Rurality in Post Industrial Society

Paper prepared for the conference ‘New Forms of Urbanization: Conceptualizing and Measuring Human Settlement in the Twenty-first Century’, organized by the IUSSP Working Group on Urbanization and held at the Rockefeller Foundation’s Study and Conference Center in Bellagio, Italy, 11-15 March 2002. Paper 14 THE NATURE OF RURALITY IN POST INDUSTRIAL SOCIETY By David L. Brown and John B. Cromartie? Draft 2/15/02 INTRODUCTION Urbanization is a dynamic social and economic process that transforms societies from primarily rural to primarily urban ways of life (Hauser, 1965). Few would dispute this definition, but how useful is it for examining the spatial reorganization of population and economic activities in postindustrial societies where a large majority of people, jobs, and organizations are concentrated in or dominated by urban agglomerations? The essence of this question hinges on our ability to differentiate between what is rural and urban in postindustrial societies. While this may have been a relatively straightforward task during the late 19th and early to mid 20th centuries, it has become an exceedingly complex question in the context of postindustrialization. We acknowledge the helpful comments of Calvin Beale, Kai Schafft, Laszlo Kulcsar, and the conference organizers Tony Champion and Graeme Hugo. Brenda Creeley prepared the manuscript. Early social scientists saw urbanization and industrialization as being reciprocally related. One process could not proceed without the other. While most scholars understood that urban and rural were not ent irely discrete categories, relatively clear lines could be drawn to distinguish urban from rural communities and distinct ways of life associated with each. In addition, early social scientists were convinced that the transformation from rural to urban-industrial society would be accompanied by a wide range of negative social outcomes. In fact, this concern is generally credited with motivating the rise of the new discipline of Sociology (Marx, 1976; Durkheim, 1951; Weber, 1968; Wirth, 1938). The social and economic organization of community life has been thoroughly transformed by technological and institutional changes since the mid 20th century. Accordingly, notions of what constitutes urban and rural communities that grew out of the era of industrialization may no longer offer a reliable lens with which to view contemporary settlement structures. They may no longer provide a reliable delineation of what is urban and what is rural, and consequently we may not be able to determine whether the level of urbanization is advancing, declining, or remaining constant. As a consequence, our analyses of population redistribution may bear little connection to the reality of spatial reorganization. The large literature on counter-urbanization, to which we are both contributors, may be missing the mark because it depends on data systems and geo-coding schemes that reflect a prior era of socio-spatial organization. Hence, our purpose in this paper is to propose a multidimensional approach for conceptualizing rurality that reflects the demographic, social, economic and institutional realities of postindustrial society. We agree with Halfacree (1993: p. 4) that â€Å"†¦the quest for an all-embracing definition of the rural is neither desirable nor feasible,† but we believe that social science can and should develop conceptual frameworks and geo-coding schemes to situate localities according to their degree of rurality. Since rurality is a multidimensional concept, the degree of rurality should be judged against a composite definition that includes key social, economic and demographic attributes. This approach rejects the notion of rural as a residual (after urban has be en measured). The operationalization of rurality should be flexible enough to differentiate urban from rural, while recognizing and appreciating the diversity contained within each category. Our approach to defining 2 rurality involves the material aspects of localities, but we acknowledge the validity of other approaches. As Halfacree and others have observed, rurality can be defined as a social representation. Or as he puts it, â€Å"the rural as space, and the rural as representing space† should be distinguished (1993: 34). We do not propose to debate the relative merits of the material and representational approaches in this paper. Each has a respected tradition in social science. Our sociodemographic approach is inspired by previous work of Paul Cloke, 1977 and 1986, while the social representation approach’s pedigree includes Moscovici, 1981, Giddens, 1984, and many other highly respected scholars. We feel that these approaches are complementary rather than competitive. As Martin Lewis has observed, â€Å"In the end, only by combining the insights of the new geography with those of the traditional approaches may human relatedness be adequately reconceptualized† (1991: 608). However, we emphasize the socioeconomic approach in this paper because of its utility for informing statistical practice essential to the quantitative empirical study of urbanization. Why Do We Need To Know What Is Rural In Postindustrial Society? At the most basic level, urbanization cannot be understood without also examining the nature of rurality. Perhaps it is axiomatic, but urbanization cannot proceed in postindustrial society unless rural people and communities persist and are at risk of â€Å"becoming urban. While there is copious evidence that rural-urban differences have diminished during the latter half of the 20th century, important differences have been shown to persist structuring the lives people live and the opportunities available to them (Brown and Lee, 1999; Fuguitt, et al. , 1989). In addition, what we believe about rural people and communities sets the agenda for public policy. The American public, for example, holds a strong pro-rural and/or antiurban bias that provides continuing support for agricultural and rural programs (Kellogg Foundation, 