Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Introduction

At least since von Thunen published Isolated State in 1826 (1966), researchers have studied mechanisms of population growth at the urban-rural fringe. The body of modern theory began to develop in earnest in the 1970s (e.g., Berry, 1970). The 1980s saw the theoretical development of spread-backwash theory (i.e., Gaile, 1980) and deconcentration/restructuring theories (e.g., Frey, 1993), with which this chapter is concerned. By the early 1990s, more descriptive, data-driven work elucidated the mechanisms of household and firm relocation to the urban fringe (e.g., Clark & Kuijpers-Linde, 1994). Soon thereafter, more predictive models of deconcentration and restructuring were developed (e.g., Renkow & Hoover, 2000), adding to the growing theory that urban fringe locations develop first by household relocation, with firm location following.

Scholars have continued to expand and refine this research by studying variation in growth patterns found internationally and within different demographic segments of the population. For example, research finds that while urban land conversion associated with population deconcentration is significantly higher in the United States than elsewhere, population growth at the urban-rural fringe is occurring in both “more developed” and “less developed” countries worldwide (e.g., Schneider & Woodcock, 2007). Additionally, life course stage affects migration patterns along the urban hierarchy (e.g., Plane, Henrie, & Perry, 2005). The continued improvement of deconcentration models expands demographic research by more precisely explaining the intricacies of population change in transitioning urban, rural, and suburban places.

Since 2000, scholars have re-popularized place-based economic development policies, primarily for isolated rural locations (e.g., Partridge & Rickman, 2006). Simultaneously, econometric modeling and policy research have focused on transportation, technology, infrastructure, and commuting-based development strategies for urban fringe locations with substantial linkages to the urban core (e.g., Renkow, 2003). The conclusions of this research favor policies managing or encouraging growth regionally (Partridge, Bollman, Olfert, & Alasia, 2007). Between these studies lie places neither geographically isolated nor significantly tied to central cities. As shown in this chapter, these counties sometimes neighbor metropolitan areas and experience suburbanization with the metropolitan counties at the same or at varying degrees of intensity or magnitude; sometimes these counties truly fall outside the geographic range of suburbanization.

This chapter focuses on these places by creating a model of deconcentration and restructuring reflecting spatial heterogeneity in regional growth around Chicago, IL, USA. This is accomplished by constructing a traditional Ordinary Least Squares (OLS) model of deconcentration, then using the coefficients from a Geographically Weighted Regression (GWR) to derive subregions within the study area. These subregions are incorporated into the OLS model, allowing measurement of the variation in and the spatial extent of deconcentration across the region. The results of this model are used to inform a discussion of rural and regional development policy for the United States, Canada, and Europe, the predominant sites of related studies. Since any temporal difference in movement to the periphery between firms and households must be accommodated through commuting, commuting is the lens through which deconcentration and restructuring are analyzed.

The remainder of this chapter follows in five sections. The following section provides background literature on the hypotheses of deconcentration and restructuring. The section entitled “Chicago, IL MSA Study Region” gives an overview of the study region and the components of its growth between 1990 and 2000. The form of the econometric analysis and background on the data and methods are provided in the section “Econometric Analysis”. The results section follows. Overall, the model shows spatial heterogeneity in the mechanism of deconcentration and a spatial extent that reaches beyond MSA boundaries. The final section discusses the policy implications.

Background

This section provides an overview of the literature on the theories of deconcentration and restructuring and popular metrics for their measurement. This literature supports the hypothesis that deconcentration and restructuring vary in magnitude across a region.

Deconcentration and Restructuring

Though they focus primarily on the role of information technology, Audirac and Fitzgerald (2003) provide an excellent overview of the concepts of deconcentration and restructuring. This review draws heavily on the sources identified there, including direct quotations to introduce each term.

As Audirac and Fitzgerald introduce it, “In the deconcentration group…we find works in the human ecology tradition of urban sociology and microeconomic neoclassical approaches in location decision theories” (2003, p. 482). This theory is straightforward: technology and infrastructure reduce the cost of travel and communication, allowing households to move to the periphery of a region. Peripheral areas afford larger lots and homes with the full range of bucolic amenities (Rouwendal & Meijer, 2001). Brian Berry (1970) was among the earliest scholars to discuss deconcentration. He posited its development on the compression of time (see also Fishman, 1990) and space, as permitted by technology, and the mobility of social classes, which would lead to increased education attainment and mobility. At its root, deconcentration stems from atomistic decision making about commuting, lifestyle amenities, and access to employment.

