Rationalizing Rural Area Classifications for the Economic Research Service: A Workshop Summary (2016)

4

Changes in Society and Economy and Their Impact on Rural Area Classifications

This chapter summarizes the fourth session of the workshop, which was a panel discussion on the the changes in society and economy during the past few decades that have transformed the nation’s settlement system and contributed to the need for reconsidering rural-urban classifications systems. Panelists were Bruce Weber (Oregon State University), David Plane (University of Arizona), David Brown (Cornell University), Linda Lobao (Ohio State University), and Jeff Hardcastle (University of Nevada, Reno, and Nevada state demographer). After the panelists’ statements, they responded to each other. The panel discussion was followed by an open floor discussion. James Fitzsimmons (U.S. Census Bureau) moderated the session discussions.

STATEMENT BY BRUCE WEBER

Weber stated that rural classification systems are created to help characterize how context varies across space and how spatial development patterns affect social and economic outcomes. He noted that a rural classification system should be able to capture both urban hierarchy and the spatial aspects of these patterns. The system should show how these patterns relate to the social and economic consequences that policy makers care about: jobs and income, poverty, regional inequality, and economic mobility.

Forces/Changes Affecting Cost of Distance and Spatial Development Patterns

Weber said that four forces affect both cost of distance and spatial development patterns: (1) Changes in information and communication technology that facilitate speedy and inexpensive transfer of ideas across space; (2) innovations in production technology and productivity-enhancing investments that are embedded in the innovations and concentrate production spatially; (3) transportation investments that speed transfer of people and products and movement of innovation across space; and (4) agglomeration economics that favor concentration of people and production across urban space.

These four forces have resulted in globalization and changes in urbanization across the rural and urban continuum, he noted. To organize his perspectives a little differently, information and communication technology (ICT), innovations in production technology, and transportation investments have reduced the cost of distance and fostered global trade and development, which results from the interaction between technology and transportation investments. These same forces, he said, have led to urbanization and rural depopulation. ICT and transportation investments have reduced costs of distance and fostered urbanization but have also contributed to urban deconcentration and rural depopulation. According to Weber, the country has moved from technology to spatial development patterns.

As Weber explained, spatial development patterns have consequences and may generate socioeconomic outcomes that vary across places:

• Urban size, the place in the urban hierarchy, and distance from urban centers lead to differential job and income growth and diversification (Partridge et al., 2007b).
• Spatial development patterns also affect variations in poverty, income inequality, and intergenerational economic mobility.
• Census data, as well as research by Chetty et al. (2014), show cross-regional disparities in income and wealth and that upward mobility varies across space.
• Research by the Economic Research Service (ERS) has shown that food insecurity varies across space.

A useful rural-urban classification system would help illuminate spatial patterns and how socioeconomic outcomes vary with them, Weber said.

Implications for a Rural Classification System

Weber said that policy makers and citizens want to know whether spatial variation in key socioeconomic outcomes across urban-rural space is consistent with social norms regarding fair distribution of resources, equality of access, and opportunity. Providing policy-relevant information about rural-urban differences in resources, access, and socioeconomic outcomes requires a rural-urban classification that conforms to common understandings of rural and urban characteristics; is rich enough to capture important spatial patterns in urban and rural development; and captures rural-urban differences in key social and economic outcomes.

He stated that these needs basically lay out the key features of a desirable rural classification system that can distinguish urban and rural areas along such key dimensions as size of population, density, and remoteness from urban centers. Further, he said, it is important for a classification system to capture important spatial development patterns such as urbanization, urban deconcentration, and rural depopulation. Finally, it is also important to accurately characterize rural-urban and spatial differentials in the key social and economic outcomes of interest.

He observed that the ERS Rural-Urban Continuum Codes (RUCC) have some of these features. The RUCC allow one to observe interesting spatial differentials on intergenerational upward mobility, for example. On average, absolute upward mobility appears to be higher in more rural areas. It is highest in small-town nonadjacent nonmetro counties and completely rural nonadjacent nonmetro counties, Weber noted. Within metro areas, absolute mobility is lowest in medium-sized metro areas.