2002; RUPRI, 1995; Willits, et al. 1990), and quite possibly promotes population deconcentration (Brown, et al. , 1997). However, research has demonstrated that this pro-rural bias is based on nostalgic positive images of rural places, and a misunderstanding of the social and economic realities of rural life (Willits, et al. , 1990). What people value in rural communities is often formed â€Å"at a distance,† through literature, art and music, not through actual experience. As John Logan (1996: 26) has observed, â€Å"A 3 large share of what we value is the mythology and symbolism of rural places, rather than their reality. Accordingly, more reliable research-based information about the social and economic organization of rural areas, their role in national society, polity and economy, and their relative share of a nation’s population and economic activity will provide a stronger bas is for public policy. Bringing beliefs about rural areas into closer connection with empirical reality will improve the fit between rural problems and opportunities, public priorities, and the targeting of public investments. HOW CAN THE NATURE OF POSTINDUSTRIAL RURALITY BE DETERMINED? The Conventional Approach: Rural-urban classification in most national statistical systems typically involves two mutually exclusive categories. In most highly developed societies, (North America, Western Europe, Oceania, and Japan) the rural-urban delineation is based solely on population size and/or density (United Nations, 1999). It is not that government statisticians don’t understand that rurality is a variable not a discrete dichotomy, that the rural-urban distinction is somewhat arbitrary regardless of the population size or density threshold chosen, or that neither the rural nor the urban category is homogeneous. However, given their responsibilities for monitoring basic aspects of social organization and social change, and for providing data tabulations to the public, to businesses, and to other government agencies, the elemental need is to develop a geographic schema that makes intuitive sense, and where between category variability exceeds internal differentiation. It has not been realistic to expect statistical agencies to adopt a complex multidimensional delineation of rurality given the realities and politics of statistical practice in which budget constraints, and competition between stake holder groups determine which items are included on censuses and other large scale public surveys, and which variables are routinely included in tabulations and data products. However, the development of GIS techniques, and new advances in small area data collection and availability suggest that more flexibility and variability in geo-coding may be possible in the future. Hence, while we do not necessarily expect statistical agencies to adopt our multidimensional approach, we believe that it raises important questions about conventional methodologies for assessing the level and pace of urbanization in highly developed nations. 4 OMB’s New Core Based System: A Step In The Right Direction: The public availability of summary tape files from censuses and other nationwide surveys, provides significant opportunities for inquiry by university-based and government scientists into the extent and nature of rurality in postindustrial societies. In effect, analysts can design their own residential categorization schemes to examine various aspects of settlement structure and change. And, innovative research experimenting with alternative categorization systems can eventually contribute to changes in official statistical practice. For example, 25 years of research by social scientists in the USDA’s Economic Research Service (ERS) and in academia is ar guably responsible for persuading the U. S. Office of Management and Budget (OMB) that an undifferentiated nonmetropolitan category is not defensible (Duncan and Reiss, 1956; Butler and Beale, 1994). As early as 1975, ERS was recommending that the nonmetropolitan category be disaggregated according to the degree of urbanization. In a major publication released in that year, Hines, Brown and Zimmer showed that more populous nonmetropolitan counties, especially those adjacent to metropolitan areas, were more similar to metropolitan areas than to their nonmetropolitan counterparts. OMB has now modified its official geo-coding scheme to recognize diversity within nonmetropolitan America. OMB has instituted a â€Å"core based statistical area classification system† that recognizes that both metropolitan and nonmetropolitan territory can be integrated with a population center. The new CBSA classification system establishes a micropolitan category as a means of distinguishing between nonmetropolitan areas that are integrated with centers of 10,000 to 49,999 population, and nonmetropolitan territory that is not integrated with any particular population center of 10,000 or more inhabitants (OMB, 2000). 2 Metropolitan counties contain 79 percent of the U. S. opulation and 21 percent of its land area in the new classification scheme while the 1 In the United States and some other postindustrial countries, two residential categorizations are used: urban vs. rural and metropolitan vs. nonmetropolitan. Some writers use these concepts interchangeably, but even though their respective shares of the nation’s total population have tracked quite closely during recent decades, they are different concepts. What is similar between them, however, is that rural and nonmetropolitan are both residuals that are left over once urban settlement is accounted for. Hence, the rural population includes all residents of places of less than 2,500 and persons who live outside of urbanized areas while the nonmetropolitan population includes all persons who live outside of metropolitan counties (counties containing or integrated with a place of 50,000 persons). 2 Social scientists have also objected to the use of counties as building blocks for the nation’s metropolitan geography, but the new OMB standards have retained counties in the new classification system (Morrill, Cromartie and Hart, 1999). 