Conversely, the restructuring school “has its intellectual roots in Marxist political economy and regulation theories… .Since theories in this school are vastly heterogeneous, it can simply be said that they emphasize economic and spatial restructuring resulting from (1) technological change, which is the result of, and the transformational force affecting, the (capitalist) mode of production, and (2) the role of the state in shaping the conditions for economic growth (capital accumulation)” (Audirac & Fitzgerald, 2003, p. 483). One of the more consistent themes in the restructuring literature is the transformation of the urban hierarchy from one based on global ports to one based on global centers of command and control with the spatial dispersion of standardized or “less intellectual” (Storper, 1997) activities and back-office functions (Audirac & Fitzgerald, 2003; Coffey & Bailly, 1992; Sassen, 1994, 2002; Scott, 1988). Unlike the deconcentration literature, restructuring studies “reflect the regulation regimes and the interests of corporate and public-sector actors” (Audirac & Fitzgerald, 2003, p. 484).

A popular conceptual measurement for deconcentration and restructuring is the relationship between in-migration and out-commuting within a jurisdiction (usually the county). A positive relationship between in-migration and out-commuting is called “complementarity”; the inverse is “substitution” (Evers, 1989; Renkow & Hoover, 2000). Conceptually, if households are moving into counties and continuing to work elsewhere (complementarity), deconcentration is occurring. Periphery lifestyle amenities have outweighed commuting costs. Households moving into counties to replace commuting to those counties (substitution), are following corporate spatial movement decisions. Complementarity and substitution are conceptual measurements for the theoretical constructs of deconcentration and restructuring.

In constructing typologies, academics group observations to allow analysis of empirical data. Although productive, categorization of observations obscures within-group variability. In reality, deconcentration and restructuring (and complementarity and substitution) happen simultaneously within regions; as discrete concepts they are the polar ends of a spectrum of more plausible scenarios. Even while population deconcentration may dominate regional expansion, some households move nearer to work and some firms move into unsettled areas. Classifying a region as deconcentrating or restructuring as a whole masks the heterogeneity within the region.

Theoretically, there are many hybrid perspectives. Deconcentration and restructuring can be seen as simultaneous results of the interaction of information technology and development (Amirahmadi & Wallace, 1995). Deconcentration suggests that workers move to the suburbs for lifestyle amenities (e.g., Hirschorn, 2000). Restructuring argues that corporations move for profit gains. A hybrid theory suggests that while the New Economy catalyzes the spatial reorganization of metropolitan companies, some firms move to the periphery for the lifestyle amenities (Beyers, 2000), an atomistic approach to corporate decision making (see also Henton & Walesh, 1998). Spatial variation in lifestyle amenities and infrastructure provision are only two examples of the many potential forces suggesting the theoretical spatial heterogeneity of deconcentration and restructuring.

Chicago, IL MSA Study Region

This paper focuses on the Chicago-Naperville-Joliet CMSA plus its surrounding nonmetropolitan counties (Fig. 22.1). Only the smallest selection of counties surrounding Chicago excludes counties with obvious linkages to at least one other MSA. The selection shown in Fig. 22.1 extends far enough from Chicago to be bounded by smaller MSAs to which the Chicago fringe counties likely have linkages. Including these counties provides a coherent view of the relationship between commuting and migration for counties at the urban fringe outside Chicago. The sample extends roughly 110 miles outward from Chicago. Of the 65 counties in the region, 12 are in the Chicago CMSA, five are in the Milwaukee-Racine, WI CMSA, and 15 are spread across another 11 MSAs. In addition to Chicago’s economic engine, each of these MSAs exerts growth effects, potentially including population deconcentration.

Fig. 22.1
figure 22_1_189979_1_En

Study region

Full size image

The region’s outlying metropolitan counties grew the fastest by a wide margin between 1990 and 2000, at 17.0% (Table 22.1). Through the 1990s, four counties in the study region converted from nonmetropolitan to “outlying metropolitan” status. The fastest-growing county in the region (McHenry, IL), converted from “outlying metropolitan” to “central metropolitan” status over the decade. Of the ten fastest-growing counties in the region, four were central metropolitan, four were outlying metropolitan, and two were nonmetropolitan in 1990. These trends resemble the national experience, where suburban population growth outpaced central city growth between 1990 and 2000 (Pisarski, 2006). Figure 22.2 shows population change by county over this period. The strongest growth occurred to the west of Chicago and north into Wisconsin. Interestingly, while Chicago maintained its rank as the third largest city in the U.S. over the decade, the region as a whole and most counties in it (45 of 65) grew slower than the nation, which grew by 13.2% over the decade.

