparative advantage, but income will be unstable as the region will be sensitive to short term economic shocks. Diversified economies are less sensitive to short term economic shocks and provide a more stable economic base (Henry, Dranbenstott and Gibson, 1988: Gilchrist and St. Louis, 1991). I Employment entropy for the state increased, implying that employment is becoming more evenly distributed between industrial sectors (Table 3.15). Establishment entropy decreased in Michigan between 1978 and 1987 for all county types, resulting in an industrial structure which was slightly more concentrated in 1987 than in 1978. Table 3.15. Michigan Metropolitan Nonmetropolitan Nonmetro Rural Nonmetro Urban Table 3.15 indicates that the majority of change EMPLOYMENT 1978 1987 74.77 76.44 73.96 76.21 75.25 75.95 77.66 78.33 72.83 73.56 CHNG 2.00 2.25 0.70 0.67 0.73 Change in Relative Entropy ESTABLISHMENT 1978 1987 CHNG 79.87 78.16 -1.71 79.67 77.76 -1.91 79.79 78.92 -0.87 79.55 79.03 -0.52 80.02 78.81' -1.21 employment and establishment entropy occurred in the metropolitan counties. in Absolute change in relative entropy for individual counties within each county type shows that the change in employment entropy for metropolitan counties ranged from -3.14 to 5.38. The majority of these counties experienced an increase in employment diversity (Appendix A). Entropy values for nonmetropolitan county varied considerably, as did the direction and magnitude of change in entropy values. Absolute change in employment entropy in nonmetropolitan urban counties ranged between -3.13 to 7.48, while absolute change in employment entropy for nonmetropolitan rural counties ranged from -9.59 to 15.22 (AppendiX'A). 'Absolute change in relative entropy for individual counties showed that establishment diversity decreased between 1978 and 1987 in all metropolitan counties, except St Clair. most nonmetropolitan counties, some nonmetropolitan While establishment diversity also decreased for counties, such as Missaukee, experienced large increases in 42 establishment entropy. .Absolute change in establishment entropy for metropolitan counties ranged from -7.70 to 0.24. Absolute change in establishment entropy ranged from -4.06 to 1.71 for nonmetropolitan urban counties, and -10.15 to 8.41 for nonmetropolitan rural counties (Appendix B). The previous section demonstrated that Michigan is experiencing a sectoral shift from manufacturing to services. Changes in employment entropy indicate that employment is becoming more evenly distributed between industrial sectors. While changes in metropolitan entropy values were fairly homegeneous across counties, large variations existed in the direction and magnitude of changes in nonmetropolitan entropy values. However, Spearman correlation coefficients between relative entropies in 1978 and 1987 were 0.829 for establishment entropies and 0.825 for employment entropies, indicating a high similarity in industrial structure for both years. 3.3 Employment Specialization Industrial specialization in Michigan counties was evaluated using location quotients. Since the major focus of this study is the shift from manufacturing to service industries, location quotients were only calculated for the following industrial sectors: manufacturing, wholesale and retail trade, and FIRE for 1978 and 1987. Wholesale and retail trade were selected as a proxy for lower order services, and FIRE was selected as a proxy for higher order 43 services. The location.quotient for a county is determined by: L0 = (Xi/Yi)/(Xj/Yj) where Xi is the number of individuals employed in an industry in'a county, Yi is the total number of employees in a county, Xj is the number of individuals employed in an industry in the state of Michigan, and vi is the total number of employees in the state. A location quotient greater than one implies that a county is more specialized in, or has a higher level of employment in that sector than the state average. A location quotient less than one implies that a county has a lower level of employment in that sector than the state average (Isserman, 1977: Hammond and McCullagh: 1978). An examination of the location quotients for these industrial sectors show that in 1978, 52 of the 83 counties in Michigan had a higher level of employment in wholesale and retail trade than the state average. That number increased to 54 in 1987 (Figure 3.1). The majority of these counties were nonmetropolitan. In 1978, 67.2% of all nonmetropolitan counties had higher levels of employment in wholesale and retail trade than the state average (Table 3.16). That percentage increased to 73.8% in 1987. Fifty ' percent of the metropolitan counties had higher levels of employment in wholesale and retail trade than the state average in 1978, but that share dropped to only 40.9% in 1987 (Tables 3.16 and 3.17). womue Hamumm ocm mammmaonz "wucwfluooo cofiumooq .H.m madman one; Exam 6.3 2322.3 "3552.0 553:.— 45 Thirty-six counties had a higher level of employment in manufacturing than the state average in 1978: and 37 counties did in 1987 (Figure 3.2). These counties were predominantly metropolitan or adjacent nonmetropolitan urban counties. Half of all metropolitan counties had higher than state average levels of employment in manufacturing for both years. Forty-one percent of all nonmetropolitan counties had higher than state average levels of employment in manufacturing in 1978, and 42.6% did in 1987. So although metropolitan counties suffered higher job losses in manufacturing than nonmetropolitan counties, manufacturing was still more concentrated in and around metropolitan counties. Nineteen counties in the state had a higher level of employment in FIRE than the state average for both years (Figure 3.3). 36.4% of all metropolitan counties had location quotients for FIRE greater than one in 1978; that share dropped to 31.8% in 1987 (Tables 3.16 and 3.17). Few nonmetropolitan counties had higher levels of employment in FIRE than state average. Nonmetropolitan counties with higher than state average employment in FIRE tended to be regional centers remote from metropolitan counties and amenity counties with real estate development potential. msfiusuomuscmz "wucmfiuozo newumooq .N.m wusowm no- so D 3 - no as - 3 I ”5383:52 "35325 553.5 47 wumuwm Hmmm. 0cm .wocmusmcH .wocmcwh "mucmfiuoso sewumooq .m.m whom?“ 2.3. B 3-2 8.1-3 I 3.7. 83mm 13% use .3523:— .oo:a=_.m Hecate—.0 :33qu 48 Table 3.16. Geographic Distribution of Counties with Location Quotients Greater than One in 1978. Manufac- . Trade turing FIRE County Type ' No. % No. % No. % Metropolitan 11 50.0 11 50.0 4 36.4 Nonmetropolitan 41 67.2' 25 41.0 11 18.0 Nonmetro Rural 34 75.6 13 28.9 11 24.4 Nonmetro Urban 7 43.8 12 75.0 0 0.0 Table 3.17. Geographic Distribution of Counties with Location Quotients Greater than One in 1987. Manufac- Trade turing FIRE County Type No. % No. % No. % Metropolitan 9 40.9 . 11 50.0 7 31.8 Nonmetropolitan 45 73.8 26 42.6 12 19.7 Nonmetro Rural 37 82.2 14 31.1 11 24.4 Nonmetro Urban 8 50.0 12 75.0 1 0.1 While some change occurred in the location quotient for individual counties (Appendix C), an examination of the change in average location quotients per county type shows that little change occurred in the relative specialization of employment within Michigan between 1978 and 1987 (Table 3.18). Table 3.18. 