In closing, Weber noted that Isserman (2005) observed that (1) core-based statistical areas (CBSA) capture integration of rural and urban in an urban-centered system, and (2) the Census rural-urban classification system captures separation of urban and rural. According to Weber, this led Isserman to conclude that CBSA and metro/nonmetro classifications were fundamentally misguided as a way of capturing “rural.” He said Lichter and Brown (2011), Irwin et al. (2009), and others have noted that the boundaries between urban and rural space have blurred as the functions of urban and rural space become increasingly overlapping.

He reminded the audience about Halfacree’s observation (see Chapter 3) that perhaps the idea of locating rural in space does not make sense. He said these cautions are important to keep in mind as classification systems are reevaluated.

STATEMENT BY DAVID PLANE

Plane focused on the geographer’s concepts of place and situation. He said he would deconstruct the topic of rural areas (or rural classification systems) from these two different locational concepts.

Place-Based Perspectives

Plane noted that one way to define a geography uses the perspective that places have meaning; they are derived from their setting or milieu, or the totality of their physical and cultural attributes. A traditional place-based perspective on rural includes a spectrum, and he said he would focus on three aspects. First is simply that a rural settlement has low density, which seems to play a major role in many definitions. A second aspect is that rural economies traditionally have featured resource-based, primary-sector livelihood systems. The third is that rural people, as distinct from urban people, are “local people.” In some parts of the world, this characteristic may be still true, but in the United States, it may be disappearing, Plane said.

Plane noted that it is time to reconceptualize how density is measured. Density has always been a property of areal units, but other options are now possible. It is possible to start from a location, go out to some threshold such as 50 or 100 miles, and calculate the density at that point on the Earth’s surface, as demonstrated by the ERS approach to Frontier and Remote Area (FAR) Codes. He observed that a time threshold, such as driving time, might be even better than a mileage threshold. A threshold is needed to calculate a density, but the result could be aggregated up on a per capita basis, he suggested.

In terms of livelihoods, rural areas are not all about agriculture anymore. In the United States, many manufacturing companies seek out rural areas. Plane referred to a map from the Chicago Federal Reserve Bank that showed the locations of auto assembly plants in rural areas in the South as an example. Plane observed people are not fixed geographic points, though the Census Bureau still uses the concept of the master address file. People move around, and he suggested rural people’s activity spaces may be larger than those of urbanites. He commented that a metro area is defined to be a place where a lot of people’s activity patterns overlap. Plane also noted many people have multiple residences and operate their lives out of multiple locations, something that will be more common in the future.

From a place-based perspective, both a strength of, and a problem with, federal rural definitions is the requirement for consistency, Plane said. The United States has different historical settlement patterns, and

these settlement patterns, the space economy, and political geographies are highly variable across regions. Rural definitions, in contrast to urban, are largely not place-based or defined as separate rural areas. They are either rurality or leftover nonmetro territory, and he asked what would constitute a meaningful definition of rural. People in rural areas do not necessarily identify with a single place in the same way that people in a metropolitan area do, he said.

Plane said most rural classification systems do not use the regionalization methods of formal, uniform, homogeneous regions versus functional, nodal, spatial interaction-based regions to the extent that they could in coming up with definitions, with some exceptions over the decades:

• Formal regions: The State Economic Areas were a collectively exhaustive/mutually exclusive county-based classification based on economic structure.
• Functional regions: The Bureau of Economic Analysis Economic Areas, now used only by OECD, are nonmetro portions of urbanized area hinterlands, plus nonmetro-centered nodal regions that predated the micropolitan area concept.
• Micropolitan Areas: Micropolitan areas are an attempt to define less-than-metro areas. However, Plane stated that they do not result in a uniform class of entities. In the East, they are leftover spaces between the interstices of metro areas. In the West, a micropolitan area could be an important central place with its own fairly big hinterland.

