Intercity Passenger Rails: Facilitating the SpatialSpillover Effects of Population and Employment

Growth in the United States, 2000–2010Bishal Bhakta Kasu1 and Guangqing Chi2

Abstract: This research examines the association that intercity passenger rails have with population and employment growth at the countylevel in the continental United States from 2000 to 2010. This research adopts an integrated spatial regression approach that incorporates bothspatial lag and spatial error dependence. The data come from the US Census Bureau, the Bureau of Transportation Statistics, the LandDevelopability Index, and the National Atlas of the United States. Population change and employment change are regressed on intercitypassenger rails, controlling for 14 socioeconomic variables. Intercity passenger rails are measured by the number of intercity passenger railterminals in each county. The results suggest that the associations that intercity passenger rails have with population and employment changeare both direct and indirect. Intercity passenger rails have a negative and direct association with population and employment change from2000 to 2010. The Great Recession during this period may have compelled people to move out of their home county in search of jobs; havingintercity passenger rails facilitated this process. The results also indicate that intercity passenger rails have a positive and indirect associationwith population and employment change. Population change and employment change in one county affect population and employment inadjacent counties. This indirect association shows the spatial spillover effect of population and employment growth through passenger rails.The indirect association does not come from within the county; rather, it is a disperse effect from neighbors. This research suggests thatintercity passenger rails, although built long ago, still play an important role in facilitating the spread of change and the integration of localcommunities into a larger regional economy. DOI: 10.1061/(ASCE)UP.1943-5444.0000477. © 2018 American Society of Civil Engineers.

Author keywords: Intercity passenger rail; Transportation; Population change; Employment change; Spatial econometrics; Spatialspillover effect.

Introduction

Railroads are considered one of the most important innovations ofAmerican economic growth (Fogel 1962; White 2008). They influ-enced the rise of corporations, the development of agriculture, andthe growth of manufacturing and interregional trade. The railroadsalso played a significant role in the patterns of early settlement,migration, population growth, and urbanization of the United States(Fogel 1962; Hedges 1926; Jenks 1944; Kirby 1983; White 2008).

Railroads flourished in the United States during the 1840s, andthe prosperity lasted throughout the nineteenth century (Itzkoff1985). It was a prime time for railroads when they were the prin-cipal modes of long-distance transportation, carrying goods andpeople. Passenger trains dominated the locomotive world at thattime, but now the glory has vanished. Several factors played im-portant roles in the demise of passenger rails. Some of these werethe unequal distribution of public funds, absence of a dedicated

funding source, competition with private vehicles, availability ofa strong intercity passenger bus and aviation network, failed mar-keting, inadequate infrastructures, and high fares (Hurst 2014;Itzkoff 1985). However, recently intercity passenger rail ridershiphas grown quickly in the United States, making it one of the fastestgrowing modes of transportation (Puentes et al. 2013).

The experiences of many countries in Europe and Asia show thatpassenger rails exert a positive impact on urban development andredevelopment (Okada 1994) because they are helpful in increasingemployment (Loukaitou-Sideris et al. 2013; Topalovic et al. 2012),enhancing economic productivity (Ryder 2012), boosting real estatemarkets (Loukaitou-Sideris et al. 2013), and increasing tourism(Loukaitou-Sideris et al. 2013; Okada 1994; Ryder 2012). In 2009,the US government recognized the economic vitality of passengerrails and considered the expansion and development of a passen-ger rail system as a stimulus to the macroeconomy (Goetz 2012;Grunwald 2010; Hurst 2014). Provision of federal funding for theimprovement of the passenger rail system was made through theAmerican Recovery and Reinvestment Act of 2009, and it was ex-pected that this provision would fight the economic crisis at thattime by creating new jobs and promoting economic growth (Gama2017). Even though in 2017 the Trump administration proposed adifferent mechanism for federal funding, the administration hasmade rebuilding and modernizing infrastructure, including rail-roads, a high priority (Friedberg 2018; White House 2018, 2017).

Most research on railroads in the United States focuses onfreight trains; intercity passenger trains are not currently considereda primary mode of passenger transportation. Furthermore, of thewritings on passenger rails, most are motivated by a vested interest,and very few are objective (Levinson 2012). Proponents focus on

1Postdoctoral Researcher, Dept. of Agronomy, Horticulture andPlant Science, South Dakota State Univ., 1390 College Ave., Brookings,SD 57007. Email: bbkasu@gmail.com

2Associate Professor, Dept. of Agricultural Economics, Sociology, andEducation, and Director of the Computational and Spatial Analysis Core ofthe Population Research Institute and Social Science Research Institute,Pennsylvania State Univ., 112E Armsby, University Park, PA 16802(corresponding author). ORCID: https://orcid.org/0000-0003-0888-7964.Email: gchi@psu.edu

Note. This manuscript was submitted on October 31, 2017; approved onMay 10, 2018; published online on September 18, 2018. Discussion periodopen until February 18, 2019; separate discussions must be submitted forindividual papers. This paper is part of the Journal of Urban Planning andDevelopment, © ASCE, ISSN 0733-9488.

© ASCE 04018037-1 J. Urban Plann. Dev.

J. Urban Plann. Dev., 2018, 144(4): 04018037

Dow

nloa

ded

from

asc

elib

rary

.org

by

Gua

ngqi

ng C

hi o

n 09

/18/

18. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

the social and economic benefits, and opponents highlight the hugecosts as well as the impracticality of the infrastructure, as the size ofthe country is large and population density is low (Levinson 2012;Peterman et al. 2009). Therefore, an understanding of the relation-ship between intercity passenger trains and population and employ-ment change is crucial. This study is important in this contextbecause it examines the association that intercity passenger railshave with population and employment change (growth or decline).

It should be noted that this study focuses on the associations thatpassenger rails have with growth rather than the bidirectional causalrelationships between rails and growth. Specifically, the relation-ships are separated into a direct association that passenger railshave with growth and an indirect association through spatial spill-over effects to be achieved in an integrated spatial regression ap-proach at the county level using continental US data from 2000 to2010. The results suggest that intercity passenger rails, althoughbuilt long ago, still play an important role in facilitating the spreadof growth and the integration of local communities into a largerregional economy.

Literature Review

Transportation is one of the factors that link economy and popu-lation by providing access to different geographical areas (Lichterand Fuguitt 1980; Thompson and Bawden 1992; van den Heuvelet al. 2014). In both strong and weak economies, transportationplays a role in the distribution (and redistribution) of populationand employment (Ory and Mokhtarian 2007). Many scholars be-lieve transportation is an essential factor for economic growth aswell as for the social well-being of communities (Lichter andFuguitt 1980). Research shows that areas with access to transpor-tation infrastructure have higher average economic growth rates(Briggs 1981; Ozbay et al. 2006; van den Heuvel et al. 2014).Access to transportation has a positive association with employ-ment growth, labor supply, and the willingness of individuals tosupply labor. The economic impact is not necessarily limited toareas adjacent to rail terminals; neighboring counties could alsobenefit from increased levels of accessibility. Such developmentalimpacts of transportation have been discussed for a long time invarious academic disciplines, including sociology, economics, ruraland urban development, and geography (Boarnet and Haughwout2000; Chi et al. 2006).

