Human Capital and Population Growth in Nonmetropolitan U.S. Counties

Abstract

Researchers have consistently shown that the stock of human capital in an area, measured as the share of the adult population with a college degree, is a strong predictor of future population growth. This article examines this relationship for U.S. nonmetropolitan counties and posits that student migration for higher education may play an important role. Students often move to an area for college and then stay in the area after their education is complete, causing the area’s educated population to grow. Empirical evidence suggests that student migration explains nearly all the greater in-migration to highly educated nonmetropolitan counties. Implications for nonmetropolitan brain drain are discussed.

Keywords

population growth migration human capital nonmetropolitan counties college

Local population growth has been a subject of considerable interest to both academics and policymakers and is an important indicator of an area’s desirability. For both metropolitan and nonmetropolitan areas, a growing population signals that an area is a desirable place to live and work. Researchers have suggested that a number of characteristics have important effects on an area’s population growth by affecting its desirability to current and prospective residents. Early researchers such as Carlino and Mills (1987) recognized the importance of local employment prospects and natural amenities. More recent research has investigated the relationship between the local human capital stock and local population growth. Overwhelmingly, these researchers have found strong and consistent evidence that the share of adults in an area with a college degree is positively correlated with local population growth for cities (Glaeser, Scheinkman, & Shleifer, 1995; Simon, 1998), counties (Clark & Murphy, 1996; Partridge, Rickman, Ali, & Olfert, 2008c), and metropolitan areas (Glaeser & Saiz, 2004; Glaeser & Shapiro, 2003; Gottlieb & Fogarty, 2003; Simon & Nardinelli, 2002).₁ Similarly, Berry and Glaeser (2005), Waldorf (2009), and Whisler, Waldorf, Mulligan, and Plane (2008) suggest that an area’s ability to attract and retain a well-educated workforce depends positively on its existing level of human capital.

This relationship between the share of college-educated workers and local population growth is thought to be due, in part, to the many other beneficial effects that educated workers have on their local labor markets. For one, researchers have suggested that educated workers make other workers in the area more productive. Consistent with this reasoning, a number of studies have shown that wages in an area increase with the local level of human capital, even after controlling for a worker’s own level of human capital (e.g., Iranzo & Peri, 2009; Moretti, 2004a; Rauch, 1993).2 This notion also receives support from the literature on the geographic localization of patent citations and its relationship with university research (Adams, 2002; Audretsch & Feldman, 1996; Jaffe, 1989; Jaffe, Trajtenberg, & Henderson, 1993). Other researchers have also suggested that educated workers improve the quality of life in the local area by facilitating the existence of endogenous consumer amenities, such as parks, theaters, museums, and tolerance of others (Florida, 2002; Shapiro, 2006).

A separate explanation advanced recently by Winters (2011) is that the connection between human capital and population growth is largely because of the growth of areas that are centers of higher education.3 Areas that have major colleges and universities receive considerable in-migration from students pursuing higher education. This is especially true for flagship state universities that bring in students from all over the state and from other states and countries (Alm & Winters, 2009). Many of the students who move to an area for higher education are likely to leave the area after their education is complete, but a significant number may develop location-specific human capital and stay in the area after their schooling is complete. The ability of centers of higher education to retain ex-students causes both population growth and an increase in the average level of human capital. Winters investigates this explanation for metropolitan areas by separately examining the effect of human capital on the migration rates of those enrolled in higher education and those not enrolled in higher education. Winters finds that the majority of the differential in-migration to high human capital metropolitan areas is attributable to persons enrolled in higher education.

This article builds on the work of Winters (2011) and others by investigating the importance of student migration in the relationship between human capital and population growth for nonmetropolitan counties in the United States. Nonmetropolitan areas are often different from metropolitan areas in important ways, and thus a separate study that focuses on nonmetropolitan areas seems warranted. In particular, nonmetropolitan areas may face greater challenges in attracting and retaining educated workers. Young, college-educated workers may often be attracted to large metropolitan areas because they offer higher wages and thicker markets for skilled labor (Chen & Rosenthal, 2008; Elvery, 2010; Plane & Jurjevich, 2009). Artz (2003) and others argue that “brain drain” is a very real problem for many nonmetropolitan areas of the United States. On the whole, nonmetropolitan areas are not keeping up with metropolitan areas in the share of their populations with college degrees, and this trend threatens the economic survival of these areas. Carr and Kefalas (2009) go even further to suggest that rural brain drain and the hollowing out of Middle America are harmful to the nation as a whole. Importantly though, there is considerable diversity in brain drain outcomes among nonmetropolitan areas. A number of nonmetropolitan counties have actually experienced a “brain gain” and seen their college graduate shares increase at even faster rates than the average metropolitan county (Artz, 2003). By examining the role of student migration in nonmetropolitan county population growth, this article offers insight into the importance of higher education institutions in attracting educated persons to nonmetropolitan areas.

The empirical results suggest that virtually all the relationship between human capital and nonmetropolitan county in-migration between 1995 and 2000 is due to the in-migration of students enrolled in higher education. Persons not enrolled in higher education are not moving to high human capital counties in large numbers. The results, therefore, suggest that the local level of human capital causes nonmetropolitan counties to grow because of students moving in for higher education and then staying after their education is complete. Furthermore, when we remove “college towns” from the analysis, the effect of the local human capital level on net migration is virtually zero. That is, the growth of high human capital nonmetropolitan counties is driven by college towns. These results also suggest that nonmetropolitan areas with a strong presence of higher education institutions are likely to be much more successful at growing their share of college graduates than areas without good access to higher education.

Theoretical Framework

Following a large literature beginning with Sjaastad (1962), this article views migration as an investment in human capital. Individuals maximize expected utility and choose to move to a new area if it gives them greater utility than their previous location, and the utility difference is large enough to compensate for the costs of moving. If moving costs are sufficiently small, individuals will move to the location that gives them the highest possible utility. Important though, the attractiveness of a particular location will vary across individuals. For example, several researchers have suggested that the young and the elderly are attracted to different locations (e.g., Chen & Rosenthal, 2008; Clark & Hunter, 1992; Conway & Rork, 2006; Plane & Heins, 2003; Plane & Jurjevich, 2009). Young people are generally attracted to areas with high wages and a strong labor market while the elderly often prefer areas with nice natural amenities but a low cost of living. Similarly, individuals will differ in the benefit they receive from living in an area with a high level of human capital.