49 Change in Average Location Quotients LQMF1 LQMF Change in County Type 1978 1987 LQMF Michigan 0.92110 0.96305 0.04195 Metropolitan Co. 1.01257 1.06875 - 0.05618 Nonmetropolitan 0.95599 1.01037 0.05438 Nonmetro Rural 0.81322 0.83066 0.01744 Nonmetro Urban 1.09876 1.19008 0.09132 LQTR1 LQTR Change in County Type 1978 1987 LQTR Michigan 1.15216 1.12192 -0.03024 Metropolitan Co. 1.03963 1.03886 0.00078 Nonmetropolitan 1.15084 1.11437 -0.03647 Nonmetro Rural 1.23898 1.19326 -0.04572 Nonmetro Urban 1.06269 1.03547 -0.02722 LQFI1 LQFI Change in County Type 1978 1987 LQFI Michigan 0.90630 0.86297 -0.04333 Metropolitan Co. 1.00377 0.94646 -0.05731 Nonmetropolitan 0.83298 0.80280 -0.03018 Nonmetro Rural 0.91327 0.86603 -0.04724 Nonmetro Urban 0.75269 0.73956 -0.01312 1 LQMF is the location quotient for manufacturing employment. LQTR is the location quotient for wholesale and retail trade employment. LQFI is the location quotient for FIRE. On average, counties were slightly more specialized in manufacturing in 1987 than in 1978, with the largest change occurring in nonmetropolitan urban counties (Table 3.18). Metropolitan counties were the only county type to experience a positive change in average location quotients for wholesale and retail trade. As the figures in Table 3.18 indicate, this change was slight. Nonmetropolitan counties on average became relatively less specialized in 50 wholesale and retail trade in 1987 than they were in 1978. All county types on average became less specialized in FIRE in 1987 than they were in 1978. T-tests indicate that a significant difference existed in the location of counties specializing in FIRE employment for both years (Table 3.19). A significant difference existed between metropolitan and nonmetropolitan counties in specialized manufacturing employment in 1978, but not in 1987 (Table 3.19). No significant difference existed in wholesale and retail trade employment specialization between metropolitan and nonmetropolitan counties for either year. Table 3.19. Metropolitan Nonmetropolitan Variation in Employment Specialization, 1978 and 1987. 1978 Nonmetro Metro F’ Prob > F' LQMF 0.888 1.012 2.63 0.0164 LQTR 1.192 1.039 1.89 0.1070 LQFI 0.871 1.003 2.35. 0.0105 1987 Nonmetro Metro F' Prob > F' LQMF 0.924 1.068 1.47 0.327 LQTR 1.152 1.039 1.38 0.417 LQFI 0.833 0.947 2.23 0.016 3.4 Summary Between 1978 and 1987, per capita income increased in all Michigan counties, except St. Clair. This increase in income was accompanied by an increase in metropolitan- nonmetropolitan variations in per capita income. 51 Although the number of establishments increased in all industrial sectors during this time period, employment in some sectors decreased. Technical changes have reduced the number of workers needed in some sectors (Harrison and Blustone, 1990): and many of the new establishments were smaller, employing fewer workers than older establishments, which were relocating and/or downsizing. So even though counties were able to attract new businesses, many of them still lost jobs. . i The highest job losses occurred in manufacturing, with .metropolitan counties experiencing the highest losses. Service sector employment increased, but the service sector in Michigan was still smaller than the national aggregate in 1987. Most of Michigan’s high tech manufacturing and FIRE employment was located in metropolitan counties, while retail trade employment comprised a larger percentage of nonmetropolitan employment. Employment diversity increased in MiChigan during the study period. While changes in metropolitan entropy values were fairly homegeneous across counties, large variations existed in the direction and magnitude of changes in nonmetropolitan entropy values. Although the industrial structure in Michigan is changing, a high similarity existed between the industrial structure present in 1978 and the industrial structure present in 1987. Significant metropolitan nonmetropolitan differences exists in the location of counties specialized in 52 manufacturing and FIRE employment. Counties specialized in manufacturing and FIRE employment are predominately - metropolitan. Although a higher percentage of nonmetropolitan counties are specialized in wholesale and retail trade employment than metropolitan counties, no significant difference exists in the location of counties specialized in wholesale and retail trade. Va re 513 CHAPTERPOUR ANALYSES Although Michigan is experiencing a sectoral shift from ' manufacturing to services, manufacturing continues to be an important employment sector. Significant metropolitan- nonmetropolitan variation exists in employment specialization in manufacturing and FIRE industries. No significant difference existed betweeen metropolitan and nonmetropolitan income per capita in 1978, but a significant difference did exist in 1987. Regression analysis is used to examine the association between manufacturing and service employment specialization and income per capita in Michigan counties for 1978 and 1987. The initial model is specified as follows: where INC is income per capita, LQMF is the location quotient for manufacturing employment, LQTR is the location quotient for wholesale and retail trade employment, LQFI is the location quotient for FIRE employment, b0, b1, and b3 are the regression coefficients, and e is the error term. First, metropolitan-nonmetropolitan variation in the above association is examined for 1978 and 1987. Second, temporal variation in the above association is examined. Stepwise regression procedure is used to obtain the most parsimonious spatial and temporal models. 53 54 The correlation matrix for the regression variables show that the location quotient for wholesale and retail trade employment (LQTR) is the only variable inversely correlated with income per capita (INC) for both years (Table 4.1 and 4.2). The location quotients for employment in FIRE (LQFI) and manufacturing (LQMF) are positively, correlated with income per capita for both years. The location quotients for wholesale and retail trade and FIRE employment are inversely correlated with the location V quotient for manufacturing employment. Table 4.1. Pearson Correlation Matrix for 1978 Equations. INC LQMF LQTR LQFI INC 1.000 0.193 -0.355 0.037 LQMF 0.193 1.000 -0.687 -0.427 LQTR -0.355 -0.687 1.000 0.535 LQFI 0.037 -0.427 0.535 1.000 Table 4.2. Pearson Correlation Matrix for 1987 Equations. INC LQMF LQTR LQFI INC 1.000 0.188 -0.307 0.066 LQMF 0.188 1.000 -0.702 -0.474 LQTR -0.307 -0.702 1.000- 0.434 LQFI 0.066 -0.474 0.434 1.000 - 4.1 Spatial variation The metropolitan-nonmetropolitan variation in the association between levels of specialization in manufacturing and service employment and income per capita 55 (see Equation 1) is examined using a dummy variable and the interaction between the dummy variable and the independent variables. The spatial model is specified as follows: INC = boo + 130114 + bloLQMF + bllM.LQMF + bzoLQTR + b21M.LQTR + b30LQFI + b31M.LQFI + e (2) where M is the dummy variable; M is 0 for nonmetropolitan counties and M is 1 for metropolitan counties: M.LQMF, M.LQTR, and M.LQFI are interaction terms between the dummy variable and the independent variables described above: boo, b01, blo, bllr b20r b21, b30, and b31 are the regression coefficients, and e is the error term. This model will be estimated using stepwise regression procedure to obtain a parsimonious model. If in the estimation process, any one of the coefficients associated with the interaction term is significant, then the model can be regarded as showing spatial variation. 