5 ercentages are exactly reversed for nonmetropolitan territory. The nonmetropolitan population is almost evenly split between micropolitan and noncore based areas, although the former category contains 582 counties while the latter has 1668. The data in tables 1-3 show substantial diversity between micropolitan and noncore based areas, and demonstrate the importance of distinguishing between these two types of counties. To begin with, the average mi cropolitan county has 45,875 persons compared with only 15,634 persons in the average noncore based area. The data in Table 1 also show that micropolitan counties have 43 persons per square mile while only 12 persons live on each square mile of noncore based territory. [Table 1 here] Table 2 compares social and economic characteristics of persons living in various types of U. S. counties. In each instance these data show regular patterns of decline as one moves from the largest metropolitan counties to noncore based counties. For example, almost half of all metropolitan persons have attended college compared about one third of nonmetropolitan residents, but only 31 percent of noncore based adults have been to college compared with 37 percent of persons living in micropolitan counties. Metropolitan workers are more dependent on jobs in service industries while their nonmetropolitan counterparts depend more heavily on farming and manufacturing, although these differences are not strikingly large. Within the nonmetropolitan category, however, dependence on farming is over twice as high in noncore based counties compared with micropolitan areas, and small but consistently smaller percentages of noncore based employees work in manufacturing, retail and services jobs. Similarly, professional, technical managerial and administrative occupations comprise a much larger share of metropolitan than nonmetropolitan jobs, and a larger share in micropolitan than in noncore based counties. Data on earnings per job (displayed in the bottom panel of Table 2) show that noncore based workers earn less than their micropolitan counterparts in all industrial categories, and their earnings are consistently the lowest of any county type in the U. S. [Table 2 here] We have also examined whether micropolitan areas are more â€Å"metropolitan† than noncore based counties with respect to the presence of various services and facilities typically associated with metropolitan status (Beale, 1984). We conducted a mail survey 6 f the heads of county government in a 10 percent random sample of noncore based areas, and in 20 percent of micropolitan and small metropolitan areas. We have only received about 40 percent of the questionnaires from the county executives at this time, so the data in Table 3 are provisional. 3 However, these preliminary results reveal that central counties of small metropolitan areas are clearly differentiated from both nonmetropolitan categories. In all t welve instances the presence of these â€Å"metropolitan functions† is most prevalent in small metropolitan counties, and least available in noncore based areas. Micropolitan areas, however, appear to be more similar to small metropolitan areas than to noncore based counties. Hence, OMB’s new system seems to be a step in the right direction from the undifferentiated nonmetropolitan residential. It does a good job of distinguishing between metropolitan and nonmetropolitan areas, and between micropolitan and noncore based areas outside of the metropolitan category. [Table 3 here] While we applaud the OMB’s new classification system as a step toward recognizing rural diversity, we believe that it is just that, one step. We recommend that social science research further examine the multidimensional nature of rurality in order to enhance understanding of the extent of urban and rural settlement and urbanization in postindustrial societies, and to guide future modifications of official statistical geography. A MULTIDIMENSIONAL APPROACH TO CONCEPTUALIZING RURALITY IN POSTINDUSTRIAL SOCIETIES As mentioned earlier, our multidimensional approach elaborates and extends earlier work by Paul Cloke (1977; 1986). The basic notion is that while urban and rural have intrinsic meaning, both concepts derive much of their analytical power when compared with the other. Low population density, for example, has important meaning in and of itself, but its meaning is further clarified when low rural density is compared with the high ratio of persons to space found in urban regions. 4 Cloke’s objective was to develop a 3 We are now involved in the refusal conversion process and hope to obtain at least a 60 percent response rate. Moreover, attitudes about urban and rural areas are formed on the basis of the attributes people believe characterize such areas, but these attitudes also reflect people’s opinions of how rural and urban areas differ 4 7 quantitative statement of rurality that could be used as a basis for comparative studies among rural areas, and between them and urban areas. He used principal components analysis to identify nine variables associated with rural-urban location. Principal components loading scores were then used as weighting criteria to form an index of rurality. The resulting scores were arrayed in quartiles ranging from extreme rural to extreme non-rural, and each of England’s and Wales’ administrative districts was assigned to one of these four categories. In 1986, Cloke replicated his 1971 index. His second study showed that while most districts were classified in the same rural-urban category in both 1971 and 1981, some districts changed categories over the decade, and the nature of rurality itself was marginally transformed over time. He found that the variables differentiating rural from urban areas in 1981 were somewhat different than those used in the initial analysis. In particular, population decline and net out migration were important rural attributes in 1971, during a period of population