Demographics and Economy in the 2000s

In terms of population change, the population of the United Statesjumped from 281.4 million in 2000 to 308.7 million in 2010, in-dicating a 9.7% growth (US Census Bureau 2011). But populationgrowth was not uniform across the country. During this period,most of the growth occurred in the South and the West, accountingfor 84.4% of the total population growth. In 2010, most of the pop-ulation (83.7%) lived in 366 metropolitan areas; the rest (16.3%)were in nonmetropolitan areas. Major population growth (83%)between 2000 and 2010 occurred in suburban areas (HousingAssistance Council 2012), indicating a continuation of the subur-banization process that occurred during the 2000s. Populationgrowth was uneven not only at the regional level but also at thecounty level. The counties that gained population are concentratedalong the coasts (Pacific, Atlantic, and Gulf) as well as along thesouthern borders. The counties that lost population are concentratedin Appalachian, Great Plains, Mississippi Delta, Great Lakes, andnorthern border areas.

One of the most important events of the 2000s economy was theGreat Recession, in which US economic activity slowed and the

amount of goods produced and services offered fell significantly(US Bureau of Labor Statistics 2012). The country faced oneof the longest and most severe recessions since World War II(Brown 2009). During this decade, the housing market deteriorated,most states saw employment declines, and the unemployment rateescalated. Rural counties in particular experienced dramatic in-creases in unemployment rates (Housing Assistance Council 2012).The Great Recession reshaped employment (and population) dis-tribution throughout the country (Hertz et al. 2014; Rickman andGuettabi 2015; US Bureau of Labor Statistics 2012).

Previous Research

Earlier works on the relationship between transportation and pop-ulation growth were conducted from the perspective of human ecol-ogy (Duranton and Turner 2012; Lichter and Fuguitt 1980; Markand Schwirian 1967; Schnore 1957; White 2008). The human eco-logical perspective essentially argues that demographic change isthe response to changes in the available technologies and localenvironments. Even though there are multiple studies on transpor-tation from the human ecological perspective, those works do notexplore the relationship between transportation and populationgrowth (or decline) in a systematic way. Some works are fromthe perspective of the impact of transportation (highways) on pop-ulation growth during the 1970s, but the results of these works areambiguous (Voss and Chi 2006), partly because of their limitedscope and failure to adopt a holistic approach (Chi 2010; Voss andChi 2006; White 2008). For example, the studies are limited to in-terstate highways, to one stage of highway development, to ruralareas, or to only one direction (e.g., the impact of transportation onpopulation growth and not the other way).

The impacts of transportation on population can be direct andindirect (Chi 2010; Voss and Chi 2006). Direct impacts include theimposition of rights-of-way on residential housing, agriculturallands, and natural wilderness (Coffin 2007; Moore et al. 1964).This impact is mostly negative, causing demolition of residentialhousing and perhaps affecting the population composition of thearea. The indirect impacts stem from the growth or decline in theeconomy, changes in employment opportunities, and changes inthe physical environment. Access to the transportation infrastruc-ture plays an important role in these economic changes, which areultimately linked with population distribution and redistribution(Boarnet and Haughwout 2000; Lichter and Fuguitt 1980).

Theories of Transportation

The role of transportation in population and employment changeshas long been debated in the context of urban development, sub-urban sprawl, central cities’ decline, and inter-/intrametropolitanaccessibility (Boarnet and Haughwout 2000). The relationships be-tween transportation and change (growth or decline) in populationand employment are well described in the regional economics lit-erature, especially with regard to growth pole theory. A growth poleis an urban location that is the hub of economic growth, constantlyinteracting with surrounding areas for the distribution or redistrib-ution of growth (Darwent 1969; Thiel 1962). This theory has twomain concepts—spread and backwash. Spread refers to a situationwhere the growth of one place causes growth in the surroundingareas, and backwash refers to when growth in a location occursat the cost of surrounding areas’ development. These concepts iden-tify the geographic relationships between the urban area and adja-cent rural areas in terms of economic growth and development(Henry et al. 1997); if growth is more dependent on transportation,the effects of spread and backwash will be stronger (Chi 2010).

© ASCE 04018037-2 J. Urban Plann. Dev.

J. Urban Plann. Dev., 2018, 144(4): 04018037

Dow

nloa

ded

from

asc

elib

rary

.org

by

Gua

ngqi

ng C

hi o

n 09

/18/

18. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

The study of the relationship between transportation andpopulation growth is becoming more specific. Contemporary re-search explores the impact on population change from differentperspectives—for example, the role of transportation (as an agentto redistribute population across locations), the double causal rela-tionship (the impact of transportation on population change andvice versa), various developmental stages (preconstruction, con-struction, and postconstruction of highways), and the expansion ofthe transportation infrastructure, focusing on highways (Chi 2010;Chi et al. 2006; Voss and Chi 2006). These studies incorporate for-mal spatial dimensions, a research method that has been neglectedin the past.

Impacts of Railroads

Because the research on intercity passenger rails is limited, thescope of this literature review is expanded to include passenger railsin general. Some studies have examined the impacts of railroadson population and employment changes (Atack and Margo 2011;Bollinger and Ihlanfeldt 1997; Levinson 2008a, b; White 2008).The results of such studies show great variation in the impacts ofrailroads on local economic development. For example, Bollingerand Ihlanfeldt (1997) found no impact of the Metropolitan AtlantaRapid Transit Authority (MARTA) on population and employmentchanges in station areas in Atlanta; most likely, MARTA does notincrease accessibility effectively because the city is already servedby a well-established network of highways. But their study did findan alteration in public and private employment compositions: eventhough the total employment did not change, public-sector employ-ment increased in the vicinity of transit stations (Bollinger andIhlanfeldt 1997).

Railroads have a positive effect on employment growth, and onoffice and housing construction, that eventually alters populationcomposition (Casson 2013; Levinson 2008a, b). Such impact varieswith location; for example, central cities see a rise in business com-plexes that increases the concentration of jobs, while suburbanareas experience an increase in housing complexes that spurs pop-ulation growth (Israel and Cohen−Blankshtain 2010; Levinson2008a, b). Commercial development increases land value, makingdowntown a very expensive place to live, and migrants select theperiphery or suburban areas. Under such conditions, passenger railsoffer fast, comfortable, dependable, and stress-free travel at peakoffice hours to the suburban populations of commuters who workin metropolitan downtowns (Pucher and Renne 2003).

A vast literature on passenger rails outside the United Statesshows the influence of railroads on local and regional growth (Chen2012; Knowles 2012; Kotavaara et al. 2011; Mejia-Doranteset al. 2012). Knowles (2012) shows that the railway has helpedCopenhagen’s economic growth by attracting substantial invest-ment in housing, retail, education, and leisure facilities, as wellas creating thousands of new jobs. Similarly, a study by Mejia-Dorantes et al. (2012) shows the economic impact of the Madridmetro line in Spain: it positively impacted the economic activityand changed the mix of business establishments in the Alcorconmunicipality, and it is associated with an increase in retail activitiesover time, which displaced manufacturing firms within the territory.In Finland, accessibility to transportation infrastructure, includingrailroads, influenced population change (Kotavaara et al. 2011).However, the relationships of transportation accessibility and pop-ulation change vary by geographic scale: at the regional level, trans-portation accessibility including railroads increases the (overall)population, while it has the opposite effect at the urban or locallevel.