Situation-Based Perspectives

The metro concept is about an overlay of individual’s daily activity spaces, Plane said. He referred to his work with Daoqin Tong (Tong and Plane, 2014), which looks at people’s activity patterns regardless of where they are going, rather than as a relationship between a core and a periphery. The key conceptual theory to rural people’s activity patterns, Plane said, is central place theory (Christaller, 1933), which concerns the situational aspects of rural households with respect to a hierarchy of other places. According to Huff (1976), rural consumers obtain their goods and services differently depending on where their homes are located within the hierarchically nested meshes of market areas. Moreover, Plane said, real-world activity patterns of rural people are even more complex than in the optimal world of Christaller’s 1933 central place theory. Where a person is located matters in terms of his or her likely activity patterns. If anything, he said, central place theory understates the number of places

that rural people go as part of their daily lives, either in person or by telephone.

In closing, Plane stated that a revival of delineations based on place-based rural areas would be useful, as well as more research on situation-based rural regions and rural residents’ activity spaces. He noted many federal definitions do not embed critical concepts of hierarchies. People live within nested hinterlands of multiple places, he said, whereas definitions are trying to get a mutually exclusive, collectively exhaustive set of regions that are only one level of the hierarchy. Plane stressed people live at multiple scales in terms of their relations with places up the hierarchy. He suggested a revival in interest in central place theory, with more research to identify current critical functions since the thresholds, market areas, ranges, and other characteristics are likely to be very different than when Christaller first did his work.

STATEMENT BY DAVID BROWN

Brown stated the issues discussed throughout the workshop are about people and not just places. He focused on increased diversity of the U.S. rural population. Brown referred to Louis Wirth’s “Urbanism as a Way of Life” (Wirth, 1938). He said Wirth provided a sociological definition of the city, with three defining attributes being large, dense, and heterogeneous. By implication, rural areas were smaller, less dense, and less diverse. Brown noted size and density have been discussed, and he said he would discuss diversity as measured by population composition by race, ethnicity, and age.

Race and Ethnicity

Brown noted that historically, the overall nonmetropolitan U.S. population has been less racially and ethnically diverse than the metro, although not across the board. Two examples are in the nonmetro South (African Americans) and parts of the nonmetro Southwest (Hispanics). There has been a lot of redistribution of race and ethnic groups in recent years, including in rural or nonmetropolitan areas. Overall, the Native American population has grown faster than whites or Blacks in nonmetro areas since 1980, especially in particular regions. Hispanic populations increased dramatically in the Midwest, Southeast, and Pacific Northwest through immigration and higher fertility among in-migrants. Since 2000, racial and ethnic minorities have become a larger share of the population in over 80 percent of nonmetropolitan counties.

A redistribution of Hispanics from the Southwest into other parts of the nonmetropolitan United States does not mean increased integration

between Hispanics or other racial and ethnic minorities and the majority population, Brown said. Research by Duncan (2000) and Pfeffer and Parra (2009), for example, indicates that ethnic minorities and Anglos often live in separate social and institutional worlds. Research by Parisi et al. (2011) demonstrates enduring residential segregation in nonmetro America.

Age Composition

Changing age composition is another aspect of increasing nonmetro diversity, Brown said. Clearly the nonmetropolitan population is aging along with the rest of the nation, but according to the 2005-2009 American Community Survey (ACS), the nonmetro population is aging more rapidly than the metro. About 12 percent of the metro population is 65 years and above, compared to 15.8 percent in nonmetro areas. However, for non-core areas, the percentage is even higher—14.8 percent in micropolitan areas and 16.5 percent in non-core-based areas.

There are two different contexts for extreme nonmetropolitan population aging, Brown said: places that are destinations for older migrants and natural population decrease. Natural population decrease has become much more prevalent in the United States in both metro and nonmetro areas. Brown noted that retirement destinations can become natural-decrease counties if older in-migration is not accompanied by the in-migration of younger people.