Fig. 22.2
figure 22_2_189979_1_En

County population change, 1990–2000

Full size image
Table 22.1 Population growth in the study region, 1990–2000
Full size table

Across the study region between 1990 and July 1999, the population grew by 1,245,416 net people through natural growth (births minus deaths) and lost 66,206 people on net via migration. The region lost over half a million people (net of –500,824) via domestic migration and gained (on net) 434,618 through international migration. Though these sources of change and their magnitudes seem surprising, they are not unusual; this pattern occurred in the eight largest U.S. cities between 1995 and 2000 (Pisarski, 2006). In megacities, the net migration rate is positive only in the age bracket 25–29 years, indicating the role of household formation and childbearing on regional demographic change (Plane et al., 2005). Evidence of lifecycle-related movement up and down the urban hierarchy appears in the fact that households moving into Chicago in the 1990s were smaller and earned less than households moving out of the Chicago MSA (Yu, 2009). Traditionally, young individuals or couples move to the city, start building careers and families, then out-migrate with higher incomes and larger households than when they arrived. Consequently, the roles of migration and commuting at the urban-rural fringe become increasingly important metrics for mechanisms of regional growth.

Given the empirical support for the theory of deconcentration over regional restructuring (e.g., Renkow & Hoover, 2000), it may seem likely that much of the spatial expansion of economic activity is done through commuting. Yet this is only part of the picture. An analysis of Bureau of Labor Statistics, Quarterly Census of Employment and Wages data from 1990 and 2000 shows strong growth in the number of business establishments across the region, with the strongest growth in the outlying metropolitan counties (Table 22.2; using Office of Management and Budget [OMB] 1999 definitions). Growth in the number of private establishments actually outpaced population growth in each of the three types of counties. Although previous research (and the conclusions of the analysis herein) find deconcentration rather than restructuring, clearly business movement toward the periphery influences regional growth.

Table 22.2 Number of establishments by county type, 1990–2000
Full size table

Bureau of Economic Analysis data demonstrates the magnitude of the economic consequences of commuting. Residents of the region who commuted outside their home county (to counties within or outside the region) earned nearly $92.9 billion, moving that money into the home county. Workers commuting to a county within the region (though not necessarily living in the region) moved nearly $94 billion across county lines (Bureau of Economic Analysis, REIS, Table CA91). It is critical for municipalities to analyze commuting to capture more of that multibillion dollar practice.

Econometric Analysis

Deconcentration theory posits that with decreasing transportation costs, people can afford more land, and so choose to commute to work. In restructuring, industry faces changing economic constraints and opportunities that motivate increased distance from the central city; workers follow (Audirac & Fitzgerald, 2003; Clark & Kuijpers-Linde, 1994; Renkow & Hoover, 2000). The deconcentration/restructuring model investigates the relationship between commuting and migration in a county at one point in time. Simultaneous out-commuting and in-migration indicate that deconcentration has occurred.

Using Renkow and Hoover (2000) as a starting point, the ability to out-commute from a county is assumed to face budget constraints where household earnings cannot exceed household expenditures, including commuting costs. Therefore, net commuting is modeled as a function of net migration and the following budget constraints: wage differential, housing cost differential, educational attainment differential, and distance. The econometric model uses county-county pairs as the unit of analysis, with the data lending itself to conclusions at that level of geography. Importantly, this deconcentration/restructuring model varies from a household location model, which would include all varieties of locational and housing amenities.

Equation (22.1) gives the empirical form, with the variable definitions and data sources following in Table 22.3.