Following Winters (2011), this article argues that students pursuing higher education are more likely than nonstudents to move to a high human capital area, because they are often centers of higher education. Several college town counties rank near the top of the list of nonmetropolitan counties in the share of adults with a college degree. These include, but are not limited to, counties that contain universities such as Washington State University, the University of Wyoming, Kansas State University, Montana State University, Mississippi State University, Oklahoma State University, and Southern Illinois University Carbondale. Consequently, persons enrolled in higher education may make up a large share of the in-migration to high human capital counties, because they are the ones who most benefit from moving to a college town.

Many of the students who move to college towns in large numbers are likely to leave the area once they complete their education. They will often move back to previous residences or move on to new locations.4 However, previous literature suggests that many people who move to an area for higher education will stay there after they have completed school (Blackwell, Cobb, & Weinberg, 2002; Groen, 2004; Groen & White, 2004; Hickman, 2009; Huffman & Quigley, 2002).5 This likely occurs because students often develop location-specific human capital in the area where they attend college (Berry & Glaeser, 2005; DaVanzo, 1983; Krupka, 2009). After a few years of college, recent in-migrants may have gained human capital that makes them more productive locally than elsewhere. This might include networks with professors and peers as well as relationships with local employers through student work and internships. Individuals may also develop location-specific human capital in consumption; they may have created valuable friendships and acquired a taste for local amenities (e.g., live college sports) in the area where they attended college.

If location-specific human capital frequently motivates recent in-migrants to remain in the area where they completed their education, areas with a considerable university presence are likely to grow faster than those without, and student migration will play an important role in this growth. Furthermore, the growth will be from highly educated individuals, and the average level of human capital in the area is also likely to increase. Thus, the growth of areas may be strongly affected by their ability to attract and retain individuals pursuing higher education.

Empirical Methods

This study examines the relationship between human capital and population growth by estimating several migration equations for nonmetropolitan counties in the United States between 1995 and 2000.6 More specifically, the following in-migration, out-migration, and net migration equations are estimated,

\begin{array}{l} M_{in} = β_{in} S + - Γ_{in} Z + μ_{in}, \\ μ_{in} = λ_{in} W μ_{in} + ε_{in}, \end{array}

(1)

\begin{array}{l} M_{out} = β_{out} S + Γ_{out} Z + μ_{out}, \\ μ_{out} = λ_{out} W μ_{out} + ε_{out} \end{array}

(2)

\begin{array}{l} M_{net} = β_{net} S + Γ_{net} Z + μ_{net}, \\ μ_{net} = λ_{net} W μ_{net} + ε_{net} \end{array}

(3)

where M_in is the county in-migration rate, M_out is the out-migration rate, M_net is the net migration rate, S is the share of the adult population with college degrees in a county, Z is a matrix of other variables found in previous literature to affect local population growth; µ_in, µ_out, and µ_net are spatially autocorrelated error terms; W is a spatial weighting matrix; λ_in, λ_out, and λ_net are spatial error coefficients; and ε_in, ε_out, and ε_net are mean zero error terms. For the results presented, W is structured as a row-standardized matrix with weights based on the inverse of their distance to a given county and a cutoff of 100 miles; that is, nearer counties are given greater weight than more distant counties and counties more than 100 miles away are given zero weight. The main results, however, are robust to several alternative specifications of the spatial weight matrix. Because of the spatial error correlation, the equations are estimated using maximum likelihood (ML) instead of ordinary least squares (OLS; Anselin, 1988).7

Because high human capital areas are often centers of higher education, β_in is expected to be positive, suggesting that a disproportionately large number of people are moving to high human capital counties. Similarly, if many people leave an area after completing their education, we also expect human capital to be associated with high rates of out-migration, so β_out is expected to be positive as well. However, some people will likely stay in an area after completing their education causing β_in to exceed β_out and the local population to grow. Therefore, it is expected that β_net will also be positive.

The migration equations above are first estimated for the entire population to gauge the overall effects. Migration equations are then estimated separately for persons 16 years and older by whether they are currently enrolled in higher education. Analysis of Census 2000 County-to-County Migration files reveals that 15% of all persons 16 years and older living in a different county in 2000 than in 1995 were enrolled in higher education in 2000. If the relationship between human capital and population growth is largely due to persons enrolled in higher education, then the effect of human capital on the in-migration rate of those enrolled should be disproportionately large. As a further test of the hypothesis advanced, this article also estimates separate migration equations for 5-year age groups. The expectation is that persons in their peak college-going years may represent a disproportionately large share of the relationship between the local human capital level and in-migration.

Data

The migration data in this article come from the Census 2000 County-to-County Migration files. This data set includes only internal migrants; international in-migrants and out-migrants are not included. A person is considered a migrant if he or she lived in a different county in 2000 than he or she did in 1995. Although defining a person’s place of residence is straightforward for most people, it is a bit more complicated for some young people away at college. In a sense, some such students have two residences: the place where they reside while attending school and the place where they reside in between school sessions, often being their parents’ residences. The Census questionnaire instructed respondents that students away at college are to be counted as residents of the places where they attend college and not their parents’ residences. Unfortunately though, there could still be some misreporting of residences, especially prior residences of persons who were enrolled in college 5 years prior to the census. If so, the effects of the college share on out-migration may be understated, and the effects of the college share on net migration may be overstated. To partially address this concern, we also briefly look at the effect of the share of adults with a college degree on population growth.

Gross in-migration to an area is computed as the sum of all persons in a county who lived in another county in the United States 5 years prior. Gross out-migration is the sum of all persons who lived in a given county in 1995, but lived in a different county in the United States in 2000. Net migration is computed as gross in-migration minus gross out-migration. Gross in-migration rates for nonmetropolitan counties are computed by dividing gross in-migration by the estimated population of the county in 1995. Gross out-migration rates and net migration rates are computed similarly. When the migration rates are computed separately by enrollment status or by age, the total population of the county continues to be used as the population base. This is done to allow for easier interpretation of each group’s contribution to the overall migration rate.

Data for the explanatory variables in the regressions come from several sources. Time-varying explanatory variables are measured as of 1990, so that they are not affected by migration during the period under consideration. The share of the adult population in a county with a college degree comes from the U.S. Counties database. Several additional explanatory variables are also computed from U.S. Counties data. These include the county population, per capita income, and the share of employment in manufacturing. The quantity and quality of natural amenities are expected to affect nonmetropolitan migration (Gottlieb & Joseph, 2006; McGranahan, 1999; McGranahan, Wojan, & Lambert, 2010; Rappaport, 2007), and data on several natural amenities are obtained from the U.S. Department of Agriculture (USDA) Economic Research Service (ERS). These include the mean January temperature, mean January sun hours, mean July temperature, mean July relative humidity, the percentage of the county area that is covered by water, and a topography score that ranges from 1 to 24 with higher values indicating a more mountainous topography. Partridge, Rickman, Ali, and Olfert (2008a, 2008b, 2008c) suggest that population growth for nonmetropolitan areas may depend on proximity to urban areas, in part because close proximity makes rural-to-urban commuting a more viable option (Partridge, Ali, & Olfert, 2010). Therefore, this article also includes as explanatory variables the population of the nearest metropolitan area, the distance to the nearest metropolitan area, the distance to the nearest metropolitan area with a population of at least 250,000, the distance to the nearest metropolitan area with a population of at least 500,000, and the distance to the nearest metropolitan area with a population of at least 1,500,000. Finally, indicator variables for the Midwest, South, and West regions are also included.