concentration, but not in the 1981 analysis after the relative rates of rural-urban population change and net migration had reversed in favor of the periphery. The 1981 revision included 8 variables. Positive variable loadings on five of the eight factors indicated that they corresponded to urban characteristics (high level of housing occupancy, high percentage of workers outcommuting, high percentage of women in childbearing ages, high level of household amenities, and high population density) while negative loadings on the remaining three variables corresponded to rural characteristics (high involvement in extractive industries, disproportionate number of older persons, and distance from an urban area of 50,000 population). It is important to point out at this juncture that neither Cloke nor we are geographic determinists, e. g. , we do not contend that the type of environment people live in has an independent causal effect on their attitudes and behavior. On the other hand, we believe that spatial locality is more than simply a setting in which social and economic relationships occur. Our position is that a person’s place of residence in a nation’s settlement system can shape social and economic outcomes, and can have a profound impact on life chances (Brown and Lee, 1999). While a growing number of social from each other. Accordingly, the public’s overall positive attitude toward rural people and areas is a combination of â€Å"pro-rural† and â€Å"anti-urban† attitudes. 8 scientists agree that space should be incorporated into social theory and research, there is little agreement on the manner in which space enters into social behavior. The debate hinges on the question of whether spatial arrangements are an elemental cause of social behavior, or whether space acts in a more contingent manner. Our position is consistent with the latter view; that space has an important but contingent causative role in social relations. Hence, we see value in distinguishing rural from urban areas because we contend that rural-urban variations in socioeconomic status, for example, can only be understood by taking into account how contingent characteristics of rural and urban places modify the access to opportunities. In other words, we are saying that local social structure contextualizes social and economic behavior. We do not question the existence of fundamental social relationships, but we observe that these relationships are modified by spatial variability in social and economic contexts. Linking back to the status attainment example, education is positively related to income in all locations, but the strength of this relationship varies across local labor markets depending on their industrial and occupational structures. Education matters everywhere, but returns to education are higher in some spatial contexts than in others depending on the availability of well paying jobs and on the nature of the stratification system (Duncan, 1999). Dimensions of Rurality in the United States at the Turn of the Century: Cloke’s approach to defining rurality was largely inductive. His choice of variables was not shaped by a clearly defined theoretical framework for distinguishing rural from urban, although they were suggested by the literature as being important aspects of the sociospatial environment. Neither do we claim that our approach emanates from a wellcrafted theory of rurality, but we do start with a clear premise about four distinct dimensions that comprise rural environments in postindustrial societies. We then choose indicators for each domain that have been shown in the research literature to vary across rural-urban space. The concept of rurality we are proposing involves ecological, economic, institutional, and sociocultural dimensions. In this section of the paper we discuss each of these four dimensions in turn, and propose a set of indicators that could be used to empirically develop a composite measure of rurality. We follow Willits and Bealer (1967) in observing that a composite definition of rurality involves both the attributes of rural areas themselves, and the attributes of persons residing in such areas. Figure 1 shows 9 the four dimensions of rurality, indicators of each dimension, and the contrasting rural vs. rban situation for each indicator. Our approach indicates the attributes that define rurality, and it does so in a comparative framework vis a vis urbanity. [Figure 1 here] The Ecological Dimension: Population size, population density, spatial situation within a settlement system and natural resource endowments are included in this dimension. As indicated earlier, conventional statis tical practice typically emphasizes this approach. Urban vs. rural delineations are usually defined by a size and/or a density threshold, while metropolitan vs. onmetropolitan delineations use size and density criteria to identify central cities and measures of geographic access such as physical distance or commuting to signify the interdependence of peripheral areas. Hope Tisdale’s (1942) influential article provides one of the clearest theoretical statements for the size/density delineation, while central place theory is the primary theoretical basis for considering geographic location vis-a-vis other places in a settlement system (Berry, 1967). The ecological dimension also includes a consideration of the natural environment. As shown in Table 1, 79 percent of land in the United States is found outside of officially recognized metropolitan areas, and 61 percent is located in noncore based areas. While this tells volumes about density, it also indicates that most of America’s natural resources are located in its rural territory. Energy, minerals, land for agricultural production, water, and habitat for wild life are all found disproportionately in the rural sector, and this is an important aspect of the nation’s rurality during the postindustrial era. The Economic Dimension: This dimension concerns the organization of economic activity in local economies. It focuses on what people do for a living, the size and composition of local economies, and the linkages between local economic activities and national and global capital. Until the mid 20th century, rural and agriculture while not synonymous were very closely related, and definitions of rural were heavily influenced by measures of dependence on agriculture and other extractive industries. Rural economies were small and undifferentiated both in terms of establishments and workers, and localities had a relatively high degree of economic autonomy. 