Table 22.3 Variables for geographically weighted regression
Full size table

The empirical form given is

$$C_{ij} = f\left(M_{i},W_{ij},D_{ij},H_{ij},E_{ij} \right)$$
((22.1))

Where

C ij =:

net number of workers commuting from county i to county j, normalized by the employed population of county i

M i =:

net migration into county i in the previous period, normalized by the population in county i in the previous period

W ij =:

wage in county j minus wage in county i (*1000)

D ij =:

distance between counties i and j, using population-weighted centroids

H ij =:

standardized housing cost in county j minus county i

E ij =:

four-year college degree attainment rate in county j minus county i

Net commuting is normalized by the employed population of county i to scale the value of commuting. The wage, distance, and housing variables are included as significant budget constraints in the decision to migrate or commute. The wage data represents wages at the place of employment rather than residence. This figure is the relevant one in modeling commuting since people commute to earn a wage offered somewhere other than the home county. College education attainment includes all county residents 25+ who have earned a four-year degree or higher; those with some college or associates degrees are not counted as having attained a four year degree. The difference in educational attainment provides a crude measure of skill mismatch between counties, since labor demand is often skill-specific.

Finally, rather than using the difference in median housing costs for all units between counties, HUD’s Fair Market Rent statistics (Department of Housing and Urban Development, 2000) allow an estimate of the difference in housing costs for similar units. This marks a departure from the literature, where traditionally housing prices have been compared across geographic units at the median, without respect to characteristics (e.g., McMillen, 2004a). Using Fair Market Rents allows a control for the size and general quality of housing units. This is important considering the key demographic that moves into and out of megacities—young people and new households, respectively. Housing units of equal price in a central city and a suburb or smaller city are unequal, with unit size being one of several key distinctions (Pisarski, 2006). Housing size needs present budget constraints in residential location choices.

This paper relies primarily on three databases: the U.S. Census of Population and Housing (2000a), the U.S. Census Transportation Planning Package (Census, 2000b), and migration data from the Internal Revenue Service (IRS, 1990–2000). The county is the unit of analysis for two primary reasons. First, though the Census databases provide information for a finer level of geography, the IRS files are available only at the county level. Additionally, CTPP commuter flow data contains a trade-off between spatial resolution and data disclosure. In densely populated areas, data nondisclosure is minimal; however, in regions inclusive of less densely settled areas, nondisclosure below the county level inhibits analysis.

Additionally, to reduce the error in estimated distance traveled for commuters between counties i and j, block group population for 2000 was used to estimate the population-weighted centroid for each county. In some counties, large portions of the population live around a dominant town or city, which may or may not be in the center of the county. Thus, the population-weighted centroids are more likely to be closer to the points of origin and destination for commuters than are the geographic centroids of the counties. The full range of regression variables by data source used is given in Table 22.3.

The econometric analysis is completed in four stages. In the first stage, an OLS regression is carried out to create a baseline for comparison to the literature and against which to interpret the subsequent model. In stage two, the initial model is converted to a GWR. In stage three, a clustering algorithm uses the GWR coefficients to define subregions in the study area. Finally, dummy variables for the subregions are interacted with the variables in the empirical specification and re-tested via OLS to test the hypotheses that deconcentration varies with space and has a spatial limit beyond the MSA border. Although the analysis ultimately relies on a standard spatial regime approach, the definition of subregions incorporates an original application of the GWR method. The use of regimes provides a proxy for spatial heterogeneity.

Observations include the set of ij county pairs that had nonzero net commuting. The set includes only the observations with positive net commuting, as is established in the literature to avoid selection bias (Renkow & Hoover, 2000). Finally, non-neighboring ij pairs were excluded (using a second-order, first-order inclusive, queen weights matrix). Invoking a spatial limit helps to eliminate observations with commute flows so small as to be within a reasonable margin of error. The final sample size is 388 ij pairs. As its dependent variable, the model uses the log of net commuting to ensure linearity.

Geographically Weighted Regression (GWR)

Within R, the function gwr.sel (spgwr) assisted in the selection of the GWR bandwidth, which is 94012.66 m. On average, this bandwidth covers 13.7 counties including county i. GWR is a technique used “to examine the spatial variability of regression results across a region and so inform on the presence of spatial nonstationarity” (Fotheringham, Charlton, & Brunsdon, 1998, p. 1907). Its general form, GWR can be expressed as:

$$y_i = a_0 + \sum {a_k }\left(u_i ,v_i \right)x_{ik} + \varepsilon _i $$
((22.2))

where u and v are coordinates of the ith point, allowing a continuous surface of parameter values. This technique produces localized regression diagnostics (Fotheringham et al., 1998). To allow calibration of the model, points nearer to point i are given more weight in the estimation of the parameter value for point i,

$$\hat a\left(u_i ,v_i \right) = \left[X^T W\left(u_i ,v_i \right)X\right]^{ - 1} X^T W \cdot \left(u_i ,v_i \right)y.$$
((22.3))