In China, the development of intercity passenger trains positivelycontributes to regional economic growth by reducing travel timebetween cities for millions of commuters (Chen 2012). However,the benefits are not universal and equally distributed. Predictive andobservational studies show that large industrialized cities receivemore benefits than small and intermediate-size cities (Loukaitou-Sideris et al. 2013). Large cities see growth in employment, the realestate market, and tourism. These economic impacts of passengerrails eventually alter the compositions of employment and popula-tion at the local level as well as at the regional level. Passenger rails,with the help of revolutionary development in information technol-ogy (or a digital network), connect businesses of multiple urbanareas that contribute to polycentric urban growth, which is evolvedfrom but different than earlier assumptions of monocentric urbangrowth (Audirac 2005; Mejia-Dorantes et al. 2012).

Among the studies carried out on railroads in the United States,some are explicitly done on passenger rails, but others do not dis-tinguish between freight and passenger trains. Moreover, thesestudies are based on limited geographical areas, such as a city orregion. No study has been done at the national scale. The researchpresented in this article is likely the first to offer a systematic ex-ploration of the associations that intercity passenger rails have withpopulation and employment changes at the national scale. This re-search thus contributes to broadening the scholarly understandingof the relationships between passenger rails and population andemployment changes.

Methods

Data

For this study, the associations that intercity passenger rails havewith population and employment changes are examined at thecounty level for the continental United States. The intercity passen-ger rail data were obtained from the Intermodal Passenger Con-nectivity Database (IPCD) (Research and Innovative TechnologyAdministration 2012), a national-level database for the passengertransportation system. Although freight and passenger rails run onthe same tracks, the IPCD database provides information on pas-senger rail terminals that have intercity service facilities, whichmakes it possible to separate passenger from freight rail terminals.The extracted data are complete but only have information for inter-city passenger rail terminals and do not provide the quality andquantity of service (Fig. 1).

Data for county-level population and employment were ob-tained from the decennial censuses of 2000 and 2010 (Table 1)(US Census Bureau 2000, 2010). Counties are considered in thisresearch because they are important governmental units where so-cial and economic data are rich, easily available, and generally con-sistent over time (White 2008). The exception is that in 2001,Broomfield County, Colorado, was created from parts of Adams,Boulder, Jefferson, and Weld counties; the average proportion ofdemographic and socioeconomic data for these counties is usedto generate corresponding data for Broomfield County. Severalgovernment programs related to agriculture, social welfare, educa-tion, taxes, and transportation construction and maintenance operateat the county level. The classification of metro and nonmetrocounties was based on the 2000 definition of the US Office ofManagement and Budget (2003) and US Census Bureau (2013).

Population and employment growth are also influenced byland use and development (Chi and Ho 2018). In this research,the variable Land Developability Index (Chi and Ho 2013) capturesthis concept; it is controlled for, along with other socioeconomic

© ASCE 04018037-3 J. Urban Plann. Dev.

J. Urban Plann. Dev., 2018, 144(4): 04018037

Dow

nloa

ded

from

asc

elib

rary

.org

by

Gua

ngqi

ng C

hi o

n 09

/18/

18. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

variables. The Land Developability Index can be understood as thepotential for land development and conversion in a geographicalarea. It is calculated on the basis of geophysical characteristics(water, wetland, and slope), the amount of built-up lands (residen-tial, commercial, and industrial areas; transportation infrastruc-ture), cultural lands (e.g., Indian reservations), and federal and statelands.

Analytical Approach

The analysis starts with the standard regression method. Ordinaryleast squares (OLS) regression models are estimated to examine thegeneral associations of intercity passenger rails with population andemployment changes. Population change and employment changeare the dependent variables. Population change is expressed as thenatural log of the 2010 census population over the 2000 census

population (Table 2). Similarly, employment change is measuredby the natural log of the 2010 employment over the 2000 employ-ment (Table 2). The natural log helps to achieve better linearity withthe independent variables. The visual representations of popula-tion and employment change during the 2000s are shown in Figs. 2and 3, respectively.

The explanatory variable is the number of intercity passenger railterminals. The rationale behind choosing intercity rail terminals isthat there is a close association between the number of intercity pas-senger rail terminals and the volumes of population they serve. Ingeneral, both the size of the population and the number of intercitypassenger rail terminals are greater in metropolitan than in nonmet-ropolitan counties. The greater the number of terminals, the biggerthe population they serve. It would also be important to considerthe quantity and quality of the service that each terminal provides.Unfortunately, the authors do not have access to such information.

Table 1. Variable descriptions and data sources

Variables Descriptions Data sources

Demographic characteristicsPopulation change Natural log of the ratio of 2010 population over 2000 census population Decennial Censuses 2000 and 2010Employment change Natural log of the ratio of 2010 population (age ≥16) in labor force over

2000 population (age ≥16) in labor forceDecennial Censuses 2000 and 2010

Population density Number of persons per square miles in 2000 Decennial Census 2000Young Percentage young (age 15–19) in 2000 Decennial Census 2000Old Percentage old (age ≥65) in 2000 Decennial Census 2000Whites Percentage Whites in 2000 Decennial Census 2000Blacks Percentage Blacks in 2000 Decennial Census 2000Hispanics Percentage Hispanics in 2000 Decennial Census 2000Female-headed households Percentage female-headed households with own children under 18 years

old in 2000Decennial Census 2000

Bachelor’s degree Percentage of population (age ≥25) with bachelor’s degrees or higherin 2000

Decennial Census 2000

Intercity passenger rail Number of intercity passenger rail terminals Intermodal passenger connectivity databaseSocioeconomic conditions

Employment Percentage of population (age ≥16) in labor force in 2000 Decennial Census 2000Household income Median household income in 2000 Decennial Census 2000

Land development Land developability index Chi and Ho (2013)Metro Metropolitan county (1 = Yes, 0 = No) United States Census Bureau

Fig. 1. Passenger rail routes, United States.

© ASCE 04018037-4 J. Urban Plann. Dev.

J. Urban Plann. Dev., 2018, 144(4): 04018037

Dow

nloa

ded

from

asc

elib

rary

.org

by

Gua

ngqi

ng C

hi o

n 09

/18/

18. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

Therefore, this study utilizes the number of terminals per county,which is publicly available. Also, many intercity passenger rails ex-isted long before 2000, and their growth impacts might have fullymaterialized in earlier decades. Focusing on rails that opened shortlybefore 2000 would better capture passenger rail systems’ associa-tion with population and employment growth. However, the authorsdo not have access to such information. The number of passengerrail terminals varies across counties, and on average each county has

0.22 terminals (Table 2). The descriptive statistics are provided inTable 2.

The OLS models include 12 demographic and socioeconomiccontrol variables, including population density in 2000 (the numberof people per square mile); percentages of young (15–19 yearsof age) and old (65 years of age and above) in the population;percentages of non-Hispanic Whites, non-Hispanic Blacks, andHispanics in the population; percentage of households that are

Table 2. Descriptive statistics (N ¼ 3; 109)

Variables Median MeanStandarddeviation

Percentile(10%)

Percentile(90%)

Dependent variablesPopulation change (ln) 0.03 0.04 0.12 −0.08 0.19Employment change (ln) 0.04 0.05 0.13 −0.10 0.21

Independent variableNumber of intercity passenger rail terminals 0.00 0.22 0.87 0.00 1.00

Control variablesPopulation density 43.25 245.75 1,681.36 4.64 343.72Young 14.97 15.08 1.81 13.05 17.19Old 14.40 14.81 4.11 10.00 20.20Whites 89.30 81.62 18.69 54.20 97.80Blacks 2.10 9.14 14.65 0.20 31.20Hispanics 1.80 6.21 12.05 0.60 16.00Female-headed households 5.80 6.26 2.34 3.90 9.20Bachelor’s degree 14.50 16.51 7.80 9.30 26.70Employment 61.70 60.94 7.04 51.60 69.30Household income 40,597 42,043.86 9,821 31,746 53,676Land developability 79.53 70.75 26.56 27.33 96.99Metro 1.00 0.67 0.47 0.00 1.00

Fig. 2. Population change from 2000 to 2010 at county level in United States. Population change is measured as natural log of 2010 census populationover 2000 census population.