4.2 Temporal variation Temporal variation in the association in Equation (1)_ is examined first by estimating Equation (1) for 1978 and 1987. Next, a model is developed to explain how the association between the levels of employment specialization and real income per capita changed over time. The model is specified as follows: 56 INC = boo + b01T + bloLQHF +-b11T.LQMF + bzoLQTR + b21T.LQTR + baoLQFI + b31ToLQFI + e (3) In this equation T is the dummy variable. T is 0 for 1978 and 1 for 1987: and T.LQMF, T.LQTR, and T.LQFI are interaction terms between the dummy variable and the independent variables. This model will also be estimated using stepwise regression procedure to obtain a parsimonious model. If in the estimation process, any one of the coefficients associated with the interaction term is significant, then the model can be regarded as showing temporal variation. 4.3 Results If the regression coefficient associated with the location quotient for employment is significant and positive, then employment specialization in that sector is associated with higher levels of income per capita. If the regression coefficient associated with the location quotient for employment is significant and negative, then employment spcialization in that sector is associated with lower levels of income per capita. Spatial Variation The estimated results of equation (2) are as follows. The numbers in parenthesis are the t statistics. 57 121& (4) INC = 7174.843 -933.740LQTR + 820.451M.LQFI + 1150.204M.LQMF (15.33) (-2.45) - (2.87) (3.48) R2 = .576 1252 (5) INC = 7654.541 + 4772.836M - 4069.010MLQTR + 1870.201MLQFI (50.54) (3.47) (-2.43) (2.14) R2 = .464 In 1978, specialization in wholesale and retail trade employment associated with lower levels of income per capita in general. Employment specialization in manufacturing and FIRE in metropolitan counties associated with higher levels of income per capita in 1978 (Table 4.3). Table 4.3. Metro-Nonmetro Variation in the Association Between Employment Specialization and Income per Capita. 1978 County Type Variable Metropolitan Nonmetropolitan Intercept 7174.843 7174.843 LQMAN 1150.204 - 'LQTRAD - -933.740 LQFIRE 1150.204 - 1987 County Type Variable Metropolitan Nonmetropolitan Intercept 12427.33? 7654.541 LQTR -4069.040 - LQFI 1870.201 - 58 In 1987, employment specialization in wholesale and retail trade was negatively associated with income per capita in metropolitan counties, while employment specialization in the FIRE sector associated with higher levels of income per capita in these counties. No nonmetropolitan variables were selected in the stepwise regression procedure in 1987. Temporal Variation The estimated results of equation (1) are as follows. The numbers in parenthesis are the t statistics. 1212 (6) INC = 8969.066 - 203.992LQMF - 2694.895LQTR + 1047.995LQFI R2 = .200 1221 (7) INC = 9996.206 + 119.236LQMF - 2644.316LQTR + 1302.512LQFI (6.12) (.20) (-2.66) (2.10) R2 = .143 Specialization in wholesale and retail trade employment associated with low income per capita for both years. Specialization in FIRE employment, which is a high order service sector, associated with high income per capita. No clear association existed between specialization in 59 manufacturing employment and income per capita for 1978 or 1987. However, the results from Equation (3) show that specialization in manufacturing employment in 1987 does help explain changes in income per capita over time. INC = (15.16) (-4.68) + 656.549T.LQMF (2.21) R2 = .362 8604.055 - 2305.956LQTR + 786.112LQFI + 1051.187T.LQFI (2.10) (3.11) (8) Specialization in manufacturing and FIRE employment associated positively with income per while specialization in wholesale and employment associated negatively with capita over time, retail trade income per capita over time. Table 4.4. Temporal Variation in the Association Between Employment Specialization and Real Income per Capita. Variables Time 1 (1978) Time 2 (1987) Intercept 8604.055 8604.055 LQTR -2305.956 - LQFI 786.112 1837.299 LQMF - 656.549 The regression analysis results indicate that spatial variation does exist in the association of employment specialization and income per capita between metropolitan and nonmetropolitan counties. Employment specialization in 60 wholesale and retail trade associated with lower per capita income in both county types in 1978, while employment specialization in manufacturing and FIRE associated with higher income per capita in metropolitan counties. In 1987 metropolitan counties specialized in wholesale and retail trade sector employment associated with lower levels of income per capita, while metropolitan counties specialized in finance, insurance, and real estate sector employment associated with higher levels of income per capita. No nonmetropolitan variables were selected in the stepwise regression procedure in 1987. This may be due to the higher levels of nonemployment income in nonmetropolitan counties for that year. The increase in the percentage of income derived from nonemployment decreased the correlation between income and earnings per capita to .810. While a correlation of .810 is still high, a regression model using earnings per capita as the dependent variable may have yielded different results. The temporal analysis indicates that counties specializing in wholesale and retail trade employment will have lower income per capita, while counties specializing in manufacturing and FIRE employment will have higher income per capita. Specialization in manufacturing and FIRE employment was more important in explaining variations in income per capita in 1987 than in 1978. CHAPTER FIVE CONCLUSIONS This study examined trends in industrial transformation in Michigan’s metropolitan and nonmetropolitan counties between 1978 and 1987. Analyses of the spatially and temporally varying associations of sectoral employment specialization with income per capita provided an understanding of the metropolitan-nonmetropolitan variation in the effects of growth in the service sector. Growth trends in employment and establishments indicate that Michigan is experiencing a sectoral shift from manufacturing to service industries. Although the number of manufacturing establishments grew in Michigan during the study period, the number of manufacturing employees decreased, indicating that new manufacturing establishments have not guaranteed a growth in Michigan's manufacturing employment. Many new manufacturing plants are smaller employing fewer workers than older establishments: and many older establishments are relocating and/or downsizing. In spite of this decline, manufacturing is still the largest employment sector in the state. Service sector employment increased