Areas of natural decrease represent a diverse category, he said. They occur in a variety of geographic locales, with a heavy concentration in the Plains but spreading widely throughout the nation, in both metropolitan and nonmetropolitan areas. According to Johnson (2013a, 2013b), natural decrease is much more prevalent in nonmetropolitan areas than metropolitan areas.

Natural decrease occurs in various places at different historical times. Two different pathways lead to natural decrease: (1) chronic out-migration of young adults and their children produces a distorted age structure, resulting in more deaths than births even if the fertility rate of those people is at or above average; and (2) net in-migration of older people without concurrent or subsequent in-migration of persons of childbearing age. In much of Europe, low fertility rates are a main cause of natural population decreases, but age composition is the issue in the United States. The model situation for U.S. nonmetropolitan areas is out-migration at younger ages and in-migration at older ages. That is certainly the case with natural decrease counties, and more so the longer and more persistent the natural decrease is over that period of time. But the natural decrease category, like any other category, is very heterogeneous, noted Brown, and examining the heterogeneity within these categories is important.

Natural population decrease is not necessarily a problem unless it persists for a significant period of time, Brown said. In such instances, it can undermine an area’s potential for future population growth, and it may reduce a community’s ability to retain essential institutions and economic activities. Brown said it is important to recognize that natural decrease is the result of chronic out-migration of young adults and long-term population decline. Hence, from a policy perspective, chronic demographic decline would be a more appropriate target variable than natural decrease, which is itself an outcome of long-term decline and distorted age structure.

Conclusion

Brown highlighted four conclusions and questions:

1. Nonmetropolitan regions are increasingly diverse. They were never as homogeneous as originally imagined.
2. The increasing diversity of the nonmetropolitan population reflects economic transformations, increased longevity, and changes in opportunity structures. Dynamic global-local relationships expose nonmetro areas to new challenges and offer new opportunities in areas such as capital, labor mobility, and information technology. The increasing diversity also reflects increased longevity, and changes in life course processes.
3. Differing social and economic aspects of life in rural America are not mutually exclusive. An important question is how to recognize this lack of mutual exclusivity in statistical systems that categorize areas.
4. Statistical geographies that appear to be homogeneous, such as the natural decrease areas of the United States, are often highly diverse. How can within-variability be minimized and between-variability be maximized in substantively meaningful ways?

STATEMENT BY LINDA LOBAO

Lobao summarized some major changes that have occurred over the past several decades that would contribute to the need for reconsidering rural classification systems. She noted the following changes with different subnational geographic expressions/rural implications across the nation:

• Economic structure: Continuing rounds of restructuring are reflected in the quality and quantity of employment. Some exam-

• ples are long-term declines in manufacturing and coal mining; growth in some rural areas in the new energy economy from the oil and gas industry; growth in care-work, such as personal services, health, education, and social services; and growth in the financial sector. These changes have varying spatial expressions. For example, services such as care-work are less spatially varied than the financial sector, coal mining, and oil/gas mining.

• Changes in the state: Continuing decentralization has occurred where responsibilities are passed downward to subnational governments. She said this has increased the importance of state and local governments in influencing growth and redistribution across places and populations, but also creates a patchwork of inequality due to the varying capacities of these governments and hence in the nation’s social safety net. Local governments are particularly critical to the well-being of places and populations. If size of government is measured by employment, local governments are about 63 percent of all government employment.
• Institutional arrangements: Institutional arrangements have changed dramatically over the past several decades. Examples are shifts in the balance of power between business and labor, declines in unionization, attempts to dismantle public sector unions, and the weakening ability of the state to protect its citizens with the social safety net.
• Continuing growth of income inequality and other economic polarization: As noted by Andrew Cherlin (2014), the ability of the traditional working class to generate income sufficient for family livelihoods and family stability has declined.

Lobao said that these changes are conceptualized by some social scientists as a package that defines the current period of national development. Social scientists characterize this as the “neoliberal period,” a period in which state and market relationships have changed to some degree fundamentally.