Summary statistics for the main variables used in this study are presented in Table 1. The sample includes 2,004 nonmetropolitan counties in the contiguous United States. Several variables have been rescaled to aid the presentation of results. The reported summary statistics reflect the scaling used in the regressions to follow. The mean in-migration and out-migration rates for the total population are both about 0.18 indicating that on average, about 18% of a nonmetropolitan county’s population lived in a different county in 2000 than they did in 1995. The summary statistics also indicate that for the average county, migration of those not enrolled is a considerably larger portion of overall migration than migration of those who are enrolled.8 However, the relative importance of college students in overall migration varies considerably across counties.

Table 1.

Summary Statistics

Variable	Mean	SD	Min	Max
In-migration rate (total)	0.179	0.066	0.039	0.630
Out-migration rate (total)	0.183	0.051	0.086	0.507
Net migration rate (total)	−0.003	0.062	−0.307	0.427
In-migration rate (enrolled)	0.014	0.028	0.000	0.336
Out-migration rate (enrolled)	0.025	0.012	0.000	0.151
Net migration rate (enrolled)	−0.011	0.028	−0.146	0.249
In-migration rate (not enrolled)	0.132	0.051	0.028	0.532
Out-migration rate (not enrolled)	0.128	0.037	0.058	0.400
Net migration rate (not enrolled)	0.004	0.050	−0.223	0.404
Share with bachelor’s degree	0.117	0.047	0.037	0.534
Population in millions	0.022	0.020	0.000	0.174
Per capita income in $1000s	10.182	1.825	3.417	26.755
Manufacturing share	0.144	0.114	0.000	0.615
January temperature/100	0.315	0.123	0.011	0.634
January sun/100	1.530	0.335	0.480	2.600
July temperature/100	0.756	0.056	0.555	0.879
July humidity/100	0.543	0.148	0.140	0.800
Topography/100	0.091	0.066	0.010	0.210
% Water area/100	0.035	0.097	0.000	0.750
Population of nearest metro in millions	0.277	0.405	0.040	4.123
Distance to nearest metro/1000	0.060	0.037	0.010	0.253
Distance to metro with/population >250,000/1000	0.107	0.087	0.020	0.579
Distance to metro with/population >500,000/1000	0.132	0.092	0.020	0.629
Dist. to metro with/population >1,500,000/1000	0.194	0.103	0.030	0.629
West	0.140	0.347	0.000	1.000
South	0.430	0.495	0.000	1.000
Midwest	0.384	0.486	0.000	1.000

Note. Sample includes 2,004 nonmetropolitan counties. All migration rates are for internal migration between 1995 and 2000 and are computed with the total population in 1995 as the base. Migration rates for those enrolled and not enrolled in higher education are for the population aged 16 years and older. Time-varying explanatory variables are measured as of 1990.

Empirical Results

The migration equations are first estimated for the entire population. These results are presented in Table 2. The first column presents the results for the in-migration rate, the second column presents the results for the out-migration rate, and the third column presents the results for the net migration rate. As expected, the share of adults in the county with a college degree has a strong positive correlation with the in-migration rate. In other words, more people are moving to highly educated places than less educated places. The coefficient estimate of 0.571 suggests that increasing the college share by 0.1 increases the in-migration rate by 0.057. This is a considerable effect. According to the results in the second column, the local human capital level is positively correlated with the out-migration rate as well with a coefficient of 0.436. On net though, the local human capital level brings in more people than it pushes out, and the college-educated share has a strong positive correlation with the net migration rate with a coefficient estimate of 0.123. This result is consistent with findings from previous literature that a more educated populace causes an area’s population to grow.

Table 2.

Migration to and From Nonmetropolitan Counties, 1995-2000

	In	Out	Net
Share with bachelor’s degree	0.571** (0.036)	0.436** (0.024)	0.123**, (0.038)
Population in millions	−0.448** (0.073)	−0.441** (0.048)	0.000 (0.077)
Per capita income in $1000s	−0.002* (0.001)	−0.001 (0.001)	−0.001 (0.001)
Manufacturing share	−0.110** (0.015)	−0.052** (0.010)	−0.050** (0.016)
January temperature/100	0.103** (0.034)	0.060** (0.019)	0.052 (0.031)
January sun/100	0.018* (0.007)	0.010* (0.004)	0.009 (0.007)
July temperature/100	−0.182** (0.042)	−0.080** (0.024)	−0.121** (0.034)
July humidity/100	−0.083** (0.022)	−0.076** (0.012)	0.005 (0.020)
Topography/100	0.074* (0.029)	−0.036* (0.018)	0.109** (0.029)
% Water area/100	0.022 (0.016)	−0.008 (0.010)	0.034* (0.017)
Population of nearest metro in millions	0.006* (0.003)	0.000 (0.002)	0.006 (0.003)
Distance to nearest metro/1000	−0.293** (0.066)	0.018 (0.040)	−0.279** (0.065)
Distance to metro with population >250,000/1000	−0.048 (0.059)	0.044 (0.035)	−0.091 (0.056)
Distance to metro with population >500,000/1000	0.000 (0.055)	0.028 (0.032)	−0.024 (0.052)
Distance to metro with population >1,500,000/1000	−0.136** (0.035)	−0.035 (0.020)	−0.100** (0.032)
West	0.034* (0.016)	0.011 (0.009)	0.029* (0.014)
South	0.037** (0.014)	−0.003 (0.008)	0.041** (0.013)
Midwest	0.033* (0.013)	0.005 (0.007)	0.031** (0.012)
Constant	0.292** (0.034)	0.223** (0.018)	0.065** (0.023)
Spatial error (λ)	0.614** (0.004)	0.487** (0.010)	0.496** (0.008)
Adjusted R²	.426	.562	.261

Note. Sample includes 2,004 nonmetropolitan counties. Standard errors are in parentheses.

p ≤ .05. **p ≤ .01.