© ASCE 04018037-5 J. Urban Plann. Dev.

J. Urban Plann. Dev., 2018, 144(4): 04018037

Dow

nloa

ded

from

asc

elib

rary

.org

by

Gua

ngqi

ng C

hi o

n 09

/18/

18. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

female-headed with children under 18; percentage of the popula-tion with bachelor’s degrees or higher; percentage of the populationthat is employed; median household income; Land DevelopabilityIndex; and the metropolitan or nonmetropolitan status of the county(Table 1).

Spatial regression models are used to control for spatial depend-ence in the OLS model residuals. Spatial lag and spatial error arethe two most common forms of spatial dependence (Chi and Zhu2008). For this study, spatial lag dependence occurs when popula-tion and employment changes in a county are affected by changesin population and employment in its neighboring counties; spatialerror dependence refers to situations where model residuals arespatially correlated. To address spatial effects, a neighborhoodweight matrix is needed. Four different spatial weight matriceswere established for each model. Rook and queen contiguityweight matrices with Orders 1 and 2 were created and tested. Thishelped with the comparison of results of different weight matricesand selecting the most appropriate one. The weight matrix that isthe most appropriate is the one that produces a high value of spatialautocorrelation along with a high level of statistical significance(Chi 2010).

In this study, three spatial regression models were used: a spatiallag model, a spatial error model, and a spatial error model with lagdependence (SEMLD). The assessment of the three different spa-tial regression models was based on the values of log likelihood,Akaike’s information criterion (AIC), and Schwartz’s Bayesian in-formation criterion (BIC). The appropriate model has the highestlog likelihood value and the lowest AIC and BIC values (Chi andZhu 2008).

It should be noted that in the spatial lag model and the SEMLD,the spatial lag term (i.e., the coefficient that the spatially lagged

dependent variable has on the dependent variable) captures thespread effect of the growth from the neighboring counties. Thishelps separate the direct association that rails have with growth fromthe indirect association via the spread effect.

Results

Exploratory Spatial Data Analysis

The spatial dependence of population change and employmentchange can be measured by Moran’s I statistic (Moran 1948). Apositive value of spatial dependence indicates that counties withhigh (or low) values of a certain attribute are surrounded by countieswith high (or low) values, and a negative spatial dependence sug-gests that counties with high (or low) values of a certain attribute aresurrounded by counties with low (or high) values. Moran’s I forpopulation change and employment change are both relatively high,at 0.46 and 0.41, respectively (Figs. 4 and 5), based on the first-orderqueen contiguity weight matrix.

The spatial dependence of population change and employmentchange is further illustrated by the local indicators of spatial asso-ciation (LISA) at the county level (Anselin 1995). Figs. 6 and 7display the spatial dependence of population change and employ-ment change, respectively, based on the LISA statistic and the first-order queen contiguity weight matrix. They are in four categoriesby the combinations of high-high (i.e., high-growth countiessurrounded by high-growth counties), low-low (i.e., low-growthcounties surrounded by low-growth counties), low-high (i.e., low-growth counties surrounded by high-growth counties), and high-low(i.e., high-growth counties surrounded by low-growth counties),

Fig. 3. Employment change from 2000 to 2010 at county level in United States. Employment change is measured as natural log of 2010 employmentover 2000 employment.

© ASCE 04018037-6 J. Urban Plann. Dev.

J. Urban Plann. Dev., 2018, 144(4): 04018037

Dow

nloa

ded

from

asc

elib

rary

.org

by

Gua

ngqi

ng C

hi o

n 09

/18/

18. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

showing only those counties where the local Moran’s I statistic isstatistically significant at the 0.05 level based on a randomizationprocedure. High-high population-growth counties are mainly inthe Wyoming–Utah–Arizona region, northeastern Florida, and met-ropolitan areas of the District of Columbia, Atlanta, and Dallas-Houston-San Antonio. Low-low population-growth counties are

mainly in the two Dakotas, Nebraska, Kansas, and the MississippiDelta region. The local spatial dependence of employment changeexhibits a different pattern: the high-high employment-growthcounties are primarily in the southwestern states and lower Texas,and the low-low employment-growth counties are primarily in theentire lower Michigan area, the Appalachian region, and parts ofMississippi, Louisiana, Arkansas, and Tennessee.

Spatial Regression Models

To examine the associations that passenger rails have with popu-lation and employment changes, first OLS regression models arefitted. The statistically significant Moran’s I indicates the existenceof significant spatial dependence in the residuals of the OLS regres-sion models (the second columns of Tables 3 and 4). This suggeststhe violation of the OLS independence assumption. From a meth-odological perspective, the issue of spatial dependence needs tobe addressed, as statistical inference without the considerationof spatial dependence, if it exists, may lead to unreliable conclu-sions (Chi 2010; Chi and Zhu 2008). Therefore, three spatial re-gression models—a spatial lag model, a spatial error model, anda SEMLD—were fitted for population change and employmentchange. The SEMLD is the best for interpreting the regression co-efficients of population change because the value of the log like-lihood is the highest and the values of AIC and BIC are the lowest(Table 3). The SEMLD results show a significant association ofintercity passenger rails with population change, and the effectis negative. It indicates that intercity passenger rails help populationoutflow at the county level. For every additional intercity passengerrail terminal, a county experienced a 0.3% population decline inthe 2000s.

Similarly, results from Table 4 indicate that the SEMLD is thebest model to interpret the regression coefficients of employmentchange because the value of the log likelihood is the highest and thevalues of AIC and BIC are the lowest. The SEMLD results suggestthat intercity passenger rails have a negative association with em-ployment change; intercity passenger rails play a role in taking em-ployed people out of the county. The results show that for everyadditional intercity passenger rail terminal, a county experienceda 0.4% employment decline in the 2000s.

In response to the Great Recession during the studied period,people might have moved out of their county in search of jobs(Rickman and Guettabi 2015). During this period, employmentgrowth was weak and uneven, and population growth and housingmarket bubbles had occurred but then burst. Areas that were de-pendent on the construction industry and that had high shares ofemployment in retail and food service were especially hard hit dur-ing the recession (Gabe and Florida 2013; US Bureau of LaborStatistics 2012). The recession period also saw high levels of masslayoffs (US Bureau of Labor Statistics 2012). Employers were in-volved in thousands of mass layoff actions, forcing workers toleave industries. It is probable that intercity passenger rails, as anadditional means of transportation, served in the movement of thosepeople who were looking for jobs in other places.

These results are similar to those of another railroad study con-ducted at the county level. White (2008) found a negative impact ofrailroads on the early-twentieth-century population change in theGreat Plains region. That research shows that railroads helpedmove people out of more densely populated counties and broughtpeople into counties with lower population densities. Hence, rail-roads served both roles—they helped in population growth and inpopulation decline.

In the SEMLD, both spatial lag and spatial error effects arestatistically significant. The spatial lag effects come from the

Fig. 4. Moran’s I scatterplot of population change, 2000–2010. Thefirst-order Queen’s contiguity weight matrix is used.