in all service categories in Michigan during the study period, thereby comprising a larger proportion of the workforce in 1987 than in 1978. The largest increases in service sector employment occurred in the retail and other services sectors. Other 61 62 services includes services such as personal, health, recreation, and hotel. However, since many service sector jobs are part-time and County Business Patterns does not distinguish between full-time and part-time employment, the contribution of service sector jobs to the total employment base may be overestimated. Michigan's industrial structure varied between metropolitan and nonmetropolitan counties. Counties specializing in manufacturing employment in Michigan were predominantly metropolitan or adjacent nonmetropolitan urban counties, while counties specializing in wholesale and retail trade employment were more likely to be nonmetropolitan. Many manufacturing firms which decentralized production moved production facilities out of state, rather than to nonmetropolitan counties within the state. While nonmetropolitan counties were more likely to be specialized in wholesale and retail trade than metropolitan counties, no significant difference existed in the level of specialization in wholesale and retail trade between metropolitan and nonmetropolitan counties. Metropolitan counties were more likely to be specialized in finance, insurance, and real estate than nonmetropolitan counties. Significant metropolitan nonmetropolitan variation existed in the level of specialization in finance, insurance and real estate for both years. Income per capita was higher in metropolitan counties than nonmetropolitan counties for both years, but this 63 difference was statistically significant only in 1987. Regression results showed that counties specializing in manufacturing and FIRE associated with higher income per capita, while counties specializing in wholesale and retail trade associated with lower income per capita. The regression analysis also indicated that spatial variation exists in the association of employment specialization and per capita income between metropolitan and nonmetropolitan counties. Employment specialization in wholesale and retail trade associated with lower income per capita in both county types in 1978, while employment specialization in manufacturing and FIRE associated with higher income per capita only in metropolitan counties. In 1987, metropolitan counties specializing in wholesale and retail trade sector employment associated with lower levels of income per capita, while metropolitan counties specializing in FIRE employment associated with higher levels of income per capita. No significant association existed between employment specialization and income per capita in nonmetropolitan counties in 1987. This may be due to higher levels of nonemployment income in nonmetropolitan counties for that year. In 1987, almost half of the per capita income in nonmetropolitan rural counties came from nonemployment income, compared to one third for nonmetropolitan urban counties, and 29% for metropolitan counties. This increased share in nonemployment income may 64 account for the diminished role of economic activities in contributing toward higher levels of per capita income. Results suggest that growth in the number of wholesale and retail establishments in Michigan will increase the employment base. However, since these jobs are low-wage and often part-time, they are not likely to contribute significantly to per capita income growth. The results from. the regression analysis confirm that counties specializing in wholesale and retail trade employment are associated with lower levels of income per capita relative to other counties in the state. Counties specializing in manufacturing and FIRE, on the other hand, associated with relatively higher levels of income per capita. This suggests that increasing employment in these sectors would Contribute to per capita income growth, particularly in the metropolitan counties. Michigan did experience an increase in the number of manufacturing establishments during the study period. However, this increase in the number of establishments did not prevent a decline in the number of manufacturing employees. The FIRE sector is a small, but growing portion of the Michigan workforce. The literature (Goe, 1990: Deavers, 1991; Mazie and Killian, 1991) and data suggest that metropolitan counties have a comparative advantage in attracting establishments in this sector. Twenty-four percent of nonmetropolitan rural counties also had higher levels of employment in FIRE than the state average, which suggests 65 that nonmetropolitan counties can attract employment in this sector also. However, nonmetropolitan counties are likely to attract branch offices, which perform low order service functions, and therefore do not generate the same level of income as may be expected for this sector. The regression results showed no significant association between employment specialization in FIRE and per capita income in nonmetropolitan counties. Future research on the consequences of sectoral employment specialization can be conducted at the national scale including all U.S. metropolitan and nonmetropolitan counties. Furthermore, the service sector should be disaggregated to include low order and high order producer services, consumer services and government services. Spatially and temporally varying associations between sectoral specialization and income growth will provide an understanding of the much debated role of the service sector in economic growth and development. APPENDICES 66 Appendix A. Change in Employment Entropy by County Relative Change Employment Entropy in 1978 1987 Entropy Metropolitan Counties Bay 72.59334 75.86407 3.27 Berrien 68.45241 70.75941 2.31 Calhoun 69.72824 72.63964 2.91 Clinton 79.80365 80.49980 0.70 Eaton 79.24817 76.57311 -2.68 Genesee 65.11228 67.55609 2.44 Ingham 75.97192 78.07927 2.11 Jackson 74.66495 77.08547 2.42 Kalamazoo 72.06938 73.49707 1.43 Kent 76.22687 77.26620 1.04 Lapeer 74.97229 72.90789 -2.06 Livingston 76.42201 77.26264 0.84 Macomb 63.79807 69.18040 5.38 Midland 54.35638 60.68467 6.33 Monroe 72.10636 76.48230 4.38 Muskegon 72.30092 72.79702 0.50 Oakland 79.46150 78.85147 -0.61 Ottawa 66.35210 68.71649 2.36 Saginaw 70.24888 74.99119 4.74 St. Clair 82.08117 78.94162 -3.14 Washtenaw 70.83764 70.39898 -0.44 Wayne 73.24802 76.69096 3.44 Nonmetropolitan Urban Counties Allegan 60.92726 62.00719 1.08 Barry 65.50331 71.57562 6.07 Branch 66.30475 68.49940 2.19 Cass 59.23625 65.04807 5.81 Gratiot 76.51702 76.68765 0.17 Hillsdale 63.17194 66.26621 3.09 Ionia 61.17707 68.65535 7.48 Isabella 78.71965 76.26590 -2.45 Lenawee 64.18860 67.45609 3.27 Marquette 82.18158 79.04691 -3.13 Montcalm 63.70567 63.59099 -0.11 Newaygo 70.15641 71.19166 1.04 Shiawasse 66.10517 68.93561 2.83 St. Joseph 70.47244 72.82527 2.35 Tuscola 76.07662 75.58473 -0.49 Van Buren 68.58573 74.94815 6.36 67 Appendix A (cont'd) Nonmetropolitan Rural Counties Alcona Alger Alpena Antrim Arenac Baraga Benzie Charlevoix Cheboygan Chippewa Clare Crawford Delta Dickinson Emmet Gladwin Gogebic Grand Traverse Houghton Huron Iosco Iron Kalkaska Keweenaw Lake Leelanau Luce Mackinac Manistee Mason Mecosta Menominee Missaukee Montmorency Oceana Ogemaw Ontonagon Osceola Oscoda Otsego Presque Isle Roscommon ‘ Sanilac Schoolcraft Wexford ' 61.45122 47.27074 74.82241 64.49357 72.68775 65.31294 69.83168 67.67434 75.28262 79.51725 75.55737 69.74454 73.35298 85.99646 75.55818 73.41076 73.66464 80.83282 78.46973 78.56284 77.04513 75.90978 85.85982 73.91803 70.21590 71.85560 74.21798 83.72862 65.40521 70.11648 72.21117 66.32979 87.66909 56.61645 77.35609 70.51595 61.12360 58.54061 61.30373 74.80658 76.05949 66.67840 74.88097 58.03861 72.01909 65.09959 62.49175 82.81649 72.35568 72.46211 71.11793 72.54943 72.68956 73.95201 73.93579 72.06635 64.42448 73.47217 81.47088 73.94023 71.67717 71.96065 81.74339 76.65636 75.20527 77.13506 73.43621 82.50232 74.42386 75.01457 70.77796 74.10111 76.81700 74.44395 75.74495 71.74255 72.91581 78.07616 71.45575 77.81389 74.14078 75.60382 66.79627 66.12240 77.89336 85.92864 67.18048 78.11347 64.05317 72.10462 3.65 15.22 7.99 7.86 -0.23 5.80 2.72. 