As discussed above, there may be problems with misreporting of previous residence in the data, especially for persons who were attending college 5 years prior to the census. If so, the effects of the college share on out-migration may be understated, and the effects of the college share on net migration may be overstated. We partially address this concern by briefly looking at the effect of the share of adults with a college degree on actual population growth (results not shown). When we regress county population growth between 1990 and 2000 on the explanatory variables, we get a coefficient of 0.300 that is statistically significant at the 1% level. Cutting the effect in half to account for the longer time period gives a coefficient of 0.150, which is a little more than the coefficient for net migration between 1995 and 2000. The difference might be partially attributable to international in-migration, differences in fertility and mortality, and asymmetry between the first and second halves of the decade. Given all these considerations, the results for population growth are quite similar to the results for net migration and reduce concerns that measurement error causes major problems for the net migration regressions.

The results in Table 2 also suggest that many of the other variables have important effects on migration as well. However, we will later see that these variables often have differing effects on student and nonstudent migrants. Because the effects of these variables are a secondary interest of this article, we will hold off on discussing the results for these variables until the next subsection when we estimate separate migration equations for persons enrolled in higher education and persons not enrolled.

Migration by Enrollment in Higher Education

The results in Table 2 tell us that the share of adults in an area who are college educated is positively correlated with future population growth. However, Table 2 does not tell us much about the role of student migration in the relationship between the local level of human capital and population growth. More specifically, this article is interested in the extent to which the positive relationship between human capital and nonmetropolitan population growth is attributable to students moving to an area for higher education and then staying in the area after their education is complete. To investigate this, the migration equations are next estimated separately for persons enrolled in higher education and persons not enrolled in higher education. For these equations, the migration rates include only persons aged 16 years and older, but the results are unaffected by including persons younger than 16 years. The results are presented in Table 3.

Table 3.

Nonmetropolitan Migration by Enrollment in Higher Education, 1995-2000

	Enrolled in higher education				Not enrolled in higher education
	In	Out	Net	In	Out	Net
Share with bachelor’s degree	0.574** (0.014)	0.081** (0.006)	0.488** (0.015)	0.001 (0.028)	0.337** (0.019)	−0.334** (0.031)
Population in millions	0.148** (0.029)	−0.053** (0.013)	0.202** (0.030)	−0.488** (0.057)	−0.356** (0.038)	−0.130* (0.063)
Per capita income in $1000s	−0.009** (0.000)	0.000 (0.000)	−0.009** (0.000)	0.005** (0.001)	−0.001 (0.001)	0.006** (0.001)
Manufacturing share	0.005 (0.006)	−0.007** (0.002)	0.011 (0.006)	−0.097** (0.012)	−0.038** (0.008)	−0.053** (0.013)
January temperature/100	−0.016 (0.011)	−0.022** (0.004)	0.007 (0.011)	0.120** (0.026)	0.066** (0.015)	0.055* (0.026)
January sun/100	−0.005* (0.003)	0.002 (0.001)	−0.007** (0.003)	0.020** (0.006)	0.005 (0.003)	0.016** (0.006)
July temperature/100	0.042** (0.012)	0.010* (0.005)	0.032** (0.012)	−0.210** (0.030)	−0.088** (0.018)	−0.137** (0.027)
July humidity/100	0.012 (0.007)	0.000 (0.005)	0.011 (0.007)	−0.061** (0.017)	−0.057** (0.010)	0.004 (0.016)
Topography/100	−0.024* (0.011)	−0.020** (0.004)	−0.004 (0.011)	0.088** (0.023)	−0.011 (0.014)	0.095** (0.024)
% Water area/100	−0.018** (0.006)	−0.005* (0.003)	−0.014* (0.006)	0.038** (0.013)	−0.003 (0.008)	0.045** (0.014)
Population of nearest metro in millions	0.002 (0.001)	0.000 (0.001)	0.003 (0.001)	0.003 (0.003)	0.000 (0.002)	0.003 (0.003)
Distance to nearest metro/1000	−0.085** (0.023)	0.021* (0.010)	−0.104** (0.024)	−0.152** (0.051)	0.000 (0.032)	−0.137** (0.053)
Distance to metro with population >250,000/1000	0.013 (0.020)	0.017* (0.008)	−0.006 (0.020)	−0.052 (0.046)	0.029 (0.027)	−0.079 (0.045)
Distance to metro with population >500,000/1000	−0.002 (0.019)	−0.004 (0.007)	0.003 (0.019)	0.010 (0.042)	0.021 (0.025)	−0.009 (0.042)
Distance to metro with population >1,500,000/1000	−0.051** (0.011)	0.008 (0.004)	−0.058** (0.011)	−0.080** (0.027)	−0.044** (0.016)	−0.037 (0.026)
West	0.009 (0.005)	0.004 (0.002)	0.006 (0.005)	0.022 (0.012)	−0.003 (0.007)	0.024* (0.011)
South	0.008 (0.004)	−0.001 (0.002)	0.009* (0.004)	0.022* (0.011)	−0.002 (0.006)	0.025* (0.010)
Midwest	0.010* (0.004)	0.002 (0.002)	0.009* (0.004)	0.018 (0.010)	0.000 (0.006)	0.018 (0.010)
Constant	0.014 (0.008)	0.009** (0.003)	0.004 (0.008)	0.224** (0.023)	0.182** (0.013)	0.042* (0.018)
Spatial error (λ)	0.446** (0.014)	0.332** (0.016)	0.416** (0.014)	0.601** (0.006)	0.494** (0.010)	0.495** (0.009)
Adjusted R²	.477	.462	.426	.403	.500	.251

Note. Sample includes 2,004 nonmetropolitan counties. Standard errors are in parentheses.

p ≤ .05. **p ≤ .01.

Looking first at the in-migration equations, the positive relationship between the share of adults with a college degree and in-migration rates appears to be entirely because of persons enrolled in higher education. The coefficient for those enrolled is 0.574 and highly statistically significant. The coefficient for those not enrolled, however, is virtually zero (0.001). These results suggest that students moving for higher education account for virtually all the greater in-migration to highly educated nonmetropolitan counties. Persons not enrolled in higher education are not moving to these areas in large numbers.

The results in Table 2 also suggested that out-migration rates are positively correlated with the local human capital level. If the relationship between in-migration and human capital is due to student in-migrants, then the positive relationship between human capital and out-migration rates is likely due in part to persons leaving after completing their education. The separate out-migration equations by enrollment in Table 3 shed light on this as well. The human capital stock is positively correlated with the out-migration of both those enrolled in higher education and those not enrolled. The majority of the effect, however, is due to persons not enrolled. This is consistent with the hypothesis that individuals move to high human capital areas for higher education, but often leave after completing their education.