One technical consideration of this approach is that it is meant to model values at i. However, the dependent variable used in this paper is the commuting flow between ij pairs of counties, meaning there are multiple data points for each sending county i. A hierarchical approach may be more ideal.Footnote 1 However, this paper uses GWR to delineate subregions on which to test the spatial heterogeneity of the mechanism of deconcentration, not as a positivistic, conclusion-drawing method. The statistical significance of subregions in the final specification is sufficient evidence that the GWR has functioned satisfactorily for the purpose of this research.

Results

Geographically Weighted Regression

The existing OLS model of commuting flows is robust and statistically significant. Not only do Renkow and Hoover (2000) report reasonable strength in their OLS models, but the straight OLS model of commuting near Chicago is strong (Table 22.4). These results are shown with White-corrected standard errors (White, 1980; R code for White correction by Gianfranco Piras and provided by Kathy Baylis). The model did not show multicollinearity. The variables common to this and the Renkow and Hoover (2000) approach show the same signs, giving a measure of external validity.

Table 22.4 OLS results from basic model, 2000
Full size table

The data was then modeled using GWR, the coefficients of which were used to cluster the sending counties (i of the ij pairs) into six groups. The choice of six groups creates reasonably spatially coherent subregions in comparison to other numbers of clusters. Figure 22.3 shows the subregions created by running a fuzzy clustering algorithm (using R) on the GWR coefficients. Only 60 of the original 65 counties are shown here; the other five (including Cook County, IL) are not in the positive half of the net commuting relationship with any neighboring county and are not included here or in the following section (Table 22.5).

Fig. 22.3
figure 22_3_189979_1_En

Subregions created by clustering the GWR coefficients

Full size image
Table 22.5 OLS output with subregional dummies
Full size table

Respecified OLS Model

Using the ij county pairs in subregion one as the comparison group, dummy variables for each region were interacted with each of the independent variables and put into a new OLS regression model. Here again, ij county pairs constitute the unit of analysis (n = 388). Results are given in Table 22.5. Table 22.5 elicits several conclusions that warrant interpretation and discussion, most notably: the spatial limits of deconcentration; the spatial heterogeneity of deconcentration; the consistency of signs across subregions; and the varying premium put on wages across space.

First, and perhaps most importantly, the model shows a spatial limit to deconcentration that is constrained yet reaches well beyond the MSA boundaries. The migration term shows statistical significance in two subregions, numbers four and six. Although these subregions most closely frame the Chicago and Milwaukee MSAs, more than half of their constituent counties lie outside MSA boundaries. Deconcentration occurs beyond the MSA, in a selection of outer-ring counties framing the region’s major cities. In contrast, the more remote portions of the study area do not demonstrate deconcentration or restructuring; growth here occurs through alternate mechanisms such as employment growth in the home county rather than through commuting to either Chicago or peripheral, lower-tier cities. As found in Ali, Olfert, and Partridge (2010), there is spatial heterogeneity in growth mechanisms according to geography and placement along the urban hierarchy.

Second, the model shows spatial heterogeneity in the magnitude of deconcentration. Subregion six has a migration coefficient of 0.45, signaling stronger suburbanization when compared to subregion four’s coefficient of 0.29. Relative to subregion four, subregion six more closely frames Chicago. Additionally, subregion six includes suburban Milwaukee. The difference in coefficients signals stronger suburbanization in subregion six, but may also signal that in-migrants in subregion four find employment more evenly through both commuting and the local economy; this suggests the possibility of growth in traditionally rural communities through partial suburban use.

Third, all independent variables show consistency of signs across subregions. This provides a measure of internal validity for the use of subregions and provides preliminary support for the hypothesis that, while the coefficients do vary, the set and general effect of budgetary constraints on commuting are consistent across the region. Finally, the variation in coefficients across subregions is minimal except in the wage term (and migration, as discussed above). The coefficient for the wage differential for subregion five is approximately double its value in subregions four and six. This suggests that the wage differential between Milwaukee County and its neighbors drives commuting more strongly than in other areas of the study region.