10 Many people continue to view rural areas through this archaic lens, even though local economies have been fundamentally restructured during the past 50 years. Direct dependence on agriculture, forestry, mining and fisheries has declined to less than one in ten nonmetropolitan workers although extractive industries continue to dominate economic activity in particular regions of the U. S. (Cook and Mizer, 1994). There is no denying that economic activities in rural and urban America have become much more similar since World War II. Not only has dependence on extractive industries declined throughout the country, but so has dependence on manufacturing, and most economic growth is now accounted for by services. However, the jobs available in rural labor markets continue to be significantly different than urban jobs. Rural manufacturing is more likely to be nondurable than urban manufacturing, and well paying producer services jobs are seldom available in rural economies. Moreover, research shows that full time rural workers earn less than urban workers regardless of their industry of employment, and that rural employment is significantly more likely to be part time and/or seasonal (Gale and McGranahan, 2001). While these rural-urban differences in employment do not adhere to the traditional farm-nonfarm contours, they show that opportunities available in rural labor markets are clearly inferior to those available in urban America, and that rural and urban areas can be differentiated with respect to how people make a living. Rural economies have traditionally been smaller than urban economies in terms of number of workers, the number and size of establishments, and the gross value of products or services sold. Of the three indicators of rural economic activity, this one has changed the least over time even though the decentralization of urban based branch plants has brought some large employers to particular rural areas. Moreover, rural economies have been much more dependent on one or a few types of economic activity than urban economies, and this too remains an important rural-urban difference. The â€Å"protection of distance† enjoyed (or suffered) by rural economies has clearly diminished in recent decades. Technological changes including all weather roads, the interstate highway system, virtually universal telephone service (now including cell phones), and the internet have greatly reduced rural isolation. This is not to deny that some important inequalities in transportation and communication infrastructure persist 11 between rural and urban areas, but for the most part the effect of physical distance has been substantially leveled by technological advances. Institutional changes, especially the increased mobility of capital, have further diminished rural economic independence. The deregulation of banking means that capital now flows easily to and from metro bank centers and the rural periphery. This has both positive and negative implications for particular rural communities, but the clear result is that rural economies are increasingly integrated within national and global structures. With this change comes a resulting decline of local autonomy and increased dependence on extra-local firms and organizations. This makes rural areas at the same time more attractive sites for certain types of external investment, and more likely to lose traditional employers because of financial decisions made elsewhere. There is little room for sentiment in the globalized economy, including sentiment for rural communities as valued â€Å"home places. † When the bottom line demands it, capital flows across national borders to production sites with low costs and few regulations, locating and relocating according to the demands of the market. The Institutional Dimension: Communities are institutionalized solutions to the problems of everyday life. Accordingly, some social scientists view communities as configurations of institutional spheres including education, religion, governance, the economy, etc. (Rubin, 1969). While we do not necessarily subscribe to this functionalist view of community organization, there is no denying that institutions are a critical aspect of local social structure, and that human beings would have little use for communities if they did not serve recurring needs. Both urban and rural areas have formal institutional sectors. Most places have some form of politics and local governance, organized religion, education, and voluntary and service organizations. Moreover, as discussed in the preceding section, sustenance and economic activity are important aspects of locality. Rural and urban areas are not so much differentiated by the presence or absence of particular types of institutions as by their diversity and capacity. For example, schools, newspapers and churches, are widespread, but most rural communities offer a narrower range of choices as to where one’s children may be educated, where to worship, and/or the media from which one obtains local news. School consolidation in rural America has resulted in fewer and larger schools. Students are often bussed long distances to school. 