Major Changes: Local Governments and Geographic Diversity

Lobao pointed to research on the growing importance of subnational governments in many nations besides the United States. The United States historically has had a more decentralized governance system when compared to Europe, but she said some analysts see increasing trends toward “fend for yourself” federalism. As noted above, local governments are already about 63 percent of all civilian government employment. Counties were the fastest-growing general purpose governments, at least up to the

recession. If the size of government is measured by employment, county governments are larger than the federal civilian government.

There are disparities across local governments in fiscal and administrative capacity and public service provision. This results in stratification of the U.S. population into places that offer high-capacity, expert government with better protection from poverty and downturn versus places that do not. For example, she said, even though county governments tend to be the major government for unincorporated areas, rural county governments have less capacity and provide fewer services.

She said variations across local governments appear to be increasing due to several factors:

• Declines continue in state and federal funding and the dumping of fiscal problems downward to localities.
• Localities become increasingly dependent upon own-source funds and facing specific fiscal problems, which makes it more difficult for local governments to provide public services.
• The fallout from the Great Recession likely has increased variations across governments. However, limited data, small sample studies, and focus on high-profile/often-urban governments may contribute to a delay in knowledge about this. No counter-cyclical response was seen in local government employment over the recession period, at least in much of the data. According to Dadyan and Boyd (2013), almost half-million local government employees were lost within two years after the recession. Analysts are debating if this is the “new” normal for local governments as opposed to coping with decline as “normal.” She said some researchers find that local resource scarcity is long term and ongoing and it has been for 30 years or more (Perlman and Benton, 2015).
• Partisan polarization has added to variations across local governments, she said. Local infrastructure requires federal intervention, she stated, but partisan polarization has disrupted the U.S. system of fiscal federalism (Kettl, 2015). If the federal level cannot deal with serious governmental issues, what will happen at the local level, she asked. Local governments are confronting these issues in different ways. Immense geographic diversity is seen in the capacities, the resources, and the specific issues that local governments face.
• The outcomes of the future system of intergovernmental relations are debated by scholars and policy makers. Some see the prom-

Local Governments and Rural-Urban Variability: Counties Across the United States

Lobao reported on results from primary data from a survey conducted in 2007/2008 in collaboration with the National Association of Counties (NACo). Even though the data were collected at the cusp of the recession, they are still the most generalizable data available on nonmetro and metro counties. The survey focused on the 46 contiguous states with county governments (Connecticut and Rhode Island lack county governments), and data were collected from county commissioners and other county officials. The survey is unique for its high response rate of 60 percent. Contrasts were presented between metro and nonmetro counties. Within nonmetro counties, comparisons also were made between micropolitan and noncore counties and between metro-adjacent and nonmetro-adjacent counties. She summarized the findings:

• Government capacity: For a list of 28 county services, there are significant metro/nonmetro variations in all but eight services.
• Economic development policies: Metro counties use more overall and a greater mix of policies that span business attraction, local business support, and workforce development. Marketing a county as a site for a prison is reported more frequently by micropolitan counties compared to other counties.
• Barriers to economic development: Metro counties report fewer barriers, except for school quality. Micropolitan (relative to other noncore and metro-adjacent counties) were less likely to report limited employers as being a problem.
• Resource shortages: Nonmetro governments have always been more limited. They report lower rises in service demands and less pressures to reduce taxes, but greater pressures from loss of state revenues. More dynamic metro governments are more likely to cut back, privatize, and use hiring freezes. Metro governments provide greater services overall, and hence when resource shortages occur, they have greater capacity to make adjustments.

Concluding Comments and Implications for Revised Rural Classification

Lobao concluded that local governments are increasingly important for analyzing and understanding the path of development across the subnational United States. Extensive diversity exists among them. This warrants more of a spatially sensitive subnational approach beyond binary rural and urban categories. However present metro-nonmetro classifications do yield some systematic patterns. Lobao indicated that their data suggest micropolitan status seems to capture some unique attributes of rural counties.