Fig. 5. Moran’s I scatterplot of employment change, 2000–2010. Thefirst-order Queen’s contiguity weight matrix is used.

© ASCE 04018037-7 J. Urban Plann. Dev.

J. Urban Plann. Dev., 2018, 144(4): 04018037

Dow

nloa

ded

from

asc

elib

rary

.org

by

Gua

ngqi

ng C

hi o

n 09

/18/

18. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

Fig. 6. Local indicators of spatial association of population change from 2000 to 2010 at county level in United States.

Fig. 7. Local indicators of spatial association of employment change from 2000 to 2010 at county level in United States.

© ASCE 04018037-8 J. Urban Plann. Dev.

J. Urban Plann. Dev., 2018, 144(4): 04018037

Dow

nloa

ded

from

asc

elib

rary

.org

by

Gua

ngqi

ng C

hi o

n 09

/18/

18. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

population and employment changes that occurred in the neighbor-ing counties. The population of each county grows by 1.027% foreach percentage point of weighted population growth in its neigh-boring counties (Table 3). Similarly, the number of workers in eachcounty grows by 1.081% for each percentage point of weightedemployment growth in its neighboring counties (Table 4). In otherwords, each county will observe 10.27% population growth and10.81% employment growth if adjacent counties gain 10% in pop-ulation and employment growths, respectively. These increases donot come from within the county; rather, they are the effect of a“gift” from the county’s neighbors. This phenomenon is consistentwith the spread effect of growth pole theory, i.e., that populationand employment increases in one area help to spur populationand employment growth in nearby areas. In this context, the spatiallag effect is an indirect effect that intercity passenger rails have onpopulation and employment growth. But this effect is not entirelybecause of intercity passenger rails; other modes of transportationcould have contributed to the process of population and employ-ment growth.

Intercity passenger rails can be best seen as a facilitator of pop-ulation and employment flows. A county with a strong economycan attract residents from other counties, and a county with a weakeconomy cannot maintain its population base. Thus, intercity pas-senger rails act as a facilitator in the process of population andemployment redistribution.

Discussion and Conclusions

Summary and Discussion

Many studies have been conducted to examine the relationshipsbetween transportation infrastructure and changes in populationand employment (Chi et al. 2006). Most research on the demo-graphic and economic impacts of rails focuses on freight rails ratherthan passenger rails. At a time when the federal government is con-sidering rebuilding and modernizing the transportation infrastruc-ture, including intercity passenger rails within the United States,this study contributes to the literature by examining the associa-tions that intercity passenger rails—an understudied transportationmode—have with population and employment changes.

This study analyzes both direct and indirect associations thatintercity passenger rails have with population and employmentchanges (growth or decline) in a systematic way by applying anOLS regression model, spatial lag model, spatial error model, andSEMLD in sequence. The systematic application of these models isone strength of this study. This approach identifies and addressesthe weaknesses of the models involved in the analysis. Likewise,the application of these models supports the identification of themost appropriate model to provide a better understanding of theassociations of passenger rails with changes in population and em-ployment. Based on the higher value of log likelihood and lower

Table 3. Regressions of intercity rail terminals on population change from 2000 to 2010

Variable OLS SLM SEM SEMLD

Explanatory variableIntercity rail terminals −0.001 (0.002) −0.003 (0.002) −1.19 × 10−4 (0.002) −0.003a (0.001)

Control variablePopulation density −4.74 × 10−6b (1.09 × 10−6) −3.31 × 10−6b (9.24 × 10−7) −5.29 × 10−7 (1.25 × 10−6) −1.23 × 10−6a (5.37 × 10−7)Young −0.009b (0.002) −0.004c (0.001) −0.002 (0.001) 6.97 × 10−4 (9.51 × 10−4)Old −0.012b (6.20 × 10−4) −0.008b (5.41 × 10−4) −0.008b (6.52 × 10−4) −0.003b (3.54 × 10−4)Whites 3.99 × 10−4 (3.15 × 10−4) 4.11 × 10−4 (2.69 × 10−4) 9.19 × 10−4c (3.35 × 10−4) 1.64 × 10−4 (1.68 × 10−4)Blacks −8.23 × 10−6 (2.84 × 10−4) −5.59 × 10−4a (2.42 × 10−4) −0.001b (3.19 × 10−4) −4.58 × 10−4c (1.53 × 10−4)Hispanics 7.84 × 10−4a (3.23 × 10−4) 3.67 × 10−4 (2.76 × 10−4) 9.67 × 10−c (3.65 × 10−4) −7.94 × 10−5 (1.70 × 10−4)Female-headedhouseholds

−0.003a (0.002) 3.68 × 10−4 (0.001) 0.006b (0.001) 8.36 × 10−4 (9.26 × 10−4)

Bachelor’s degree 6.56 × 10−4a (3.33 × 10−4) 0.001b (2.84 × 10−4) 0.001b (3.04 × 10−4) 0.001b (1.96 × 10−4)Employment −0.001b (3.82 × 10−4) −0.001b (3.25 × 10−4) −7.28 × 10−4 (3.81 × 10−4) −6.17 × 10−4c (2.12 × 10−4)Household income 3.82 × 10−6b (3.70 × 10−7) 2.29 × 10−6b (3.23 × 10−7) 4.65 × 10−6b (4.09 × 10−7) 2.53 × 10−7 (2.00 × 10−7)Land developability −2.24 × 10−4c (7.65 × 10−5) 1.37 × 10−4a (6.53 × 10−5) 2.11 × 10−4a (9.99 × 10−5) 2.53 × 10−4b (3.83 × 10−5)Metro −0.004 (0.004) −0.002 (0.003) −1.17 × 10−4 (0.003) −7.06 × 10−4 (0.003)Constant 0.223b (0.046) 0.098a (0.039) −0.081 (0.044) 0.014 (0.026)Spatial lag effects — 0.559b (0.018) — 1.027b (0.012)Spatial error effects — — 0.683b (0.017) −0.828b (0.028)

Diagnostic testMoran’s I (error) 0.379b 0.0293b −0.048b −0.023bLagrange multiplier(lag)

1,011.03b — — —

Robust LM (lag) 2.32 — — —Lag multiplier (error) 1,241.85b — — —Robust LM (error) 2,33.14b — — —

Measure of fitLog likelihood 2,872.01 3,263.72 3,375.63 3,566.46AIC −5,716.02 −6,497.43 −6,723.25 −7,102.92BIC −5,631.43 −6,406.80 −6,638.67 −7,012.29

Spatial weight matrix 1st-order queen 1st-order queen 1st-order queen 1st-order rook

Note: AIC = Akaike’s information criterion; BIC = Schwartz’s Bayesian information criterion; OLS = ordinary least squares; SEM = spatial error model;SEMLD = spatial error model with lag dependence; and SLM = spatial lag model.aSignificant at p ≤ 0.05 for a two-tail test.bSignificant at p ≤ 0.001 for a two-tail test.cSignificant at p ≤ 0.01 for a two-tail test; standard errors in parentheses.

© ASCE 04018037-9 J. Urban Plann. Dev.

J. Urban Plann. Dev., 2018, 144(4): 04018037

Dow

nloa

ded

from

asc

elib

rary

.org

by

Gua

ngqi

ng C

hi o

n 09

/18/

18. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

values of AIC and BIC, the SEMLD stands out as the best fit. Inaddition, the simultaneous application of the spatial lag and spatialerror models helps to identify indirect effects of intercity passengerrails and the potential effects of variables that are not included inthe model.