It is important to ask if the results shown in Table 22.5 indicate the significance of spatial heterogeneity in deconcentration because the independent variable values (x-bar) vary, because the coefficients vary, or due to an average effect. Using the county-level output from the GWR, Moran’s I values (calculated using a first-order Queen-based weight in GeoDa) for the β, x-bar, and βx terms for each independent variable (Table 22.6) overwhelmingly show that an average effect drives the significance of the model shown in Table 22.5. Both the coefficients and the values of independent variables vary across space. This warrants more investigation into the mechanisms of commuting (McMillen, 2004b). The coefficients may be biased toward having a spatial pattern by virtue of having been created through a GWR.

Table 22.6 Moran’s I values for GWR output
Full size table

Tables 22.5 and 22.6, taken together, reveal spatial heterogeneity in the magnitude of deconcentration and a spatial limit to its reach. Perhaps as interesting, Tables 22.5 and 22.6 reveal that regardless of the presence of deconcentration, the aggregated households within counties choose to expend their aggregate household budgets (for housing and commuting costs) similarly across space, with high relative wages in Milwaukee driving commuting more than the wage differential across the Chicago region.

Conclusion and Policy Recommendations

This work has shown spatial heterogeneity and limits in the geographic scope of population deconcentration in the region surrounding the Chicago CMSA. In its simplest interpretation, this work has three conclusions: the magnitude of deconcentration varies within a region; the geographic scope of deconcentration is much smaller than the universe of counties with proximity to both the megacity (Chicago) and lower-tier cities, like Kankakee, Illinois, and; with the exception of the wage differential between central and suburban Milwaukee, the budgetary constraints to commuting act similarly across the region, regardless of the presence of deconcentration. In particular, estimating the spatial limit of deconcentration signals the need for more in-depth understanding of the mechanisms of growth in counties at the urban fringe. These counties are neither remote nor connected via strong commuting streams to cities. Further research on urban fringe development could replicate the statistical method presented here for estimating the spatial extent of deconcentration.

This research adds a new level of precision to existing models of deconcentration and adds nuance to crude categorizations of metropolitan or nonmetropolitan counties in growth models. Scholars in the United States, Canada, and Europe have argued that place-based rural policies must recognize spatial heterogeneity and properly identify functional regions to be most effective (Partridge, Olfert, & Ali, 2009; Pezzini, 2001). Therefore, the findings of this chapter should prove useful to policymakers seeking more effective planning efforts and growth policies within regions.

This chapter has identified nonmetropolitan counties in subregions with significant ties to urban centers. Based on previous research in the United States, Canada, and Ireland, population retention or growth in rural places with substantial linkages to major urban areas can best be achieved through regional growth policies (Henry, Barkley, & Bao, 1997; Khan, Orazem, & Otto, 2001; Moss, Jack, & Wallace, 2004; Partridge et al., 2007; Partridge & Rickman, 2006). Supporting growth in urban centers that serve as employment hubs for neighboring rural counties positively impacts households in those communities by expanding employment opportunities within established commuting distances (Moss et al., 2004; Partridge & Rickman, 2006; Partridge et al., 2009; Portnov & Schwartz, 2009). These policies will be particularly effective for counties in the initial stages of deconcentration because delays in firms’ movements to the periphery result in a period of households’ increased reliance on commuting.

Based on the magnitude of the migration coefficient in this work, policymakers can estimate where counties fall in the range of deconcentration and restructuring at a particular point in time. Counties with higher migration coefficients are likely to be in early stages of deconcentration (Clark & Kuijpers-Linde, 1994; Renkow & Hoover, 2000), where households have deconcentrated but employment has not. Policymakers who correctly identify the current stage and predict the near-future stage of their county’s growth can better plan for the social service and business needs of their communities. Regional governments and planning agencies can use this information to target planning efforts towards areas likely to experience rapid growth or to encourage business development in areas well suited to becoming future employment hubs. This information is particularly relevant in countries where land use change dramatically outpaces population growth, such as the U.S., in some cases Canada, and in rare cases China (Schneider & Woodcock, 2007).