5.02 -1.33 -5.58 -3.49 -5.32 0.12 -4.53 -1.62 -1.73 -1.70 0.91 -1.81 -3.36 0.09 -2.47 -3.36 0.51 4.80 -1.08 -0.12 -6.91 9.04 5.63 -0.47 6.59 -9.59 14.84 0.46 3.62 14.48 8.26 4.82 3.09 9.87 0.50 3.23 6.01 0.09 68 Appendix B. Change in Establishment Entropy by County Relative Change Establishment Entrophy in 1978 1987 Entropy Metropolitan Counties Bay 77.79659 76.24495 -1.55 Berrien 80.17967 78.35391 -1.83 Calhoun 77.79314 70.09778 -7.70 Clinton 82.08196 80.94462 -1.14 Eaton 78.45402 76.81712 -1.64 Genesee 73.74501 71.72137 -2.02 Ingham 76.37366 73.50626 -2.87 Jackson 80.90594 78.24403 -2.66 Kalamazoo 79.17145 77.26760 -1.90 Kent 82.24072 80.59554 -1.65 Lapeer 81.32722 81.01219 -0.32 Livingston 81.24935 79.77011 -1.48 Macomb 79.93557 79.02732 -0.91 Midland 75.43796 73.52879 -1.91 Monroe 79.77753 79.37415 -0.40 Muskegon 77.49029 76.08789 -1.40 Oakland 81.33286 78.29550 -3.04 Ottawa 82.66756 81.35823 -1.31 Saginaw 77.73565 75.45896 -2.28 St Clair 78.49533 78.73376 0.24 Washtenaw 76.29162 74.32557 -1.97 Wayne 77.54782 75.93052 -1.62 Nonmetropolitan Urban Counties Allegan 82.64651 78.62428 -4.02 Barry 78.32929 77.80953 -0.52 Branch 79.38360 79.60827 0.22 Cass 78.25665 79.96177 1.71 Gratiot 78.73256 78.09178 -0.64 Hillsdale 81.20747 81.52424 0.32 Ionia 77.48483 78.19139 0.71 Isabella 79.32846 77.63201 -1.70 Lenawee 77.91652 76.30655 -1.6l Marquette 75.66988 75.39663 -0.27 Montcalm 79.65458 78.79434 -0.86 Newaygo 79.13347 78.89573 -o.24 Shiawassee 82.59376 80.86793 -l.73 St Joseph 80.16015 76.10220 -4.06 Tuscola 79.77241 78.59219 -1.18 Van Buren 80.17234 79.37759 -0.79 69 Appendix B (cont'd) Nonmetropolitan Rural Counties Alcona Alger Alpena Antrim Arenac. Baraga Benzie Charlevoix Cheboygan Chippewa Clare Crawford Delta Dickison Emmet Gladwin Gogebic 'Grand Traverse Houghton Huron Iosco Iron Kalkaska Keweenaw Lake Leelanau Luce Mackinac Manistee Mason Mecosta Menominee Missaukee Montmorency Oceana Ogemaw Ontonagon Osceola Oscoda Otsego Presque Isle Roscommon Sanilac Schoolcraft Wexford 70.84133 76.79557 80.65436 78.44437 80.77228 82.48341 73.86695 79.26653 74.87204 77.87930 76.15425 76.37662 78.06125 81.21837 76.12357 73.85649 75.12160 79.83352 78.83567 81.61666 76.42943 79.53795 83.78438 67.71945 75.17676 78.63193 78.88942 72.28150 77.71028 77.07690 78.72645 83.98276 81.45536 78.32502 77.54110 80.56881 76.09141 78.85545 73.14889 81.18424 79.80154 68.31536 79.84426 78.27778 81.23160 73.43120 77.02991 80.15906 82.48892 79.39158 78.72770 76.75248 77.82703 73.21088 73.98892 77.11653 66.22577 78.97835 79.07615 76.12769 76.83327 74.54089 80.33562 75.59984 79.73494 76.33898 77.50027 87.43943 71.61369 71.42951 77.18904 77.70478 71.77199 77.39337 76.63218 77.06365 83.14731 89.86670 77.31272 77.78664 80.55084 74.44338. 79.82524 75.13018 82.31628 80.74858 73.26179 79.42701 76.28132 78.09029 2.59 0.23 -0.50 4.04 -1.38 -3.76 2.89 -1.44 -1.66 -3.89 0.96 -10.15 0.92 -2.14 0.00 2.98 -0.58 0.50 -3.24 -1.88 -0.09 -2.04 3.66 3.89 -3.75 -1.44 -1.18 -0.51 -0.32 -0.44 -1.66 -0.84 8.41 -1.01 0.25 -0.02 -1.65 0.97 1.98 1.13 0.95 4.95 -0.42 -2.00 -3.14 70 Change in Location Quotients for Manufacturing by County Appendix C. (a) 0.50841 LQMF LQMF Change In 1978 1987 LQMF Metropolitan Counties Bay 0.92498 0.82676 -0.09822 Berrien 1.16293 1.25187 0.08894 Calhoun 1.09249 1.02741 -0.06509 Clinton 0.73004 0.86469 0.13464 Eaton 0.60112 0.42904 -0.17207 Genesee 1.27051 1.53872 0.26821 Ingham 0.87154 0.86295 -0.00859 Jackson 0.77049 0.95065 0.18016 Kalamazoo 0.98335 1.03037 0.04702 Kent 0.89987 0.97054 0.07066 Lapeer 0.90797 1.11574 0.20777 Livingston 0.77020 0.86057 0.09037 Macomb 1.29663 1.34028 0.04365 Midland 1.82571 2.13742 0.31171 Monroe 1.00867 0.96801 -0.04066 Muskegon 1.08181 1.12378 0.04197 Oakland 0.69987 0.65356 -0.04630 Ottawa 1.28276 1.46898 0.18623 Saginaw 1.10531 1.03989 -0.06542 St. Clair 0.87764 0.97228 0.09464 Washtenaw 1.08646 1.13599 0.04953 Wayne 1.02630 0.94304 -0.08327 Nonmetropolitan Counties Alcona 1.09329 1.02803 -0.06526 Alger 2.29886 1.68975 -0.60911 Allegan 1.42069 1.62374 0.20304 Alpena 0.91356 0.85452 '-0.05903 Antrim 1.19967 1.28494 0.08527 Arenac 0.85020 0.57741 -0.27279 ‘ Baraga 1.23084 1.19757 -0.03327 Barry 1.22039 1.24331 0.02292 Benzie 0.87646 1.02019 0.14373 Branch 1.24930 1.48593 0.23663 Cass 1.43758 1.59645 0.15887 Charlevoix 1.20745 1.21735 0.00990 Cheboygan 0.73299 0.60509 -0.12790 Chippewa 0.24305 0.42798 0.18494 Clare 0.72429 0.60296 -0.12133 Crawford 0.51014 0.32143 -0.18871 Delta 0.89785 0.95179 0.05394 Dickinson 0.70976 0.90726 0.19750 Emmet 0.44674 0.53029 0.08356 Gladwin 0.68416 1.02550 0.34134 Gogebic 0.62943 -0.12102 Appendix C. (a) Grand Traverse Gratiot Hillsdale Houghton Huron Ionia Iosco Iron Isabella Kalkaska Keweenaw Lake Leelanau Lenawee Luce' Mackinac Manistee Marquette Mason Mecosta Menominee Missaukee Montcalm Montmorency Newaygo Oceana Ogemaw Ontonagon Osceola Oscoda Otsego Presque Isle Roscommon Sanilac Schoolcraft Shiawassee St. Joseph Tuscola Van Buren Wexford (cont'd) 0.53074 0.74715 1.38366 0.32939 0.84526 1.33622 0.67342 0.65913 0.37201 0.24820 1.21459 0.29279 0.33332 1.23999 0.60280 0.37997 1.31757' 0.14745 1.04206 0.79756 1.32335 0.44926 1.24829 1.41660 1.11009 0.69767 0.98796 0.39790 1.50406 1.23042 0.48876 0.19074 0.45789 1.31064 0.65954 1.09229 1.52435 0.91102 1.13967 0.96444 71 0.61765 0.70804 1.55902 0.40120 1.17056 1.40841 0.63782 0.55581 0.34831 0.92559 0.47494 0.60392 0.35077 1.33750 0.91325 0.21404 1.14301 0.16293 0.96666 0.77696 1.39745 0.59525 1.54725 0.82856 1.29684 0.94114 0.82096 0.52783 1.60467 1.31664 0.69252 0.40090 0.35734 1.46521 0.77398 1.03635 1.65612 0.96730 1.06383 1.15446 0.08691 -0.03910 0.17536 0.07180 0.32531 0.07219 -0.03560 -0.10332 -0.02370 0.67738 -0.73964 0.31113 0.01745 0.09751 0.31045 -0.16593 -0.17456 0.01548 -0.07540 -0.02060 0.07409 0.14599 0.29896 -0.58803 0.18675 0.24347 -0.16700 0.12993 0.10061 0.08623 0.20376 0.21017 -0.10054 0.15457 0.11444 -0.05595 0.13177 0.05628 -0.07584 0.19001 72 Change in Location Quotients for Wholesale and Retail Trade by County Appendix C. (b) LQTR LQTR Change in 1978 1987 LQTR Metropolitan Counties Bay 1.21652 1.25135 '0.03483 Berrien 0.94524 0.89943 -0.04581 Calhoun 0.89032 0.97574 0.08541 Clinton 1.37951 1.30151 -0.07801 Eaton 1.51663 1.71129 0.19466 Genesee 0.94405 0.99164 0.04758 Ingham 1.09033 0.98983 -0.10050 Jackson 0.81776 0.98933 0.17158 Kalamazoo 1.05068 0.99064 -0.06004 Kent 1.11584 1.05660 -0.05924 Lapeer 1.26095 1.25712 -0.00384 Livingston 1.26999 1.05137 -0.21862 Macomb 0.91953 0.97562 0.05609 Midland 0.65960 0.66651 0.00692 Monroe 0.93022 0.95876 0.02854 Muskegon 0.89924 0.97590 0.07667 Oakland 1.16031 1.06152 -0.09880 Ottawa 0.89676 0.82466 -0.07209 Saginaw 1.02046 1.06939 0.04892 St. Clair 1.18665 1.05880 -0.12785 Washtenaw 0.80606 0.89871 0.09265 Wayne 0.89528 0.89918 0.00390 Nonmetropolitan Counties Alcona 1.20091 1.35332 0.15240 Alger 0.81580 0.64635 -0.16945 Allegan 0.77386 0.79614 0.02228 Alpena 1.24474 1.18933 -0.05541 Antrim 0.99054 0.86621 -0.12433 Arenac 1.35289 1.31678 -0.03611 Baraga 0.98110 1.03897 0.05787 Barry“ 0.91825 0.96095 0.04270 Benzie 0.96313 1.03589 0.07276 Branch 1.17659 1.00315 -0.17344 Cass 0.88609 0.85947 -0.02662 Charlevoix. 