The results of this study show direct as well as indirect associ-ations that intercity passenger rails have with changes in populationand employment. The major contribution of this study lies in theseparation of the direct and indirect associations through a spatiallag term capturing spread effects from neighboring counties. Inter-city passenger rails exerted direct associations with changes inpopulation and employment at the county level in the 2000s, evenafter controlling for 14 socioeconomic variables. The associationswere strong enough to influence changes in population and em-ployment independently. However, these associations were nega-tive, suggesting that intercity passenger rails helped in populationand employment outflows. The results indicate that the declines inpopulation and employment occurred by 0.3% and 0.4%, respec-tively, for each additional intercity passenger rail terminal. Intercitypassenger rails act as a facilitator of population and employmentoutflows in a weak economy.

This research also indicates the indirect associations that inter-city passenger rails have with population and employment changes.The indirect associations are measured by the spatial lag effect ofthe SEMLD. Population and employment in counties are spatiallyconnected. Changes in one county affect the neighboring counties.Neighboring counties are connected not only geographically but

also socially and economically. Intercity passenger rails, along withother modes of transportation, play an important role in facilitatingthe spread of changes and in the integration of small communitiesinto a larger regional economy. These findings support the spreadeffect of growth pole theory, which suggests that population andemployment increases in one area spur population and employmentincreases in adjacent areas. This has important implications forurban planning and transportation planning. Demographic and eco-nomic changes in nearby areas affect neighboring areas; therefore,planners for one area should also pay close attention to their sur-rounding areas during the planning process.

Another possible explanation for the negative direct and positiveindirect associations that passenger rails have on population andemployment changes could be related to the “straw” effects. Thestraw effects of transportation infrastructure lower economic pro-ductivity in lagging areas because of the increased local dependencyon the vibrant surrounding areas (Kim and Han 2016). Intercity pas-senger rails do not contribute to making a county an attractive place;rather, they help move out local population and employment. Thevariable of land developability is positive and significant for bothpopulation and employment changes, indicating that counties withhigher potential for land development have magnetic powers to at-tract additional population and employment.

In this context, intercity passenger rails can be viewed as achange agent that is neither a boom factor nor a bust factor; rather,their role is determined by the national socioeconomic context.This finding is consistent with the results of research on other

Table 4. Regressions of intercity rail terminals on employment change from 2000 to 2010

Variable OLS SLM SEM SEMLD

Explanatory variableIntercity rail terminals −0.004 (0.003) −0.004 (0.002) −2.56 × 10−4 (0.002) −0.004a (0.002)

Control variablePopulation density −4.65 × 10−6b (1.35 × 10−6) −3.08 × 10−6c (1.18 × 10−6) −1.43 × 10−6 (1.61 × 10−6) −1.23 × 10−6 (8.82 × 10−7)Young 0.005a (0.002) 0.002 (0.002) 0.002 (0.002) 0.001 (0.001)Old −0.004b (7.64 × 10−4) −0.004b (6.67 × 10−4) −0.004b (8.26 × 10−4) −0.004b (4.68 × 10−4)Whites −0.002b (3.90 × 10−4) −0.001b (3.40 × 10−4) −8.29 × 10−4a (4.21 × 10−4) −2.45 × 10−5 (2.55 × 10−4)Blacks −0.002b (3.53 × 10−4) −0.001b (3.10 × 10−4) −0.002b (4.06 × 10−4) −4.74 × 10−4a (2.27 × 10−4)Hispanics 0.001c (4.02 × 10−4) 3.52 × 10−4 (3.52 × 10−4) 0.001c (4.63 × 10−4) −3.43 × 10−4 (2.49 × 10−4)Female-headedhouseholds

−0.012b (0.002) −0.005c (0.002) −1.71 × 10−4 (0.002) 5.17 × 10−4 (0.001)

Bachelor’s degree 0.002b (4.06 × 10−4) 0.001b (3.54 × 10−4) 0.001b (3.88 × 10−4) 5.68 × 10−4a (2.82 × 10−4)Household income 1.47 × 10−6b (4.15 × 10−7) 9.83 × 10−7c (3.65 × 10−7) 3.10 × 10−6b (4.95 × 10−7) 2.44 × 10−7 (2.46 × 10−7)Land developability −3.38 × 10−4b (8.89 × 10−5) −1.85 × 10−5 (7.76 × 10−5) 8.17 × 10−5 (1.23 × 10−4) 1.66 × 10−4b (4.97 × 10−5)Metro −2.23 × 10−4 (0.005) 1.56 × 10−4 (0.004) −8.93 × 10−5 (0.004) −7.46 × 10−5 (0.004)Constant 0.293b (0.055) 0.140c (0.048) 0.044 (0.056) 0.019 (0.036)Spatial lag effects — 0.551b (0.020) — 1.081b (0.018)Spatial error effects — — 0.620b (0.019) −0.973b (0.039)

Diagnostic testMoran’s I (error) 0.32b 0.001 −0.03b −0.007Lagrange multiplier(lag)

838.45b — — —

Robust LM (lag) 6.21a — — —Lag multiplier (error) 885.49b — — —Robust LM (error) 53.25b — — —

Measure of fitLog likelihood 2,192.64 2,514.85 2,559.95 2,601.20AIC −4,359.29 −5,001.71 −5,093.90 −5,174.4BIC −4,280.74 −4,917.12 −5,015.35 −5,089.81

Spatial weight matrix 1st-order queen 1st-order queen 1st-order queen 2nd-order rook

Note: AIC = Akaike’s information criterion; BIC = Schwartz’s Bayesian information criterion; OLS = ordinary least squares; SEM = spatial error model;SEMLD = spatial error model with lag dependence; and SLM = spatial lag model.aSignificant at p ≤ 0.05 for a two-tail test.bSignificant at p ≤ 0.001 for a two-tail test.cSignificant at p ≤ 0.01 for a two-tail test; standard errors in parentheses.

© ASCE 04018037-10 J. Urban Plann. Dev.

J. Urban Plann. Dev., 2018, 144(4): 04018037

Dow

nloa

ded

from

asc

elib

rary

.org

by

Gua

ngqi

ng C

hi o

n 09

/18/

18. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

modes of transportation such as railroads and highways (Chi 2010;Levinson 2008b). Even though transportation infrastructure previ-ously was considered a growth factor, recent research studies haveshown mixed results, indicating that transportation infrastructurecould be viewed as a facilitator of change. Loukaitou-Sideris et al.(2013) also find similar impacts of passenger rails on urban eco-nomic growth. According to their study, passenger rails are more adistributive than a generative force. To realize positive economicgrowth, other factors such as the magnitude of public capital invest-ment and the quality of urban planning play vibrant roles. Duringthe Great Recession period, when employment growth was weakand uneven and employers were conducting mass layoffs, intercitypassenger rails might have helped job-seeking populations moveout of their origin counties. The findings of this study are importantduring this time when debate about whether intercity passengerrails should be rebuilt, upgraded, and expanded is growing.