Turning to the net migration results, the share of adults with a college degree is positively and significantly correlated with the net migration rate for those enrolled with a coefficient estimate of 0.488. For those not enrolled in higher education, the human capital stock is negatively and significantly correlated with net migration with a coefficient estimate of −0.334. Thus, highly educated areas are on net gaining people enrolled in higher education and losing people not enrolled in higher education. This is again consistent with the hypothesis that people move to highly educated areas for education but often leave once their formal education is done. Importantly though, the net migration coefficient for those enrolled is larger in absolute value than the net migration coefficient for those not enrolled. This suggests that some of the people who move for higher education end up staying in the area after their educations are complete. This causes highly educated areas to experience faster population growth than their less educated counterparts.

The results in Table 2 suggested that many of the other variables have important effects on migration. The results in Table 3 suggest that these variables often have differing effects on the migration of those enrolled in higher education and those not enrolled. Focusing on the net migrations equations for brevity, the results suggest that the local population increases the net migration of students but decreases the net migration of nonstudents. Per capita income has the opposite effect—it decreases the net migration of students but increases the net migration of nonstudents. The manufacturing share has a positive but insignificant coefficient for the net migration of those enrolled, but a negative and significant effect on the net migration of those not enrolled. January temperature and January sun have positive and significant effects on the net migration of those not enrolled, but for the net migration rate of those enrolled January temperature has an insignificant effect and January sun has a significantly negative effect. July temperature significantly reduces the net migration of those not enrolled, but significantly increases the net migration of those enrolled. July humidity is insignificant for both. Topography and the percentage of the area covered by water have significantly positive effects on the net migration of those not enrolled, but negative coefficients for the net migration of those enrolled, though the effect for topography is not significant. These results suggest that persons not enrolled in higher education are drawn to areas with nice natural amenities, but persons pursuing higher education are generally not. The distance to the nearest metro area has a negative and significant effect on the net migration of both those enrolled and those not enrolled, supporting the findings of Partridge et al. (2008a, 2008b, 2008c). The distance to the nearest metro with a population of at least 1.5 million also has a significantly negative effect on the net migration of persons enrolled in higher education. This suggests that proximity to a large metropolitan area is more important for student migrants than nonstudent migrants. The West region dummy is significantly positive for the net migration of persons not enrolled, the South dummy is significantly positive for both, and the Midwest dummy is significantly positive for those enrolled. Finally, the spatial error coefficient is significant in all the equations confirming that standard errors are spatially correlated. Consequently, methods that do not account for spatial error correlation such as OLS are likely to produce inconsistent standard errors and potentially incorrect inferences. The magnitude and significance of the spatial error term also suggest that there are important unobserved regional factors that affect nearby counties in similar ways. Thus, policy makers for nonmetropolitan counties might often be well served by working with other nearby counties to make their entire regions more desirable to potential migrants.

Migration by Age Group

Evidence in the previous section suggests that the relationship between the local human capital level and population growth is primarily because of persons pursuing higher education. If this is true, we also might expect the age distribution of persons moving to high human capital areas to be skewed toward persons in their primary college-going years. Similarly, we might expect the age distribution of persons leaving high human capital areas to be skewed toward age groups that are likely to have recently completed higher education or dropped out of school. This article, therefore, next examines the effect of human capital on migration by 5-year age groups. Table 4 presents the results for the effect of the share of adults with a college degree on in-migration, out-migration, and net migration rates between 1995 and 2000. Though the results are not reported, the regressions in Table 4 also include the additional explanatory variables included previously. The full results are available on request. The coefficients in Table 4 are also illustrated graphically in Figure 1.

Table 4.

Nonmetropolitan Migration by Age Group, 1995-2000

	In	Out	Net
Share with bachelor’s degree 1990
Age group (years)
5-9	−0.007* (0.003)	0.017* (0.004)	−0.021** (0.004)
10-14	−0.011** (0.003)	0.004 (0.003)	−0.015** (0.004)
15-19	0.162** (0.006)	−0.011** (0.004)	0.169** (0.007)
20-24	0.391** (0.010)	0.048** (0.007)	0.339** (0.012)
25-29	0.055** (0.005)	0.253** (0.007)	−0.193** (0.007)
30-34	0.013** (0.005)	0.075** (0.004)	−0.061** (0.005)
35-39	0.004 (0.005)	0.032** (0.003)	−0.028** (0.005)
40-44	−0.007 (0.004)	0.014** (0.003)	−0.021** (0.005)
45-49	−0.004 (0.003)	0.011** (0.003)	−0.014** (0.004)
50-54	−0.002 (0.003)	0.010** (0.002)	−0.011** (0.003)
55-59	−0.010** (0.003)	0.003 (0.002)	−0.012** (0.003)
60-64	−0.009** (0.003)	−0.002 (0.002)	−0.007* (0.003)
65-69	−0.010** (0.003)	−0.008** (0.002)	−0.002 (0.003)
70-74	−0.006** (0.002)	−0.010** (0.002)	0.004 (0.003)
75-79	−0.003** (0.001)	−0.003 (0.002)	−0.002 (0.002)
80-84	−0.002 (0.001)	−0.002 (0.002)	0.001 (0.002)
85+	0.000 (0.002)	−0.004 (0.003)	0.004 (0.003)

Note. Sample includes 2,004 nonmetropolitan counties. Regressions also include the additional explanatory variables included in Tables 2 and 3. Standard errors are in parentheses.

p ≤ .05. **p ≤ .01.

As seen in Table 4 and Figure 1, the largest effect of the local human capital level on in-migration is for persons in their peak college-going years. The effect is largest for the 20- to 24-year-old age group with a significant coefficient of 0.391, followed by the 15- to 19-year-old, and 24- to 29-year-old age groups with significant coefficients of 0.162 and 0.055, respectively. For all other age groups, the effect of human capital on in-migration is much smaller and even significantly negative for some age groups. The out-migration results by age are also as expected. The effect of human capital on out-migration is largest for the 25- to 29-year-old age group, followed by the 30- to 34-year-old and 20- to 24-year-old age groups.

The net migration results in the third column of Table 4 suggest that on net high human capital areas are gaining workers in their peak college-going years (those aged 15-19 and 20-24 years old) and losing workers in age groups who are likely to have recently finished their education (those aged 25-29 and 30-34 years old), which is also nicely illustrated in Figure 1. Importantly though, the effect of human capital on the net inflow of the younger cohorts exceeds the effect of human capital on the net outflow of the older cohorts. Examining migration by 5-year age groups, therefore, provides further support for the hypothesis that high human capital areas are growing faster than low human capital areas because many of the persons who move to an area for higher education remain in the area after their schooling is complete.