Counties, perhaps even more important than in the past in analyzing key subnational changes, combine policy-making units with population aggregates and capture some conceptualizations of “place” in terms of administrative unit, local social system, and community bond, culture, or place identification. For counties, total government and population size, distance from metro areas, and presence of a larger urban area appear to have some relationship with policy choices. She noted a caveat: These relationships may vary when different control variables are added in multivariate models. As units of analysis, administrative units can be expected to have increasing importance for national well-being. Municipalities, too, may offer an important lens.

Lobao said continuity and change exist in structural forces influencing rural America and in their geographical expression. Updating classifications with an eye to recent changes playing out across the entire subnational United States would be useful to researchers.

STATEMENT BY JEFF HARDCASTLE

Hardcastle focused his remarks on his experience in Nevada. There, he said, rural still matters, but the question is how to talk about it.

Aging of the Population

Aging is an issue in Nevada, Hardcastle said. Based on median age from the 1990, 2000, and 2010 decennial censuses, Nevada’s population is aging faster than the U.S. population. The two most aged counties are Esmeralda and Mineral Counties, which are adjacent to each other and the most isolated in the state. Douglas County has an older aging population and probably the next highest median age, but it is an amenity county attracting wealthy seniors.

Nevada’s population is changing. Hardcastle said. There are different perspectives regarding population density in the state. Populations within

the counties are very dispersed around the state and in some cases have very small population concentrations. Nevada is the least densely populated state in the country, but at the same time one of the most urbanized, although that does not quite hold on a county-to-county level. Nevada cities are growing, while rural counties are losing population. Defining populations as urban and rural based on population density does not always account for the issues discussed at the workshop, he said.

Definitions Matter

Is a county urban, rural, frontier, micro, or metro, Hardcastle queried. He observed these definitions arise in his work, and they matter. They can be analytical or operational for research purposes. They can be administrative, such as how to classify a unit of government and collect taxes. They can be programmatic, such as for grants or other programs. They can be colloquial, such as how people refer to themselves and see their place in the universe. Hardcastle said these definitions may conflict or complement each other. Someone may have a perception of being rural, but that perception may not necessarily match up to the micropolitan, metropolitan, official, or research-oriented definitions for their community. Distance and density may not always be a good indicator of ruralness or rurality.

How important is the geographic unit, Hardcastle asked. Talking about rural characteristics, he gave as an example a call he received when his population estimates were released. A reporter wrote as the headline about the estimates, “Nevada City Is Growing, Rural Counties Are Getting More Rural.” He noted the reporter’s implication that because counties were losing population and shrinking, they were becoming more rural.

He contrasted this to an idea by Popper and Popper (1987) and Popper et al. (1993) about the idea of Buffalo Commons for the Great Plains. They described the loss of population, but also the potential revival of that area through going back to a natural, more steady-state ecosystem where more of the natural grassland is reestablished. Loss of population and rural characteristics can have different interpretations, Hardcastle said. Losing population does not have to mean that the community is lost.

Hardcastle said that Warren (1978) talked about a community’s institutions being vertically and horizontally integrated. How those two points intersect is how an institution or a community is seen and what makes it unique. For instance, a hardware store could be locally owned and operated responding to local needs (horizontal integration) but get goods through a national or regional distribution chain (vertical integration). That can be compared to a national chain that establishes national

goals for the sale of goods and controls distribution (almost total vertical integration).

Governance, Knowledge, and Skill Capacity

Hardcastle noted that governance does not just mean local governments, but the larger institutional political system within a community. At times, older entrenched interests in communities have to address changing economies, economic situations, or even environmental situations. For example, he said, in one rural county, water is being depleted through mining while people still hold onto agricultural watering. There is an imbalance between the older agricultural interest and the mining interest. Also, local governments tend to have small staff and limited resources and are serviced by volunteers. They are often frustrated by having to deal with federal or state regulations, he observed.