Future Research

This study could be extended in several directions. First, the asso-ciation between intercity passenger rails and growth could be com-pared across different geographical areas, such as urban, suburban,and rural areas. This can be done through a spatial regime modelthat deals with spatial heterogeneity, allowing comparison of thedirect and indirect associations that intercity passenger rails havewith population change across urban, suburban, and rural areas.Second, future research could analyze the associations between in-tercity passenger rails and changes in population and employmentfor other decades, such as the 1980s and 1990s, as well as for thewhole period of 1980–2010. That would provide an understandingof the rails–growth relationship over a long time period. Third, fu-ture research could address the issue of intermodality. In this study,intercity passenger rails are considered in isolation from othermodes of transportation. In future research, the impact of intercitypassenger rails could be examined while controlling for the impactsof highways and airways on changes in population and employ-ment. Fourth, future research could consider the possible impactsof intercity passenger rails on social inequality as measured by ed-ucation, income, and race and ethnicity. Fifth, the causality frompopulation and employment growth to passenger rail (terminal) de-velopment, as well as the endogeneity between growth and raildevelopment, should be carefully addressed.

Acknowledgments

The authors thank Robert Boyd, Mary Emery, Jeffrey Jacquet,Meredith Redlin, and Songxin Tan for providing comments onearlier drafts of this article. This research was supported in part bythe National Science Foundation (Award 1541136), the US DOT(Awards DTRT12GUTC14-201307 and DTRT12GUTC14-201308),and the Eunice Kennedy Shriver National Institute of Child Healthand Human Development (Award P2C HD041025).

References

Anselin, L. 1995. “Local indicators of spatial autocorrelation—LISA.”Geog. Anal. 27 (2): 93–115. https://doi.org/10.1111/j.1538-4632.1995.tb00338.x.

Atack, J., and R. A. Margo. 2011. “The impact of access to rail transpor-tation on agricultural improvement: The American Midwest as a testcase, 1850–1860.” J. Transp. Land Use 4 (2): 5–18. https://doi.org/10.5198/jtlu.v4i2.188.

Audirac, I. 2005. “Information technology and urban form: Challenges tosmart growth.” Int. Reg. Sci. Rev. 28 (2): 119–145. https://doi.org/10.1177%2F0160017604273624.

Boarnet, M. G., and A. F. Haughwout. 2000. Do highways matter? Evi-dence and policy implications of highways’ influence on metropolitandevelopment. Washington, DC: Brookings Institute.

Bollinger, C. R., and K. R. Ihlanfeldt. 1997. “The impact of rapid rail transiton economic development: The case of Atlanta’s MARTA.” J. UrbanEcon. 42 (2): 179–204. https://doi.org/10.1006/juec.1996.2020.

Briggs, R. 1981. “Interstate highway system and development in nonmet-ropolitan areas.” Transp. Res. Rec. 812: 9–12.

Brown, R. A. 2009. “The 2009 economic landscape: How the recession isunfolding across four US regions.” FDIC Q. 3 (1): 26–38.

Casson, M. 2013. “The determinants of local population growth: A study ofOxfordshire in the nineteenth century.” Explorations Econ. Hist. 50 (1):28–45. https://doi.org/10.1016/j.eeh.2012.09.003.

Chen, C. L. 2012. “Reshaping Chinese space-economy through high-speedtrains: Opportunities and challenges.” J. Transp. Geogr. 22 (May):312–316. https://doi.org/10.1016/j.jtrangeo.2012.01.028.

Chi, G. 2010. “The impacts of highway expansion on population change:An integrated spatial approach.” Rural Sociol. 75 (1): 58–89. https://doi.org/10.1111/j.1549-0831.2009.00003.x.

Chi, G., and H. Ho. 2013. “Land developability: A measure of the propor-tion of lands available for development and conversion.” Accessed May18, 2017. http://www.landdevelopability.org.

Chi, G., and H. Ho. 2018. “Population stress: A spatiotemporal analysis ofpopulation change and land development at the county level in the con-tiguous United States, 2001–2011.” Land Use Policy 70 (1): 128–137.https://doi.org/10.1016/j.landusepol.2017.10.008.

Chi, G., and H. C. Ho 2013. “Land developability: A measure of the pro-portion of lands available for development and conversion.” AccessedSeptember 11, 2018. http://www.landdevelopability.org.

Chi, G., P. R. Voss, and S. C. Deller. 2006. “Rethinking highway effectson population change.” Public Works Manage. Policy 11 (1): 18–32.https://doi.org/10.1177/1087724X06292336.

Chi, G., and J. Zhu. 2008. “Spatial regression models for demographicanalysis.” Popul. Res. Policy Rev. 27 (1): 17–42. https://doi.org/10.1007/s11113-007-9051-8.

Coffin, A. W. 2007. “From roadkill to road ecology: A review of the eco-logical effects of roads.” J. Transp. Geogr. 15 (5): 396–406. https://doi.org/10.1016/j.jtrangeo.2006.11.006.

Darwent, D. F. 1969. “Growth poles and growth centers in regionalplanning—A review.” Environ. Plann. 1 (1): 5–31. https://doi.org/10.1068/a010005.

Duranton, G., and M. A. Turner. 2012. “Urban growth and transportation.”Rev. Econ. Stud. 79 (4): 1407–1440. https://doi.org/10.1093/restud/rds010.

Fogel, R. W. 1962. “A quantitative approach to the study of rail-roads in American economic growth: A report of some preliminaryfindings.” J. Econ. Hist. 22 (2): 163–197. https://doi.org/10.1017/S0022050700062719.

Friedberg, M. 2018. “Donald Trump’s plan can’t fix America’s infra-structure without focusing on our railways.” NBC News. AccessedJune 20, 2018. https://www.nbcnews.com/think/opinion/donald-trump-s-plan-can-t-fix-america-s-infrastructure-ncna846496.

Gabe, T. M., and R. Florida. 2013. “Effects of the housing boom and buston US metro employment.” Growth Change 44 (3): 391–414. https://doi.org/10.1111/grow.12013.

Gama, A. 2017. “Own and cross-price elasticities of demand for domesticflights and intercity trains in the US.” Transp. Res. Part D Transp.Environ. 54: 360–371. https://doi.org/10.1016/j.trd.2017.06.010.

Goetz, A. R. 2012. “Guest editorial: Introduction to the special section onrail transit systems and high speed rail.” J. Transp. Geogr. 22 (May):219–220. https://doi.org/10.1016/j.jtrangeo.2012.02.004.

Grunwald, M. 2010. “Can high-speed rail get on track?” Accessed July 19,2010. http://content.time.com/time/magazine/article/0,9171,2002523,00.html.

Hedges, J. B. 1926. “The colonization work of the Northern PacificRailroad.” Mississippi Valley Historical Rev. 13 (3): 311–342. https://doi.org/10.2307/1893110.

© ASCE 04018037-11 J. Urban Plann. Dev.

J. Urban Plann. Dev., 2018, 144(4): 04018037

Dow

nloa

ded

from

asc

elib

rary

.org

by

Gua

ngqi

ng C

hi o

n 09

/18/

18. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

Henry, M. S., D. L. Barkley, and S. Bao. 1997. “The hinterland’s stake inmetropolitan growth: Evidence from selected southern regions.” J. Reg.Sci. 37 (3): 479–501. https://doi.org/10.1111/0022-4146.00065.

Hertz, T., L. D. Kusmin, A. Marré, and T. Parker. 2014. “Rural employmenttrends in recession and recovery.” Accessed February 26, 2015.http://www2.econ.iastate.edu/classes/crp274/swenson/CRP566/Readings/ERS%20Rural%20Recovery.pdf.

Housing Assistance Council. 2012. Social, economic and housing condi-tions in rural America. Washington, DC: Housing Assistance Council.

Hurst, K. L. 2014. “Railroads, passenger.” In Encyclopedia of transporta-tion: Social science and policy, edited by M. Garrett. Thousand Oaks,CA: SAGE Publications.