Controlling for “College Towns”

To provide additional evidence, we next estimate the effect of the share of adults with college degrees on migration separately for “college town” and noncollege town counties. This article defines college towns using a definition similar to that suggested by Plane and Heins (2003). A county is defined as a college town if the age profile of its in-migration flows is skewed toward persons aged 15 to 24 years old.9 More specifically, a county is defined as a college town if its share of in-migrants aged 15 to 24 years is more than one standard deviation above the mean share across all nonmetropolitan counties. This definition identifies 205 nonmetropolitan counties as college towns.

Table 5 reports the results from reestimating the equations in Table 2 separately for college town and noncollege town counties. The regressions also include the additional explanatory variables in Table 2 and a spatially lagged error term, but for brevity only the results for the college share variable are presented. Panel A reports the results that only include college town counties. The percentage of adults with a college degree has a statistically significant effect on in-, out-, and net migration with coefficients of 0.770, 0.512, and 0.260, respectively. Not surprisingly, the effects for college town counties are even larger than the effects for all nonmetropolitan counties in Table 2. In particular, the net migration effect more than doubles.

Table 5.

Controlling for “College Town” Counties, 1995-2000

	In	Out	Net
A. Only including college town counties (N = 205)
Share with bachelor’s degree	0.770** (0.072)	0.512** (0.056)	0.260** (0.071)
Adjusted R²	.496	.527	.201
B. Excluding college town counties (N = 1,799)
Share with bachelor’s degree	0.138** (0.043)	0.134** (0.027)	−0.003 (0.048)
Adjusted R²	.405	.561	.283
C. Controlling for the share of the population aged 15-24 years (N = 2,004)
Share with bachelor’s degree	0.229** (0.040)	0.265** (0.026)	−0.038 (0.043)
Share of population aged 15-24 years	0.604** (0.043)	0.256** (0.029)	0.340** (0.047)
Adjusted R²	.477	.575	.296

Note. Regressions also include the additional explanatory variables included in Tables 2 and 3. Standard errors are in parentheses.

p ≤ .01.

When the sample is restricted to noncollege town counties in Panel B, the percentage of adults with a college degree continues to have a positive and statistically significant effect on both in-migration and out-migration rates, with coefficients of 0.138 and 0.134. These effects, however, are considerably smaller than the corresponding estimates for college town counties and for the full sample. Even more important, the effects on in- and out-migration for noncollege towns are of roughly equal magnitude, and the effect on net migration is virtually nil (−0.003) and not statistically significant. In other words, the human capital level appears to have no effect on net migration for noncollege town counties. The relationship between net migration and the share of adults with college degrees found in Table 2 is entirely because of net migration to highly educated college town counties.

An additional way to examine the importance of college towns is to reestimate the full sample regressions adding the share of the population aged 15 to 24 years as an additional explanatory variable. This additional variable is expected to be highly correlated with the extent to which the county is a college town and including it should greatly reduce the coefficient on the share of college graduates in the migration equations. Panel C of Table 5 reports the results of adding the share of the population ages 15 to 24 years to the full sample regressions in Table 2. The share of the population aged 15 to 24 years is significant in all three regressions with coefficients of 0.604, 0.256, and 0.340, respectively, suggesting that this variable is an important predictor of migration flows for nonmetropolitan counties. As expected, controlling for the share aged 15 to 24 years reduces the effect of the college graduate share on both in- and out-migration to 0.229 and 0.265, respectively. More important, the coefficient on the share of college graduates for net migration is small (−0.038) and statistically insignificant. Thus, when we include this additional measure of the extent to which a county is a college town, the college graduate share is unrelated to net migration.

Additional Results

Table 6 presents some additional results. In Panel A, we consider the robustness of the results in Table 3 to using 1993 metropolitan area definitions. As seen in Panel A, the coefficient estimates and significance levels are nearly identical to the corresponding estimates in Table 3. In Panel B, we examine the robustness of replacing regional fixed effects with state fixed effects. Again the results are virtually identical to the baseline results in Table 3 and are therefore quite robust to including state fixed effects.

Table 6.

Additional Results

	Enrolled in higher education			Not enrolled in higher education
	In	Out	Net	In	Out	Net
A. Using 1993 metro definitions
Share with bachelor’s degree	0.587** (0.015)	0.082** (0.006)	0.502** (0.015)	0.019 (0.029)	0.353** (0.019)	−0.330** (0.032)
Adjusted R²	.488	.456	.437	.393	.501	.242
B. Including state fixed effects
Share with bachelor’s degree	0.602** (0.014)	0.081** (0.006)	0.518** (0.015)	0.006 (0.029)	0.357** (0.019)	−0.352** (0.032)
Adjusted R²	.499	.470	.454	.405	.506	.265
C. Including an interaction term for metro proximity
Share with bachelor’s degree	0.664** (0.024)	0.095** (0.010)	0.564** (0.025)	−0.062 (0.048)	0.410** (0.032)	−0.473** (0.052)
Distance to nearest metro/1000	0.113* (0.049)	0.052* (0.021)	0.063 (0.051)	−0.289** (0.097)	0.161* (0.064)	−0.444** (0.106)
Share with bachelor’s × Distance to metro/1000	−1.542** (0.332)	−0.242 (0.143)	−1.298** (0.346)	1.064 (0.651)	−1.253** (0.436)	2.385** (0.722)
Adjusted R²	.483	.462	.430	.404	.502	.254

Note. Regressions also include the additional explanatory variables included in Tables 2 and 3. Standard errors are in parentheses.

p ≤ .05. **p ≤ .01.

In Panel C of Table 6, we examine the effect of adding to the regressions a term for the interaction between the share of college graduates and the distance to the nearest metro area. The regressions already included separate measures for these variables, but we might be interested in whether the effect of the share of college graduates depends on proximity to metropolitan areas. The results in Panel C suggest that there is indeed some interaction effect. The interaction term has a negative effect on in-migration and net migration of college students with coefficients of −1.542 and −1.298, respectively. These negative interaction effects suggest that the effects of human capital on student in-migration decline with the distance to the nearest metro area. Thus, there is a positive effect of being near a metro area on student in-migration. However, the interaction effect for persons not enrolled in higher education goes in the opposite direction. Although the interaction term does not significantly affect the in-migration of nonstudents, it does significantly affect out-migration and net migration of persons not in college with coefficients of −1.253 and 2.385, respectively. This suggests that the proximity to a metro area actually increases the outflow of recent graduates from nonmetropolitan college towns.