Regarding the economy for local rural communities, Christaller (1933) talked about the idea of an intercepted economy. Some Nevada communities are isolated, but some also are intercepted, said Hardcastle, such as Elko, which is trying to grow between Salt Lake City and Reno. Those two very large economies are competing for any services that get in there. The other communities are going to be intercepted between them. In economies that boom and bust over time, as seen in the western region, overcapacity from the boom has to be reabsorbed over time.

Culture, Technology, Identity

Hardcastle noted that changing tastes or markets often influence movement of people, such as senior citizens returning to central cities. That is an indication that preferences are changing. To market themselves, for example, some locations look at instituting public art to draw people in and get them out of their cars. Other areas sell themselves as being patriotic and a place to visit.

Hardcastle talked about looking at Nevada’s counties through four definitions of rural and urban: the National Center for Frontier Communities, the National Center for Health Statistics, the Office of Management and Budget, and the USDA’s Rural-Urban Continuum Codes (RUCC). Micropolitan areas classified as frontier are on average 47 percent rural and have an RUCC of 6. Non-core frontier areas are on average 71 percent rural, with an RUCC of 8. Micropolitan areas are less dependent on resource extraction, but extraction remains a very prominent sector. The biggest difference is that they have larger heath care, retail, and tourist sectors. Micropolitan areas serve as service centers for the surrounding region, he explained.

The outlying county in this analysis, according to Hardcastle, was Storey County with 93 percent of its population classified as rural. It is classified as medium metro, not a frontier county with a 2 RUCC designation because it is part of the Reno metropolitan statistical area. It is also the site for the Tesla gigafactory with 6,500 direct jobs. Local agencies have been looking at the plant and its impact on the region. Some comparisons for the potential impact of agglomeration economics have been made to other metropolitan statistical areas, such as Provo, Utah. More appropriate comparisons might be to combined statistical areas because the Reno-Carson City-Fernley, Nevada, combined statistical area captures the region that will be impacted. He said this project illustrates two points by other panel members: location matters for agglomeration, and impact depends on the specific location. For example the impact of the Tesla plant would be different if it had been sited in Washoe County with different multiplier effects. Also, as David Plane pointed out, there are different levels of hierarchies, and the combined statistical area concept is one of those hierarchies.

Hardcastle addressed some issues relating to technology and change. For example, mining is often thought of as building tunnels or digging holes, but technological changes have increased production and lowered the cost per worker. In contrast, cable television was developed first for the rural part of the country in especially mountainous regions. That technology has become ubiquitous and has impacted the delivery of information across the country. It is interesting, he noted, that rural areas are now getting broadband.

PANEL DISCUSSION

Brown asked Plane about reconceptualizing density by changing the denominator so rather than a geographic unit, it would be a measure of time such as travel time, and then asked about the numerator. He suggested thinking about density as the number of social interactions or transactions or something else that expresses the reality of the social and economic life that occurs in geographic regions. Plane responded that traffic might be an interesting variable. However, he observed, one needs something that can be measured.

Brown observed that Barry Wellman, a network analyst, has worked on the density of social interactions to show how people in social relationships in metropolitan areas have changed over time; what appeared to be a loss of locality was a more extensive locality within the metropolitan areas (Wellman and Hampton, 1999). Plane commented that much big data research is very much in that mode.

Hardcastle asked Lobao if school employment was included in her

figures on total county government employment. She clarified respondents were asked to exclude school employment.

OPEN DISCUSSION

Brigitte Waldorf agreed with Plane’s observations that people’s activities should be viewed in both space and time, but said she sees challenges because it affects both the denominator and the numerator of an indicator. An example of where activity in space and time becomes very important is a labor market area because people often have two or three jobs. Plane responded that the single-minded focus on commuting time is the key issue. It has its problems, but asked if there are data to go beyond that. For example, he said, people migrate because it is a constrained world and they are likely to move to places where they can find jobs. He added for other kinds of activities, maybe the actual market areas for different goods and services to identify spatiality should be looked at. Lobao pointed out the policy-making unit is also important. To influence policy makers, there should be a focus on people’s lives. She commented because of decentralization, counties and municipalities are increasingly important to making policy.