12 Similarly, while churches are present in most rural communities, the range of denominations and congregations is narrow. Clubs, service organizations, and voluntary associations are also an important part of rural community life, but the choice of organizations to join is constrained in comparison to the organizational choices available in urban environments. Rural institutions also tend to have more limited capacity than their urban counterparts. Rural governments, for example, are often constrained by part time leadership, insufficient fiscal resources, ineffective organizational structures, limited access to technical information and expertise, and limited ability to assess changing community needs (Kraybill and Lobao, 2001; Cigler, 1993). The Sociocultural Dimension: Moral traditionalism is one of the most consistent themes subsumed under the term â€Å"rural culture† (Willits and Bealer, 1967). Rural persons are often considered to be more conservative than their urban counterparts, and data from national surveys indicate this to be true in the United States. Calvin Beale (1995) has shown that 49 percent of rural respondents to a 1993 National Opinion Research Center (NORC) national survey regard themselves as religious fundamentalists compared with 33 percent of urban respondents. Similarly, a much lower percentage of rural respondents believe that abortion should be available for any reason (26 percent vs. 44 percent), and a much higher percentage of rural persons believe that homosexuality is immoral (84 percent vs. 2 percent). Beale also observed that rural voters have been more likely to support conservative candidates in recent elections even though rural persons are slightly more likely than urban persons to describe themselves as democrats. A related idea is that rural conservatism is often associated with the homogeneity of the rural population. Wirth (1938) and ot hers argued that increased population diversity was one of the dominant effects of urbanization, and one of the reasons why informal social control was likely to break down in cities. Ironically, Fisher (1975) and other critics of Wirth, argued that ethnic diversity rather than contributing to a weakening of the social order was a main reason why the strength of social relations did not diminish in cities, and why community was not â€Å"eclipsed† in urban environments. While the association between ethnic and other aspects of population diversity and social and political attitudes is still an open question, research clearly indicates that rural populations in the U. S. , while 13 increasingly diverse, remain significantly more homogeneous than urban populations (Fuguitt, et al. 1989). In addition, the rural population’s racial and ethnic diversity is not spread evenly across the landscape, but tends to concentrate in particular regions and locales (Cromartie, 1999). Hence, even though about one out of ten rural Americans is African American, few rural communities are 10 percent Black. Rather, Blacks tend either to comprise the majority or large mino rity of a rural population or an insignificant percentage. The same tends to be true with respect to other racial and/or ethnic populations. Much has been written to suggest that primary social interaction is more prevalent and more intense in rural areas, and that rural areas have a higher level of informal social control than is true in urban areas. However, these contentions, if ever true, are not supported by contemporary empirical evidence. Copious research has shown that urban persons are involved in regular and intense interaction with family, friends and neighbors, and that community has not been eclipsed in urban America (Hummon, 1990; Fischer, 1975). Moreover, research by Sampson (1999), and others has shown that social networks are quite effective in regulating social behavior in urban locales. Accordingly, primary social interaction and effective social control do not differentiate rural and urban areas in contemporary American society, and are not components of the sociocultural dimension of rurality. CONCLUSIONS How urbanized are postindustrial societies? How rapidly is the remaining rural population being incorporated within the urban category? How do rural people and rural areas contribute to and/or detract from the social and economic well being of highly developed nations? We contend that answering these questions accurately is contingent on the availability of theoretically informed definitions of rural and urban areas. Virtually every developed nation uses population size and density as the basis for its differentiation of urban and rural areas. Areas obtain urban status by reaching some threshold of population size and/or density, and commuting or some similar measure of routine social and/or economic interaction is used to determine whether peripheral areas are integrated with, and hence part of large/dense urban agglomerations. Rural areas are simply the residual—areas that fail to satisfy the urban threshold or lack routine interaction with core 14 areas. We join with many previous scholars in arguing that this approach is blind to the complex multidimensional nature of postindustrial rurality. We believe that the residual approach is inadequate for differentiating rural from urban populations, and for examining social, economic, political, ecological and other forms of diversity within the rural category itself. We have recommended a multidimensional framework for considering the nature of rurality in postindustrial society. Our approach includes conventional demographic measures, and adds information on the natural environment, economic structures and activities, the diversity and capacity of institutions, and a sociocultural domain. Our case is the United States but we believe that the situation we describe in the U. S. is similar to that in most other postindustrial societies. Our paper rejects the notion that rurality is simply a residual that is leftover once urban areas have been identified. The rural as residual approach clearly identifies the extremes or urbanity and rurality (Paris, France vs. Paris, Texas, for example), but it offers no guidance for examining settlements that fall in the intermediate zone between these extremes. We believe that the multidimensional approach to conceptualizing rurality is helpful not only for distinguishing urban from rural but also