0.78881 0.86847 0.07966 Cheboygan 1.35247 1.28452 -0.06796 Chippewa‘ 1.57363 1.39805 -0.17557 Clare 1.31143 1.50194 0.19050 Crawford 1.35466 1.53375 0.17909 Delta 1.12076 1.21597 0.09522 Dickinson 1.19257 1.11337 -0.07920 Emmet 1.26790 1.16152 -0.10638 Gladwin 1.45767 1.23117 -0.22650 Gogebic 1.38852 1.21112 -0.17740 Appendix C. (b) (cont'd) Grand Traverse 1.28662 Gratiot 1.25817 Hillsdale 0.88578 Houghton 1.38243 Huron 1.28309 Ionia 0.94818 . Iosco 1.46171 Iron 1.46858 Isabella 1.81294 Kalkaska 1.31793 Keweenaw 0.99129 Lake 1.82033 Leelanau 1.03075 Lenawee 0.98490 Luce 1.79697 Mackinac 1.45981 Manistee 0.89370 Marquette 1.25096 Mason 1.02253 Mecosta 1.49302 Menominee 0.97077 Missaukee 1.44451 Montcalm 0.98485 Montmorency 1.15479 Newaygo 1.04280 Oceana 1.39732 Ogemaw 1.21882 Ontonagon 0.59304 Osceola 0.83885 Oscoda 1.19153 Otsego 1.44389 Presque Isle 1.16993 Roscommon 1.99992 Sanilac 0.97044 Schoolcraft 1.26189 Shiawassee 1.06478 St. Joseph 0.73266 Tuscola 1.29217 Van Buren 0.99005 Wexford 1.03214 73 1.09865 1.18255 0.87468 1.31281 1.10494 0.95573 1.56620 1.47588 1.39926 1.00261 1.00082 1.39554 0.98471 0.88229 1.49470 1.38272 1.13350 1.24466 1.03668 1.40480 0.88175 1.10296 0.82811 1.21431 1.05158 1.24152 1.40465 0.78177 0.95132 1.06342 1.19361 1.27181 2.02301 0.89243 1.28181 1.19274 0.90439 1.28467 1.14721 1.03101 -0.18797 -0.07562 -0.01110 -0.06962 -0.17815 0.00755 0.10449 0.00730 -0.41368 -0.31532 0.00954 -0.42479 -0.04604 -0.10262 -0.30227 -0.07709 0.23979 -0.00629 0.01415 -0.08821 -0.08901 -0.34155 -0.15674 0.05952 0.00878 -0.15580 0.18583 0.18873 0.11247 -0.12811 -0.25028 0.10188 0.02309 -0.07801 0.01992 0.12796 0.17173 -0.00750 0.15716 -0.00114 74 Appendix C. (c) Change in Location Quotients for FIRE by County' LQFI LQFI Change In 1978 1987 LQFI Metropolitan Counties Bay 0.84212 0.83438 -0.00774 Berrien 0.77561 0.77235 -0.00326 Calhoun 1.57635 1.46651 -0.10983 Clinton 1.28099 0.79695 -0.48404 Eaton 2.80109 2.25746 -0.54363 Genesee 0.67850 0.60558 -0.07291 Ingham 1.18226 1.19083 0.00857 Jackson 0.53584 0.75404 0.21820 Kalamazoo 0.85205 0.89324 0.04119 Kent 1.02805 0.94222 -0.08583 Lapeer 0.96586 1.02754 0.06168 Livingston 1.41660 ,1.24423 -0.17237 Macomb 0.52261 0.56797 0.04535 'Midland 0.52004 0.60791 0.08787 Monroe 0.71539 0.62881 -0.08658 Muskegon 0.65320 0.75895 0.10575 Oakland 1.55020 1.41826 -0.13194 Ottawa 0.63108 0.53397 -0.09711 Saginaw 0.87582 0.80763 -0.06819 St Clair 0.81219 0.82563 0.01345 Washtenaw 0.80889 0.73427 -0.07462 Wayne 1.05824 1.15337 0.09513 Nonmetropolitan Counties Alcona 0.18490 0.36233 0.17743 Alger 0.60725 0.58761 -0.01964 Allegan 0.43248 0.40761 -0.02487 Alpena 0.98822 0.90135 -0.08687 Antrim 0.55482 0.62866 0.07384 Arenac 0.99858 0.75520 -0.24338 Baraga 0.75528 0.80386 0.04859 Barry 0.99789 0.96814 -0.02975 Benzie 0.72766 0.96117 0.23351 Branch 0.80175 0.77118 -0.03057 Cass 0.76306 0.61752 -0.14554 Charlevoix 0.79961 0.64994 -0.14966 Cheboygan 0.97085 0.82353 -0.14732 Chippewa 1.32603 1.02581 -0.30022 Clare 0.99873 0.73326 -0.26547 Crawford 0.80123 0.78092 -0.02031 Delta 0.84148 0.87914 0.03766 Dickinson 0.85979 0.58353 -0.27626 Emmet 0.81565 0.79920 -0.01645 Gladwin 1.23970 1.12450 -0.11520 Gogebic 1.00350 0.92724 -0.07626 Appendix C. (c) Grand Traverse Gratiot Hillsdale Houghton Huron Ionia Iosco Iron Isabella Kalkaska Keweenaw Lake Leelanau Lenawee Luce Mackinac Manistee Marquette Mason Mecosta Menominee Missaukee Montcalm Montmorency Newaygo Oceana Ogemaw Ontonagon Osceola Oscoda Otsego Presque Isle Roscommon Sanilac Schoolcraft Shiawassee St Joseph Tuscola Van Buren Wexford (cont'd) 1.14450 0.84250 0.64099 1.46184 0.99294 0.67911 1.29226 1.11398 0.82425 0.55824 0.92549 2.10786 0.66351 0.70639 0.85353 0.84632 0.77153 0.96952 0.73071 1.13069 0.66520 2.13929 0.57690 0.88063 0.90650 0.97064 0.43771 0.48074 0.54412 0.90068 0.77485 0.91846 1.36000 0.74434 0.49818 0.78155 0.55671 0.79302 0.77037 0.71557 75 0.88900 0.71262 0.52507 1.43329 0.89431 0.76861 1.20551 1.17019 1.05274 0.21380 1.62733 1.20603 0.55856 0.83056 0.70093 0.86602 0.83715 0.97503 1.24995 0.90015 0.72866 1.15488 0.51414 1.27399 0.79512 0.83892 0.61807 0.52543 0.66650 0.60216 0.71423 0.99408 1.26382 0.99085 0.94784 0.80636 0.51792 0.94072 0.62969 0.57245 -0.25551 -0.12989 -0.11592 -0.02855 -0.09863 0.08949 -0.08675 0.05621 0.22849 -0.34444 0.70184 -0.90183 -0.10495 0.12418 -0.15260 0.01970 0.06561 0.00551 0.51925 -0.23055 0.06346 -0.98441 -0.06277 0.39335 -0.11138 -0.13172 0.18036 0.04469 0.12238 -0.29852 -0.06062 0.07563 -0.09618 0.24652 0.44966 0.02481 -0.03880 0.14770 -0.14069 -0.14311 BIBLIOGRAPHY BIBLIOGRAPHY Aiken, Charles S., 1990. "A New Type of Black Ghetto in the Plantation South", Annuals of the Association of American Geographers, Vol. 80, No. 2, p. 223-246. Alter, Theodore R. and Richard W. Long, ed., 1988. National Perceptions and Political Significance of Rural Areas: An International Symposium, University Park, Pennsylvania: Pennsylvania State University. Amos, Jr., Orley M, 1988. "Unbalanced Regional Growth and Regional Income Inequality in the Latter Stages of Development", Regional Science and Urban Economics, Vol. 18, p. 549-566. Amos, Jr., Orley M, 1989. An Inquiry into the causes of Increasing Regional Income Inequality in the United States, The Review of Regional Studies, Vol. 19, No. 2, p. 1-12. Angel, David P. and Jeffrey Mitchell, 1991. Intermetropolitan Wage Disparities and Industrial Change, Economic Geography, Vol 67, No. 2, pp. 124-135. Apedaile, L.P., 1991. "A Synthesis of Restructuring Issues for the Development of Rural Economies", Staff Paper No. 91- 01. Edmonton: University of Alberta. Barancik, Scott, 1990. The Rural Disadvantaged: Growing Income Disparities Between Rural and urban Areas, Washington, D.C.: Center on Budget and Policy Priorities. Barkley, David L. and John E. Keith, 1991. "The Locational Determinants of Western Nonmetropolitan High Tech Manufacturers: An Econometric Analysis", Western Journal of Agricultural Economics, Vol 16, No. 2, pp. 331-344. Blackley, Paul R., 1986. Urban-Rural Variations in the Structure of Manufacturing Production, urban Studies, Vol 23, pp. 471-483. Bonnen, James T., 1990. "The Political Economy of U.S. Rural Policy: An EXploration of the Past with Strategies for the Future", Staff Paper No. 90-54. E Lansing: Michigan State University. 