Regional transportation planning and investment aimed at improving the accessibility of urban clusters to rural workers is also supported by the research (Moss et al., 2004; Renkow & Hoover, 2000; Rephann & Isserman, 1994). Traditionally rural areas, particularly areas that have experienced local structural change and increased reliance on out-commuting, will benefit from such policies (i.e., Moss et al., 2004). Coordinating zoning and environmental policies with regional transportation plans has the potential to benefit residents and businesses in both rural and urban locations (Partridge et al., 2007). Based on this work’s findings, regional policies, plans, and investments should pay careful attention to the variation and extent of deconcentration within a region to avoid inefficient outcomes.

This research has also more clearly identified places where spread effects are unlikely to occur. While the counties in the study area are not geographically isolated, the policies recommended for such places are largely applicable. Based on prior research, isolated rural counties are less likely to benefit from regional policies attached to urban centers (Henry et al., 1997; Partridge et al., 2007; Partridge & Rickman, 2006; Renkow & Hoover, 2000). Instead, community-specific programs designed to improve a community’s vitality and its residents’ quality of life are more appropriate. Critics of rural place-based policies argue that they waste public dollars by artificially suppressing out-migration from areas unlikely to achieve self-sustaining population or employment levels and by creating employment opportunities likely to be awarded to new commuters and in-migrants (Partridge & Rickman, 2006; see also Bolton, 1992). However, community-specific programs should be particularly effective in isolated rural counties because remoteness decreases competition for jobs from commuters in proximate urban areas (Partridge & Rickman, 2006). Additionally, place-based programs in isolated rural places are more likely to identify and address the specific “contextual effects” that most influence a rural community’s vitality (Blank, 2005).

Research on successful nonmetropolitan counties suggests that building local capacity in the areas of entrepreneurialism, community leadership, social-capital, and community planning leads to positive prosperity and population outcomes (Cook et al., 2009; Green, 2008; Low, Henderson, & Weiler, 2005; Partridge & Rickman, 2006; Schultz, 2004). Community leadership can stimulate population retention or growth by improving housing, focusing on quality of life issues, and coordinating economic development strategies with neighboring nonmetropolitan communities (Cook et al., 2009; Henry et al., 1997; Khan et al., 2001; Partridge & Rickman, 2006). For the smallest counties, regionally cooperative economic development policies rewarding job creation are likely to be more successful than those targeting high-wage jobs (Khan et al., 2001). From the state and federal level, education investments, technical assistance, and technical infrastructure development are also appropriate aids to these communities (Blank, 2005; Cook et al., 2009; Duncan, 1999; Fuguitt & Beagle, 1996; Garcia-Milà & McGuire, 1992; Low et al., 2005; Oden & Strover, 2002; Partridge & Rickman, 2006; Strover, Oden, & Inagaki, 2002). Such policies will improve the quality of life for existing residents and may attract some new residents; however, they are less likely to attract firms (Henry et al., 1997) and may influence future out-migration of young adults as increased educational attainment drives workers to find higher incomes in urban areas (Berry, 1970; Cushing & Poot, 2004; Plane et al., 2005).

Finally, this chapter highlights counties neither geographically isolated nor significantly tied to cities. The mechanisms of population growth and factors of location key to effective policymaking in other nonmetropolitan counties are less important factors of growth in these places, and their policies should reflect this difference. For these counties, policymakers must act prudently to select a pragmatic, flexible policy mix. Most importantly, strategic planning should serve these areas by facilitating scenario planning and proactive community dialogue about growth goals. Counties wishing to maintain a traditionally rural status may embrace place-based development strategies tied to growth management policies that protect open space and restrict population growth (Nelson & Dawkins, 2004). Counties envisioning a longer-term transition to suburban land use may pursue infrastructure and residential amenity development to facilitate commuting to proximate urban areas. These areas should also strengthen relationships and coordinated planning efforts with neighboring jurisdictions, regional governments, and metropolitan planning agencies (Scott & Storper, 2003).

In conclusion, this chapter has evaluated a mechanism of population growth at the urban-rural fringe. Primarily, this work finds a spatial limit to deconcentration that exceeds and cross-cuts metropolitan boundaries while also finding spatial heterogeneity in subregions experiencing deconcentration. Public policy work elucidates development strategies for areas with strong urban attachments and place-based policies for isolated areas. The study region includes counties neither geographically isolated nor significantly tied to cities. One-size-fits-all approaches based on regional growth dynamics are not sufficient or appropriate for these places; instead, these counties should advance policy portfolios that emphasize community goals and leverage existing community assets.