for understanding the variability of social and economic organization that occurs within both categories. As we have shown, the OMB’s new core-based statistical areas systems is a step toward recognizing important aspects of rural diversity and of focusing attention on the zone between what is clearly urban and clearly rural. We acknowledge that there is a venerable tradition in social science of examining the correlates of city size (Duncan, 1951; Duncan and Reiss, 1956), and that it is possible that rural-urban variability in ecological, economic, institutional and sociocultural attributes may simply be a reflection of inter area differences in population size. If this is the case then the conventional practice of using population size to define urbanity may be sufficient for delineating urban from rural. In contrast, if the other dimensions of social and economic activity are only weakly associated with population size then conventional statistical practice may be producing misleading information regarding urbanization and the conditions of life in rural and urban communities. This important question merits continued examination in future research. 15 Changes in a nation’s urban-rural balance have significance that extends beyond purely academic curiosity. Understanding how variability in spatial context affects opportunity structures and the quality of life contributes to producing flexible public programs that are sensitive to local needs. Misinformation about the social, economic and institutional organization of rural and/or urban areas, and about the size and composition of a nation’s population living and working in rural and urban places will result in misinformed policies. For example, if policy makers believe that most rural persons are farmers, agricultural policies will be seen as a reasonable response to rural poverty and income insecurity. But, of course, agricultural policies will not have much of an effect on rural poverty because most rural persons in postindustrial societies do not depend on farming for their livelihoods (Gibbs, 2001). Or, if research indicates that the size of a nation’s rural population has held constant over time, as is the case in the United States where about 55-60 million persons has been classified as rural since 1950, then significant public investments for rural development will be legitimized (at least from an equity perspective). But, if the measurement of rurality is too permissive, and the population that is genuinely rural has actually declined, then public resources may be targeted to the wrong populations. We realize that the multidimensional perspective we are promoting could not be easily or cheaply built into a national statistical system. But, regardless of its practicality our framework raises important questions about the sufficiency of the size/density conventions used throughout the developed world, and consequently about the state of knowledge on urbanization in postindustrial societies. Moreover, our contention that rurality should not be treated as an undifferentiated residual complements the social representational approach in which rurality is defined by how people imagine community life in everyday discourse. Both approaches focus attention on the complexity of contemporary rural life and its continuing distinctiveness in comparison with urban areas. 16 REFERENCES Beale, C. 1995. â€Å"Non Economic Value of Rural America. † Paper presented at the USDA experts’ conference on the value of rural America. † Washington, DC: USDA-ERS. ______. 1984. â€Å"Poughkeepsie’s Complaint or Defining Metropolitan Areas. † American Demographics 6(1): 28-31; 46-48. Berry, B. 1967. Geography of Market Centers and Retail Distribution. Englewood Cliffs, NJ: Prentice Hall. Brown, D and M. Lee. 1999. Persisting Inequality Between Metropolitan and Nonmetropolitan America: Implications for Theory and Policy. † Pp. 151-167 in P. Moen, D. Demster-McClain and H. Walker (eds. ) Diversity, Inequality, and Community in American Society. Ithaca: Cornell University Press. ______. G. Fuguitt, T. Heaton, and S. Waseem. 1997. â€Å"Continuities in Size of Place Preferences in the Uni ted States, 1972-1992. † Rural Sociology 62(4) : 408-428. Butler, M. and C. Beale. 1994. â€Å"Rural-Urban Continuum Codes for Metropolitan and Nonmetropolitan Counties, 1993. † Staff Report No. 9425. Washington, DC: USDA-ERS. Cigler, B. 993. â€Å"Meeting the Growing Challenges of Rural Local Government. † Rural Development Perspectives 9(1): 35-39. Cloke, P. and G. Edwards. 1986. â€Å"Rurality in England and Wales, 1981: A Replication of the 1971 Index. † Regional Studies 20: 289-306. _____. 1977. â€Å"An Index of Rurality for England and Wales. † Regional Studies 11: 31-46. Cook, P. and K. Mizer. 1994. â€Å"The Revised ERS County Typology. † Rural Development Research Report No. 84. Washington, DC: USDA-ERS. Cromartie, J. 1999. â€Å"Rural Minorities Are Geographically Clustered. † Rural Conditions and Trends 9(2): 14-19. Duncan, C. 1999. Worlds Apart: Why Poverty Persists in Rural America. † New Haven: Yale University Press. Duncan, O and A. Reiss. 1956. Social Characteristics of Urban and Rural Communities. New York: John Wiley and Sons. Duncan, O. 1951. â€Å"Optimum Size of Cities. † Pp. 632-645 in P. Hatt and A. Reiss (eds. ) Reader in Urban Sociology. New York: Free Press. 17 Durkehim, E. 1951. Suicide. New York: Free Press. Fischer, C. 1975. â€Å"Toward a Subcultural Theory of Urbanism. † American Journal of Sociology 80: 1319-1342. Fuguitt, G. , D. Brown, and C. Beale. 1989. Rural and Small Town America. New York: Russell Sage Foundation. Gale, F. nd D. McGranahan. 2001. â€Å"Nonmetro Areas Fall Behind in the New Economy. † Rural America 16(1): 44-51. Gibbs, R. 2001. â€Å"Nonmetro Labor Markets in an Era of Welfare Reform. † Rural America 16(3): 11-21. Giddens, A. The Constitution of Society. Cambridge: Polity Press. Halfacree, K. 1993. â€Å"Locality and Social Representa tion: Space, Discourse, and Alternative Definitions of the Rural. † Journal of Rural Studies 9(1): 23-37. Hauser, P. 1965. â€Å"Urbanization: An Overview. † Pp. 1-47 in P. Hauser and L. Schnore (eds. ) The Study of Urbanization. New York: John Wiley and Sons. Hines, F, D. Brown, and J. Zimmer. 