76 77 Bonnen, James T., Glenn Nelson, and Kenneth Deavers, 1988, "A Synthesis". In National Perceptions and Political Significance of Rural Areas: An International Symposium, edited by Theodore R. Alter and Richard W. Long, University Park, Pennsylvania: Pennsylvania State University. Booth, Douglas E., 1986. "Long Waves and Uneven Regional Growth" Southern Economic Journal, Vol. 53, pp. 448-460. Bound, John and George Johnson, 1992. "Changes in the Structure of Wages in the 1980's: An Evaluation of Alternative Explanations", The American Economic Review, Vol 82, No. 3, pp. 371-392. Bradbury, J.H., 1985. Regional and Industrial Restructuring Processes in the New International Division of Labour, Progress in Human Geography, Vol 9, No 1, pp. 38-63. Bureau of Economic Analysis and U.S. Department of Commerce, 1981. Local Area Personal Income 1974-1979, Vol 4, Washington, D.C.: U.S. Government Printing Office. Bureau of Economic Analysis and U.S. Department of Commerce, 1990. Local Area Personal Income 1984-1989, Vol 2, Washington, D.C.: U.S. Government Printing Office. Castle, Emery, 1987. "Policy Options for Rural Development in a Restructured Rural Economy: An International Perspective". In Agriculture and Beyond: Rural Economic Development, edited by Gene Summers, John Bryden, Kenneth Deavers, Howard Newby, and Susan Sechler, Madison: University of Wisconsin, pp. 11-27. Clark, Gordon, 1992. "Tourism and Leisure Conflicts in the British Countryside: A Research Framework and Agenda". In Geographic Perspectives on the Social and Economic Restructuring of Rural Areas: Proceedings of the International Geographic Congress Commission on Changing Rural Systems, edited by M. Duane Nellis, Manhattan, KanSas: Kansas State University. Clarke, Roger, 1985. Industrial Economics, New York: Basil Blackwell. Cochrane, Willard W., 1979. The Development of American Agriculture: A Historical Analysis, Minneapolis: University of Minnesota Press. Commins, Patrick, 1990. "Restructuring Agriculture in Advanced Societies: Transformation, Crisis and Responses". In Rural Restructuring: Global Processes and Their Responses edited by Marsden, Lowe, and Whatmore. London: David Fulton Publishers. 78 Connaughton, John E. and Ronald A. Madsen, 1990. The Changing Regional Structure of the U.S. Economy, Growth and Change, Vol 21, No. 3, pp. 48-60. Crown, William H., 1991. ”Migration and Regional Economic Growth: An Origin-Destination Model", Economic Development Quarterly, Vol. 5, No. 1, p. 45-59. Cutler, David M. and Lawrence F. Katz, 1992. "Rising Inequality? Changes in the Distribution of Income and Consumption in the 1980's", The American EConomic Review, Vol. 82, No. 2, pp. 546-551. DaVanzo, Julie, 1983. "Repeat Migration in the United States: Who Moves Back and Who Moves On?”, Review of Ebonomics and Statistics, Vol. 65, No. 4, pp. 552-559. Deavers, Kenneth L., 1991. "1980's A Decade of Broad Rural Stress", Rural Development Perspectives, Vol. 7, No. 3, pp. 2-5 0 . de Janvry, Alain, David Runsten and Elisabeth Sadoulet, 1987. "Toward a Rural Development Program for the United States: A Proposal". In Agriculture and Beyond: Rural Economic Development, edited by Gene Summers, John Bryden, Kenneth Deavers, Howard Newby, and Susan Sechler, Madison: University of Wisconsin, p. 55-93. Dillman, Don A., 1992. "Information Technologies and Rural Economic Development". In National Rural Studies Committee: A Proceedings, Corvalis: Oregon State University, pp. 76- 100. ~ Drabenstott, Mark, 1991. "Rural America in the 19903: A Tale of Two Cities". In Rural Community Economic Development, edited by NCrman Walzer, New York: Praeger Publishers, pp. . Dubin, Robin, 1991. "Commuting and Firm Decentralization", Land Economics, Vol. 67, No. 1, pp. 15-29. Edwards, Clark, 1976. ”The Political Economy of Rural Development: Theoretical Perspectives", American JOurnal of Agricultural Economics, Vol 58, No. 5, pp. 914-921. Esparza, Adrian, 1990. "Manufacturing Decline and Technology Lags in Nonmetropolitan Illinois", Growth and Change, Vol. 21, No. 4, pp. 19-32. Fik, Timothy J., Robert G. Amey, and Edward J. Malecki, 1991. "Changing Employment Profiles and Growth: An Economic Base Study of Florida Counties (1982-1987)", Growth and Change, Vol 22, No 3, pp. 86-104. 79 Fitzsimmons, Margaret, 1986. "The New Industrial Agriculture: The Regional Integration of Specialty Crop Production”, EConomic Geography, Vol. 62, No. 4, pp. 334- 353. Flora, Cornelia Butler, 1990. "Presidential Address: Rural Peoples in a Global Economy", Rural Sociology, Vol. 55, No. 2, pp. 157-177.< Garrison, Charles B. and Albert S. Paulson, 1973. "An Entropy Measure of the Geographic Concentration of Economic Activity", Economic Geography, Vol. 49, No. 4, pp. 319-324. Gilchrist, Donald A. and Larry V. St. Louis, 1991. "Directions for Diversification with an Application to Saskatchewan", JOurnal of Regional Science, Vol. 31, No. 3, pp. 273-289. Glasmeier, Amy and G. Borchard, 1989. "Research Policy and review 31. From branch plants to back offices: prospects for rural services growth", Environment and Planning A, Vol. 21, p. 1565-1583. Glasmeier, Amy and Norman Glickman, 1990. "Foreign Investment Boosts Rural Economies", Rural Development Perspectives, June-September, p. 19-25. Glasmeier, Amy and Amy Kays-Teran, 1989. "Rural Area Development: Competing for the Chips", Forum for Applied Research and Public Policy, Fall 1989, p. 53-57. Glickman, Norman J. and Amy K. Glasmeier, 1989. "The International Economy and the American South". In Deindustrialization and Regional Economic Transformation: The Experience of the United States edited by Lloyd Rodwin and Hidehiko Sazanami. Boston: Unwin Hyman, pp. 60-80. Goe, W.R., 1990. "Producer Services, Trade and the Social Division of Labour”, Regional Studies, Vol. 24, No. 4, pp. 327-324. Grubb, W. Norton and Robert H. Wilson, 1992. "Trends in Wage and Salary Inequality, 1967-88", MOnthly Labor Review, Vol. 115, No. 6, pp. 23-39. Hady, Thomas F. and Peggy J. Ross, 1990. An Update: The Diverse Social and Economic Structure of Nonmetropolitan America, Economic Research Service. Washington, D.C.: U.S. Department of Agriculture, Economic Research Service. Staff Report No. AGES 9036. Hammond, R. and P.S. McCullagh, 1978. Quantitative Techniques in Geography. Oxford: Oxford University Press. 