Conclusion

This article examines the well-established positive relationship between human capital and population growth for nonmetropolitan counties in the United States. Following Winters (2011), this article suggests that because high human capital areas are often centers of higher education, persons moving to college towns to pursue higher education may play an important role in the positive relationship between human capital and population growth in nonmetropolitan areas. Consistent with this hypothesis, this article finds that the positive relationship between human capital and in-migration is almost entirely due to persons enrolled in higher education. For persons not enrolled in higher education, there is virtually no relationship between in-migration and the local level of human capital.

After completing their education, many of those who recently moved to an area for higher education move back to a previous area or on to a new location. However, many recent student in-migrants develop location-specific human capital, such as relationships with local employers and friends and a taste for local amenities, and these local attachments often make their current location the best place to start their careers and families. As a result, many recent student in-migrants remain in an area after completing their schooling, causing centers of higher education to grow in population and increase their local level of human capital. Importantly though, this article finds that the stock of human capital only increases net migration to nonmetropolitan areas that are college towns. Nonmetropolitan areas without access to higher education institutions have much greater difficulty attracting educated workers and thus are less able to build their human capital stock and grow their population.

The question then becomes what, if anything, can policy makers do to help stem the brain drain from nonmetropolitan areas and prevent the hollowing out of Middle America (Artz, 2003; Carr & Kefalas, 2009)? Some might suggest that small, isolated, rural areas are a lost cause and the best thing we can do for their residents is to encourage them to relocate to areas with better employment prospects. Many others, however, believe that America needs its small towns, and we can and should do something to help them survive and even thrive. A number of policies have been advanced to make nonmetropolitan areas more desirable places, but there seems to be no consensus on the best approach. Examples include efforts to build up local recreational amenities, student loan forgiveness programs, better K-12 education, improving infrastructure, and access to high-speed internet, and economic development strategies that directly try to bring in industries. This article suggests that access to higher education may play a special role. Unfortunately, it is simply not possible to have a world class university in every small town in America. However, there may be large social benefits from increasing the density of university satellite campuses and community colleges in parts of states that are relatively underserved.

Online learning may present an especially promising strategy for increasing access to higher education in nonmetropolitan areas, provided that the quality of the education is not diminished. Although a number of higher education institutions, both public and private, have been involved in online learning for some time now, we may only be scratching the surface of the benefits that online learning could provide to small, remote areas. Online learning increases access to higher education but does not force residents to leave home to acquire that education. However, one of the major obstacles to online learning is that most potential students have very little experience with it and are generally unsure about it. One way to increase experience with online learning is to offer more online courses to high school students. Ideally, this could also be accomplished with Advanced Placement (AP) or dual enrollment courses that potentially allow students to earn high school and college credits simultaneously. There might also be considerable benefits from increased efforts to market online programs, especially to nontraditional students who live in remote areas. There is, of course, an important role here for university officials, but there also could be an important role for local officials. Local officials and educators could set up programs in local schools and libraries that help students navigate online higher education programs, either as individuals or in small groups. Increased information and local support for online programs would likely increase the number of nonmetropolitan residents who participate. Online learning, therefore, is a potentially important way to increase access to higher education to persons in remote areas, and there are many things that local leaders could do to make online learning more accessible.

Footnotes

Acknowledgements

The author thanks two anonymous referees for helpful comments.

The author(s) declared no conflicts of interest with respect to the research, authorship, and/or publication of this article.

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Adams

J. D.

(2002). Comparative localization of academic and industrial spillovers. Journal of Economic Geography, 2, 253-278.

Alm

Winters

J. V.

(2009). Distance and intrastate college student migration. Economics of Education Review, 28, 728-738.

Anselin

(1988). Spatial econometrics: Methods and models. Boston, MA: Kluwer Academic.

Artz

(2003, December). Rural brain drain: Is it a reality? Choices, 4th Quarter, 11-15.

Audretsch

D. B.

Feldman

M. P.

(1996). R&D spillovers and the geography of innovation and production. American Economic Review, 86, 630-640.

Berry

C. R.

Glaeser

E. L.

(2005). The divergence of human capital levels across cities. Papers in Regional Science, 84, 407-444.

Blackwell

Cobb

Weinberg

(2002). The economic impact of educational institutions: Issues and methodology. Economic Development Quarterly, 16, 88-95.

Carlino

G. A.

Mills

E. S.

(1987). The determinants of county growth. Journal of Regional Science, 27, 39-54.

Carr

P. J.

Kefalas

M. J.

(2009). Hollowing out the middle: The rural brain drain and what it means for America. Boston, MA: Beacon Press.

10.

Chen

Rosenthal

S. S.

(2008). Local amenities and life-cycle migration: Do people move for jobs or fun? Journal of Urban Economics, 64, 519-537.

11.

Clark

D. E.

Hunter

W. J.

(1992). The impact of economic opportunity, amenities, and fiscal factors on age-specific migration. Journal of Regional Science, 32, 349-365.

12.

Clark

D. E.

Murphy

C. A.

(1996). Countywide employment and population growth: An analysis of the 1980s. Journal of Regional Science, 36, 235-256.

13.

Conway

K. S.

Rork

J. C.

(2006). State death taxes and elderly migration: The chicken or the egg? National Tax Journal, 59, 97-128.

14.

Corcoran

Faggian

Mccann

(2010). Human capital in remote and rural Australia: The role of graduate migration. Growth and Change, 41, 192-220.

15.

DaVanzo

(1983). Repeat migration in the United States: Who moves back and who moves on? Review of Economics and Statistics, 65, 552-559.

16.

Drucker

Goldstein

(2007). Assessing the regional impacts of universities: A review of current approaches. International Regional Science Review, 30, 20-46.

17.

Elvery

J. A.

(2010). City size and skill intensity. Regional Science and Urban Economics, 40, 367-379. doi:10.1016/j.regsciurbeco.2010.05.00610.1016/j.regsciurbeco.2010.05.006

18.

Faggian

McCann

(2009). Human capital, graduate migration and innovation in British regions. Cambridge Journal of Economics, 33, 317-333.

19.

Faggian

McCann

Sheppard

(2007a). Human capital, higher education and graduate migration: An analysis of Scottish and Welsh students. Urban Studies, 44, 2511-2528.

20.

Faggian

McCann

Sheppard

(2007b). Some evidence that women are more mobile than men: Gender differences in UK graduate migration behavior. Journal of Regional Science, 47, 517-539.

21.

Florida

(2002). Bohemia and economic geography. Journal of Economic Geography, 2, 55-71.

22.

Florida

Mellander

Stolarick

(2008). Inside the black box of regional development: Human capital, the creative class, and tolerance. Journal of Economic Geography, 8, 615-649.

23.