1975. Social and Economic Characteristics of the Population in Metropolitan and Nonmetropolitan Counties, 1970. † Agricultural Economic Report No. 272. Washington, D. C. : USDA-ERS. Hummon, D. 1990. Common Places: Community Ideology and Identity in American Culture. Albany: SUNY Press. Kellogg Foundation. 2002. Perceptions of Rural America. Battle Creek, MI. : Kellogg Foundation. Kraybill, D. and L. Lobao. 2001. County Government Survey: Changes and Challenges in the New Millennium. Washington, DC: National Association of Counties. Lewis, M. 1991. â€Å"Elusive Societies: A Regional-Cartographical Approach to the Study of Human Relatedness. Annals of the Association of Ame rican Geographers 18(4): 605-626. Logan, J. 1996. ‘Rural America As A Symbol of American Values. † Rural Development Perspectives 12(1): 24-28. Marx, K. 1976. Capital, Vol. I. London: Penguin NLR. Morrill, R, J. Cromartie, and G. Hart. 1999. â€Å"Metropolitan, Urban, and Rural Commuting Areas: Toward a Better Depiction of the united States Settlement System. † Urban Geography 20(8): 727-748. 18 Moscovici, S. 1981. â€Å"On Social Representation. † Pp. 181-209 in J. Forgas (ed. ), Social Cognition: Perspectives on Everyday Understanding. London: Academic Press. RUPRI. 1995. 1995 National RUPRI Poll: Differential Attitudes of Rural and Urban America. † Columbia, Missouri: Rural Policy Research Institute. Rubin, J. 1969. â€Å"Function and Structure of Community: A Conceptual and Theoretical Analysis. † International Review of Community Development 21-22: 111-119. Sampson, R. , J. Morenoff, and F. Earls. 1999. â€Å"Beyond Social Capital: Spatia l Dynamics of Collective Efficacy for Children. † American Journal of Sociology 92(1): 27-63. Tisdale, H. 1942. â€Å"The Process of Urbanization. † Social Forces 20: 311-316. United Nations. 1999. World Urbanization Prospects: 1999 Revision. New York: United Nations. U. S. Office of Management and Budget. 2000. â€Å"Standards for Defining Metropolitan and Micropolitan Statistical Areas. † Federal Register 65(249): 82228-82238. (http://www. whitehouse. gov/omb/fedreg/metroareas122700. pdf. ) Weber, M. 1968. Economy and Society. New York: Bedminister. Willits, F. , R. Bealer, and V. Timbers. 1990. â€Å"Popular Images of Rurality: Data From a Pennsylvania Survey. † Rural Sociology 55(4): 559-578. ______. 1967. â€Å"An Evaluation of a Composite Index of Rurality. † Rural Sociology 32(2): 165-177. Wirth, L. 1938. â€Å"Urbanization As a Way of Life. American Journal of Sociology 44(1): 129. 19 Figure 1: A Multidimensional Framework of Rurality in Postindustrial Society Indicators Rural Areas or Populations Urban Areas or Populations Are More Likely to Be: Are More Likely to Be: Dimensions of Rurality Ecological Dimension Population Size Population Density Situation in Settlement System Natural Environ ment Economic Dimension Dependence on Industrial Activities Size of Local Economy Diversity of Economic Activity Autonomy of Local Economy Institutional Dimension Local Choice Public Sector Capacity Sociocultural Dimension Beliefs/Values Population Diversity Small Low/Scattered Peripheral Rich in Natural Resources Large High/Concentrated Central Lacking Natural Resources Extractive Nondurable Manufacturing Consumer Services Small Workforce Small Establishments Undiversified Low/Dependent Producer Services Professional Services Durable Manufacturing Large Workforce Large Establishments Diversified High Narrow/Constrained Limited/Modest Wide High Conservative Homogeneous Progressive Heterogeneous 20 Table 1: Population, Land Area, Density and Percent Rural by CBSA Category, 19901 CBSA Category U. S. Metro Large Small Nonmetro Micro Non-CBSA 1 No. Counties 3,141 891 606 285 2,250 582 1,668 Population 1,000s Percent 248,709 195,930 171,606 24,323 52,780 26,699 26,081 100 79 69 10 21 11 10 Land Area (square miles) 1,000s Percent 3,536 737 488 249 2,799 625 2,174 100 21 14 7 79 18 61 Population Per Sq. Mile 70 266 351 98 19 43 12 See OMB (2000) for discussion of procedures used to delineate CBSA county types. Source: 1990 U. S. Census of Population 21 Table 2: Comparative Profile of Metro, Micro and Noncore Based Counties, U. S. , 19901 Metropolitan Large Small Nonmetropolitan Micro Noncore Characteristic Educational Attainment Pct. Less Than High School Pct. High School Pct. College Total Total 23 29 48 23 28 49 25 32 43 31 35 34 29 34 37 34 36 31 Industry of Employment (selected) Pct. Farm Pct. Manufacture Pct. Retail Pct. Services 1 13 16 29 1 13 16 30 3 15 18 25 8 18 16 21 5 18 17 22 11 17 15 19 Occupation of Employment (selected) Pct. Manager, Professional Pct. Tech. , Sales, Admin. Pct. Labor2 Earnings Per Job3 All Jobs (000) Manufacture (000) Retail (000) Services (000) 1 2 28 33 24 29 34 24 24 30 28 20 26 34 21 27 33 18 24 36 27 36 15 24 27 37 15 25 0 27 12 16 20 25 12 15 20 27 12 16 18 23 11 14 See OMB (2000) for rules used to identify county types. Skilled and unskilled 3 Nonfarm jobs Source: 1990 U. S. Census of Population 22 Table 3: Presence of Services and Facilities by County Type, 20001 Percent Provided in County Micro 29 71 62 58 91 89 41 64 38 100 45 Service or Facility Scheduled Passenger Air Service Scheduled Inter County Bus Service Local Bus Servic e Museum2 Daily Newspaper National or Regional Hotel Franchise Four Year College Library with Multiple Branches Commercial Television Station3 General Hospital4 N 1 Small Metro 50 91 95 77 95 100 82 64 68 100 22 Noncore Based 11 31 29 23 18 44 11 34 9 74 71 Ten percent sample of noncore based counties; 20% samples of small metro and micro counties. Current response rate = small metro: 41%; micro: 75%; noncore: 42%. Art, science or natural history with focus beyond local county. With local news and advertising. With at least two of four of the following services: emergency room, physical therapy, cardiac care or MRI. 2 3 4 23