Steven Turner (Southern Rural Development Center) commented on the barriers to economic development faced by counties as shown in Lobao’s survey. He noted the respondents in the most nonadjacent, nonmetropolitan category saw no problem with the quality of their public schools. He said that is clearly their perception rather than the outsider’s perception. Lobao responded that quality was not the real issue. This is similar to a quality-of-life measure, with subjective versus objective measures. County administrators subjectively assess their schools as good.

Gregory Hooks (McMaster University) observed that pressures for funding formulas and policy call for an answer. He said he has started to wonder how often it is useful to reduce a multidimensional concept to a zero-one variable.

Hardcastle referred to an analysis of workforce development in underserved rural communities conducted by a colleague. It was hard to find data at the county level and to relate a range of ruralness and urbanness to anything meaningful about workforce development. Sometimes, definitions need to be cleaned up for policy discussions so that conceptually there is a clearer, common sense theme that people who write grants, for example, could utilize.

Weber observed that rural classification systems need to assess the impact of context on outcomes. The context people live in affects whether they are poor and whether they have a job. Density affects outcomes and distance; remoteness affects outcomes. While it is not possible to change

remoteness, density can be changed a little. What else can be changed about context? A classification system that would identify places that need investment to change undesirable outcomes would be invaluable, he suggested.

Michael Ratcliffe commented that the Census Bureau receives many questions, challenges, and complaints about where boundaries are drawn. These challenges are made because the line that defines urban may keep a community from achieving a programmatic goal. For example, in examining the criteria for urbanized areas for the 2010 Census, they were considering the correct distance for jumping across low-density, intervening territory that separates two areas of high density, the core and an outlying urban use. They adopted 2.5 miles in 2000, but a 1.5-mile distance had been used from 1950 through 1990. One community wanted it to remain at 2.5 because that maximized their urban area for funding purposes. In contrast, the rural health community favored 1.5 miles so that they could minimize the urban area, drop below 50,000, and get the funding that they felt they deserved.

Ratcliffe observed these thresholds are fairly arbitrary. He suggested a continuum that allows geographers to say whether places are definitely urban or definitely rural would be useful. Flexibility is needed from a policy standpoint, he said, so that communities can find where they fit, and policy makers and decision makers can find a way to say that a mixed area might qualify a little bit for both. He said after 25 years of dealing with challenges, he does not think the dichotomy is working anymore.

Robert Gibbs commented that he found it interesting that Brown began his presentation quoting from Wirth (1938). He recalled that in the 1960s, in the United Kingdom, the urban-rural continuum was confounded when researchers in London found all characteristics that were supposed to apply to rural communities in urban areas, and vice versa. The idea of mapping a continuum from urban to rural crumbled. He said that he was attracted to the way Lobao looked at the changing role of the state, noting the importance of trying to understand the challenges and how they might change in different places.

He asked about work on the social relations that characterize different rural places, such as The Differentiated Countryside (Murdoch et al., 2003). That work identified different types of rural places according to the dominant social relations, such as the paternalistic countryside where large landowners held the power, the clientelistic countryside where the state held the power and people tried to get money from the state, and the contested countryside where urban income was moving in and disrupting established social relationships and norms. While they would be different in the United States, social relations differ in different parts of the rural United States, and the relationship with poverty would be key.

Ken Johnson, referring to James Fitzsimmons’ earlier presentation, remarked that he appreciated the problems the federal agencies face related where the rural/urban line is drawn. He questioned whether any set of definitions could actually satisfy all users. He referred to a statement that the classifications are intended for statistical purposes only. John Cromartie responded the reason for the concern is that definitions have consequences for funding. But he suggested better communication about the fact that rural classifications, especially in the United States, are fuzzy.

Hardcastle ended the session with a suggestion that something like health care services provided in a region be included in a definition to make the concept broader and more usable. He said that he picked health care because funding seems to revolve around it, and health care can vary by both size of place and by area.