80 Harrison, Bennett and Barry Bluestone, 1990. The Great U- Turn: Corporate Restructuring and the Polarizing of America. New York: Basic Books, Inc. Hatton, Timothy J. and Jeffrey G. Williamson, 1992. "What Explains Wage Gaps between Farm and City? Exploring the Todaro Model with American Evidence, 1890-1941", Economic Development and Cultural Change, Vol. 40, No. 2, pp. 267- 294. Haynes, Kingsley E. and Zachary B. Machunda, 1987. "Spatial Restructuring of Manufacturing and Employment Growth in the Rural Midwest: An Analysis for Indiana", Economic Geography, Vol 63, No 4, pp. 319-333. Henry, Mark, Mark Drabenstott and Lynn Gibson, 1988. "A Changing Rural Economy". In Rural America in Transition edited by Henry, Drabenstott and Gibson. Kansas City: Federal Reserve Bank, pp. 15-37. Isserman, Andrew M., 1977. "The Location Quotient Approach to Estimating Regional Economic Impacts", JOurnal of the American Institute of Planners, Vol. 43, No. 1, pp. 33-41. Johnson, Kenneth M., 1989. "Recent Population Redistribution Trends in Nonmetropolitan America", Rural Sociology, Vol. 54, No. 3, pp. 301-326. Jones, Sue H., 1989. Focus on the Future: Options in Developing a New National Rural Policy. Rural Development Workshops. College Station: Texas A&M University System, Texas Agricultural Extension Service. Jones, John P. and Janet E. Kodras, 1990. "Restructured Regions and Families: The Feminization of Poverty in the U.S.", Annuals of the Association of American Geographers, Vol. 80, No. 2, pp. 163-183. Kale, Steven R., 1989. "Theoretical Contributions to the Understanding of U.S. Nonmetropolitan Economic Change", Economic Development Quarterly, Vol. 3, N01, p 58-69. Kellerman, Aharon, 1985. ”The Evolution of Service Economies: A Geographical Perspective", The Professional Geographer, Vol 37, No. 2, p. 133-143. Kirn, Thomas J., Richard S. Conway, Jr., and William B. Beyers, 1990. "Producer Services Development and the Role of Telecommunications: A Case Study in Rural Washington", Growth and Change, Vol. 21, No. 4, pp. 33-50. Kutscher, Ronald E. and Valerie A. Personick, 1986. "Deindustrialization and the Shift to Services", Monthly Labor Review, Vol. 109, pp. 3-13. 81 Lapping, M., T. Daniels, and J. Keller, 1989. Rural Planning and Development in the United States, New York: The Guilford Press. Levy, Frank and Richard J. Murnane, 1992. "U.S. Earnings~ Levels and Earnings Inequality: A Review of Recent Trends and Proposed Explanations", The JOurnal of Economic Literature, Vol. 30, No. 3, pp. 1333-1381. Lonsdale, Richard, 1982. "Non-metropolitan Industrial Employment and Rural Transformation in the United States", in The Effect of Modern Agriculture on Rural Development, by Gyorgy Enyedi and Ivan Volgyes, Elmsford, New York: Pergamon. Press. p. 281-294. Marsden, Terry, Philip Lowe, and Sarah Whatmore, ed., 1990. Rural Restructuring: Global Processes and Their Responses, London: David Fulton Publishers. Markusen, Ann R. and Virginia carlson, 1989. "Deindustrialization in the American Midwest: Causes and Responses”. In Deindustrialization and Regional Economic Transformation: the EXperience of the United States, edited by Lloyd Rodwin and Hidehiko Sazanami. Boston: Unwin Hyman, pp. 29-59. Massey, Doreen, 1984. Spatial Divisions of Labor: Social Structures and the Geography of Production, New YorkzMethuen, Inc. Mazie, Sara Mills and Molly Sizer Killian, 1991. "Growth and Change in Rural America: the Experience of the 1980s and Prospects for the 1990s". In Rural Community Economic Development, edited by Norman Walzer, New York: Praeger Publishers, pp. . McGranahan, David A., 1992. "Assessing the Research Framework and Institutional Context for Rural Development Policy", Southern JOurnal of Agricultural Economics, Vol. 24, No. 1, pp. 105-109. McNamara, Kevin T. and Lewell F. Gunter, "Off-farm Earnings: The Impact of Economic Structure", The Review of Regional Studies, pp. 1-9.(Citation incomplete) . Mead, Lawrence M., 1991. "Why are the Poor Still With Us: The New Politics of the New Poverty", The Public Interest, No. 103, pp. 3-20. Molnar, Joseph J. and Greg Traxler, 1991. "People Left Behind: Transitions of the Rural Poor", Southern Journal of Agricultural Economics, Vol. No. , p. 75-83. 82 Moriarty, B. M., 1991. "Urban Systems, Industrial restructuring, and the Spatial-Temporal Diffusion of Manufacturing Employment", Environment and Planning A, Vol. 23, No. , pp. 1571-1588. ' Noyelle, Thierry J. and Thomas M. Stanback, Jr., 1983. The Economic Transformation of American Cities, Totowa, New Jersey: Rowman 8 Allanheld. O’hUallachain, B., 1990. ”The Location of US Manufacturing: Some Empirical Evidence on Recent Geographical Shifts", Environment and Planning A, Vol. 22, No. 9, pp. 1205-1222. Patton, Wendy and Ann Markusen, 1991. "The Perils of Overstating Service Sector Growth Potential: A Study of Linkages in Distributive Services", Economic Development Quarterly, Vol. 5, No. 3, pp. 197-212. Phillips, Bruce D. and James P. Miller, 1991. ”Patterns of Employment Growth in Midwestern Metropolitan and Nonmetropolitan Counties". In Rural Community Economic Development, edited by Norman Walzer, New York: Praeger Publishers. Rainey, Kenneth D., 1976. "Forces Influencing Rural Community Growth", American Journal of Agricultural Economics, Vol. 58, No. 5, pp. 959-962. Ray, Dennis M., 1992. "Perspectives on Urban Economic Development", Entrepreneurship, Innovation, and Change, Vol. 1, No. 1, pp. 27-56. Redman, John M. and Thomas D. Rowley, 1989. State-Level Comparison of Metro and Nonmetropolitan Economic Performance, 1979-1986, Economic Research Service Staff Report AGES 89-60, Washington, D.C.: U.S. Government Printing Office. . Roepke, Howard G. and David A. Freundenberg, 1981. "The Employment Structures of Nonmetropolitan Counties", Annuals of the Association of American Geographers, Vol 71, No. 4, pp. 580-592. Smith, Eldon D., 1990. "Economic Stability and Economic Growth in Rural Communities: Dimensions Relevant to Local Employment Creation Strategy", Growth and Change, Vol. 21, No. 4, pp. 3-17. Smith, Steven M. and David L. Barkley, 1989. "Contributions of High-Tech Manufacturing to Rural Economies", Rural Development Perspectives, Vol. 5, No. 3, pp. 6-11. 83 Smith, Steven M. and David L. Barkley, 1991. "Local Input Linkages of Rural High-Technology Manufacturers", Land Economics, Vol. 67, No. 4, pp. 472-483. ~ Summers, Gene F., Francine Horton and Christina Gringeri, 1990. ”Rural Labour-Market Changes in the United States". In Rural Restructuring: Global Processes and Their Responses, edited by Terry Marsden, Philip Lowe, and Sarah Whatmore. London: David Fulton Publishers. Susman, Paul, 1989. ”Exporting the Crisis: U.S. Agriculture and the Third World", Economic Geography, Vol. 65, No. 4, p. 293-313. Swanson, Louis E., 1989. ”The Rural Development Dilemma", Resources, No. 96, pp. 14-16. Tam, Mo-Yin S. and Joseph Persky, 1982. ”Regional Convergence and National Inequality", The Review of Economics and Statistics, Vol. 64, No. 1, pp. 161-165. Tweeten, Luther and George L. Brinkman, 1976. Micropolitan Development: Theory and Practice of Greater-Rural Economic Development, Ames: The Iowa State University Press. U.S. Department of Commerce and Bureau of Census, 1980. County Business Patterns 1978: Michigan, Washington, D.C.: U.S. Government Printing Office. U.S. Department of Commerce and Bureau of Census, 1989. County Business Patterns 1987: Michigan, Washington, D.C.: U.S. Government Printing Office. Walker, Richard and Michael Storper, 1981. ”Capital and Industrial Location", Progress in Human Geography, Vol 5, No 4,. p0 473-5090 Wheeler, James O., 1990. "The New-Corporate Landscape: America's Fastest Growing Private Companies", The Professional Geographer, Vol. 42, No. 4, pp. 433-444. Wilson, William J., 1987. The Truly Disadvantaged: the Inner City, the Uhderclass, and Public Policy, Chicago: The University of Chicago Press. Young, Ruth C., 1986. "Industrial Location and Regional Change: The United States and New York State", Regional Studies, Vol 20, No 4, pp. 341-369. "ImmmmES