Franklin

R. S.

(2003). Migration of the young, single and college educated: 1995 to 2000 (Census 2000 Special Reports, CENSR-12). Washington, DC: Government Printing Office.

24.

Gibbs

R. M.

(1995). Going away to college and wider urban job opportunities take highly educated youth away from rural areas. Rural Development Perspectives, 10, 35-44.

25.

Gibbs

R. M.

(2000). College graduates in the nonmetropolitan South: Origins and prospects. Southern Rural Sociology, 16, 36-59.

26.

Glaeser

E. L.

Saiz

(2004). The rise of the skilled city. Brookings-Wharton Papers on Urban Affairs, 47-94.

27.

Glaeser

E. L.

Scheinkman

J. A.

Shleifer

(1995). Economic growth in a cross-section of cities. Journal of Monetary Economics, 36, 117-143.

28.

Glaeser

E. L.

Shapiro

J. M.

(2003). Urban growth in the 1990s: Is city living back? Journal of Regional Science, 43, 139-165.

29.

Goldstein

Drucker

(2006). The economic development impacts of universities on regions: Do size and distance matter? Economic Development Quarterly, 20, 22-43.

30.

Gottlieb

P. D.

Fogarty

(2003). Educational attainment and metropolitan growth. Economic Development Quarterly, 17, 325-336.

31.

Gottlieb

P. D.

Joseph

(2006). College-to-work migration of technology graduates and holders of doctorates within the United States. Journal of Regional Science, 46, 627-659.

32.

Groen

J. A.

(2004). The effect of college location on migration of college-educated labor. Journal of Econometrics, 121, 125-142.

33.

Groen

J. A.

White

M. J.

(2004). In-state versus out-of-state students: The divergence of interest between public universities and state governments. Journal of Public Economics, 88, 1793-1814.

34.

Henderson

J. V.

(2007). Understanding knowledge spillovers. Regional Science and Urban Economics, 37, 497-508.

35.

Hickman

D. C.

(2009). The effects of higher education policy on the location decision of individuals: Evidence from Florida’s Bright Futures Scholarship program. Regional Science and Urban Economics, 39, 553-562.

36.

Huffman

Quigley

J. M.

(2002). The role of the university in attracting high tech entrepreneurship: A Silicon Valley tale. Annals of Regional Science, 36, 403-419.

37.

Iranzo

Peri

(2009). Schooling externalities, technology, and productivity: Theory and evidence from U.S. states. Review of Economics and Statistics, 91, 420-431.

38.

Jaffe

A. B.

(1989). Real effects of academic research. American Economic Review, 79, 957-970.

39.

Jaffe

A. B.

Trajtenberg

Henderson

(1993). Geographic localization of knowledge spillovers as evidenced by patent citations. Quarterly Journal of Economics, 108, 577-598.

40.

Krupka

D. J.

(2009). Location-specific human capital, location choice and amenity demand. Journal of Regional Science, 49, 833-854.

41.

Lange

Topel

(2006). The social value of education and human capital. In Hanushek

Welch

(Eds.), Handbook of the economics of education (pp. 459-509). Amsterdam, Netherlands: Elsevier.

42.

Lendel

(2010). The impact of research universities on regional economies: The concept of university products. Economic Development Quarterly, 24, 210-230.

43.

McGranahan

D. A.

(1999). Natural amenities drive rural population change (Agricultural Economic Report No. AER781). Washington, DC: U.S. Department of Agriculture, Economic Research Service.

44.

McGranahan

D. A.

Wojan

T. R.

Lambert

D. M.

(2010). The rural growth trifecta: Outdoor amenities, creative class and entrepreneurial context. Journal of Economic Geography. Advance online publication. doi:10.1093/jeg/lbq00710.1093/jeg/lbq007

45.

Moretti

(2004a). Estimating the social return to higher education: Evidence from longitudinal and repeated cross-sectional data. Journal of Econometrics, 121, 175-212.

46.

Moretti

(2004b). Human capital externalities in cities. In Henderson

J. V.

Thisse

(Eds.), Handbook of regional and urban economics (pp. 2243-2291). Amsterdam, Netherlands: Elsevier.

47.

Partridge

M. D.

Ali

Olfert

M. R.

(2010). Rural-to-urban commuting: Three degrees of integration. Growth and Change, 41, 303-335.

48.

Partridge

M. D.

Rickman

D. S.

Ali

Olfert

M. R.

(2008a). Employment growth in the American urban hierarchy: Long live distance. B.E. Journal of Macroeconomics(Contributions), 8, 1-36.

49.

Partridge

M. D.

Rickman

D. S.

Ali

Olfert

M. R.

(2008b). The geographic diversity of U.S. nonmetropolitan growth dynamics: A geographically weighted regression approach. Land Economics, 84, 241-266.

50.

Partridge

M. D.

Rickman

D. S.

Ali

Olfert

M. R.

(2008c). Lost in space: Population growth in the American hinterlands and small cities. Journal of Economic Geography, 8, 727-757.

51.

Plane

D. A.

Heins

(2003). Age articulation of U.S. inter-metropolitan migration flows. Annals of Regional Science, 37, 107-130.

52.

Plane

D. A.

Jurjevich

J. R.

(2009). Ties that no longer bind? The patterns and repercussions of age-articulated migration. Professional Geographer, 61, 4-20.

53.

Rappaport

(2007). Moving to nice weather. Regional Science and Urban Economics, 37, 375-398.

54.

Rauch

J. E.

(1993). Productivity gains from geographic concentration of human capital: Evidence from the cities. Journal of Urban Economics, 34, 380-400.

55.

Shapiro

J. M.

(2006). Smart cities: Quality of life, productivity, and the growth effects of human capital. Review of Economics and Statistics, 88, 324-335.

56.

Simon

C. J.

(1998). Human capital and metropolitan employment growth. Journal of Urban Economics, 43, 223-243.

57.

Simon

C. J.

Nardinelli

(2002). Human capital and the rise of American cities, 1900-1990. Regional Science and Urban Economics, 32, 59-96.

58.

Sjaastad

L. A.

(1962). The costs and returns of human migration. Journal of Political Economy, 70, 80-93.

59.

Waldorf

B. S.

(2009). Is human capital accumulation a self-propelling process? Comparing educational attainment levels of movers and stayers. Annals of Regional Science, 43, 323-344.

60.

Whisler

R. L.

Waldorf

B. S.

Mulligan

G. F.

Plane

D. A.

(2008). Quality of life and the migration of the college-educated: A life-course approach. Growth and Change, 39, 58-94.

61.

Winters

J. V.

(2011). Why are smart cities growing? Who moves and who stays. Journal of Regional Science, 51, 253-270.