Israel, E., and G. Cohen-Blankshtain. 2010. “Testing the decentralizationeffects of rail systems: Empirical findings from Israel.” Transp. Res.Part A 44 (7): 523–536. https://doi.org/10.1016/j.tra.2010.03.021.

Itzkoff, D. M. 1985. Off the track: The decline of the intercity passengertrain in the United States. Praeger Pub Text. London: Praeger.

Jenks, L. H. 1944. “Railroads as an economic force in Americandevelopment.” J. Econ. Hist. 4 (1): 1–20. https://doi.org/10.1017/S002205070008400X.

Kim, J. Y., and J. H. Han. 2016. “Straw effects of new highway constructionon local population and employment growth.” Habitat Int. 53 (Apr):123–132. https://doi.org/10.1016/j.habitatint.2015.11.009.

Kirby, R. S. 1983. “Nineteenth–century patterns of railroad development onthe Great Plains.” Great Plains Q. 3 (3): 157–170.

Knowles, R. D. 2012. “Transit oriented development in Copenhagen,Denmark: From the Finger Plan to Ørestad.” J. Transp. Geogr.22 (May): 251–261. https://doi.org/10.1016/j.jtrangeo.2012.01.009.

Kotavaara, O., H. Antikainen, and J. Rusanen. 2011. “Population changeand accessibility by road and rail networks: GIS and statistical approachto Finland 1970-2007.” J. Transp. Geogr. 19 (4): 926–935. https://doi.org/10.1016/j.jtrangeo.2010.10.013.

Levinson, D. 2008a. “Density and dispersion: The co-development of landuse and rail in London.” J. Econ. Geogr. 8 (1): 55–77. https://doi.org/10.1093/jeg/lbm038.

Levinson, D. 2008b. “The orderliness hypothesis: The correlation ofrail and housing development in London.” J. Transp. Hist. 29 (1):98–114. https://doi.org/10.7227/TJTH.29.1.8.

Levinson, D. 2012. “Accessibility impacts of high-speed rail.” J. Transp.Geogr. 22 (May): 288–291. https://doi.org/10.1016/j.jtrangeo.2012.01.029.

Lichter, D. T., and G. V. Fuguitt. 1980. “Demographic response to trans-portation innovation: The case of the interstate highway.” Social Forces59 (2): 492–512. https://doi.org/10.1093/sf/59.2.492.

Loukaitou-Sideris, A., H. Higgins, M. Piven, and W. Wei. 2013. “Tracks tochange or mixed signals? A review of the Anglo-Saxon literature on theeconomic and spatial impacts of high-speed rail.” Transp. Rev. 33 (6):617–633. https://doi.org/10.1080/01441647.2013.836578.

Mark, H., and K. P. Schwirian. 1967. “Ecological position, urban centralplace function, and community population growth.” Am. J. Sociology73 (1): 30–41. https://doi.org/10.1086/224434.

Mejia-Dorantes, L., A. Paez, and J. M. Vassallo. 2012. “Transportation in-frastructure impacts on firm location: The effect of a new metro line inthe suburbs of Madrid.” J. Transp. Geogr. 22 (May): 236–250. https://doi.org/10.1016/j.jtrangeo.2011.09.006.

Moore, C. T., M. L. Mayer, H. A. Lipson, and G. Joyce. 1964.“A study of the expected economic and social impact of interstatehighways in the industrial and commercial trading area of Birmingham,Alabama—The first phase.” Dept. of Marketing, Univ. of Alabama.Accessed November 19, 2015. http://www.econbiz.de/Record/study-expected-economic-social-impact-interstate-highways-industrial-commercial-trading-area-birmingham-alabama-phase-charles-thomas-moore/10003247080.

Moran, P. 1948. “The interpolation of statistical maps.” J. R. Stat. Soc. B10 (2): 243–251.

Okada, H. 1994. “Features and economic and social effects of theShinkansen.” Jpn. Railway Transp. Rev. 3: 9–16.

Ory, D. T., and P. L. Mokhtarian. 2006. “Which came first, the telecom-muting or the residential relocation? An empirical analysis of causality.”

Urban Geogr. 27 (7): 590–609. https://doi.org/10.2747/0272-3638.27.7.590.

Ozbay, K., D. Ozmen, and J. Berechman. 2006. “Modeling and analysis ofthe link between accessibility and employment growth.” J. Transp. Eng.132 (5): 385–393. https://doi.org/10.1061/(ASCE)0733-947X(2006)132:5(385).

Peterman, D. R., J. Frittelli, and W. J. Mallett. 2009. High speed rail (HSR)in the United States. Washington, DC: Congressional Research Service.

Pucher, J., and J. L. Renne. 2003. “Socioeconomics of urban travel:Evidence from the 2001 NHTS.” Transp. Q. 57 (3): 49–77.

Puentes, R., A. Kane, and J. Kane. 2013. A new alignment: StrengtheningAmerica’s commitment to passenger rail. Washington, DC: BrookingsInstitution.

Research and Innovative Technology Administration. 2012. “Intermodalpassenger connectivity database.” Bureau of Transportation Statistics.Accessed January 27, 2014. http://www.transtats.bts.gov/DatabaseInfo.asp?DB_ID=640&Link=0.

Rickman, D. S., and M. Guettabi. 2015. “The great recession and nonmet-ropolitan America.” J. Reg. Sci. 55 (1): 93–112. https://doi.org/10.1111/jors.12140.

Ryder, A. 2012. “High speed rail.” J. Transp. Geogr. 22 (May): 303–305.https://doi.org/10.1016/j.jtrangeo.2012.03.004.

Schnore, L. F. 1957. “Metropolitan growth and decentralization.” Am. J.Sociology 63 (2): 171–180. https://doi.org/10.1086/222174.

Thiel, F. I. 1962. “Social effects of modern highway transportation.” High-way Res. Board Bull. 327: 1–20.

Thompson, C., and T. Bawden. 1992. “What are the potential economicdevelopment impacts of High–Speed Rail?” Econ. Dev. Q. 6 (3):297–319. https://doi.org/10.1177/089124249200600306.

Topalovic, P., J. Carter, M. Topalovic, and G. Krantzberg. 2012. “Lightrail transit in Hamilton: Health, environmental and economic impactanalysis.” Social Indic. Res. 108 (2): 329–350. https://doi.org/10.1007/s11205-012-0069-x.

US Bureau of Labor Statistics. 2012. “The Recession of 2007–2009.”US Bureau of Labor Statistics. Accessed February 13, 2015. http://www.bls.gov/spotlight/2012/recession/pdf/recession_bls_spotlight.pdf.

US Census Bureau. 2000. “Census decennial 2000.” Accessed March 18,2014. https://factfinder.census.gov/faces/nav/jsf/pages/index.xhtml.

US Census Bureau. 2010. “Census decennial 2010.” Accessed March 18,2014. https://factfinder.census.gov/faces/nav/jsf/pages/index.xhtml.

US Census Bureau. 2011. “Population distribution and change: 2000 to2010.” Accessed February 13, 2015. http://www.census.gov/prod/cen2010/briefs/c2010br-01.pdf.

US Census Bureau. 2013. “Cartographic boundary shapefiles-metropolitan and micropolitan statistical areas and related statisticalareas.” Accessed January 6, 2015. https://www.census.gov/geo/maps-data/data/cbf/cbf_msa.html.

US Office of Management and Budget. 2003. Metropolitan and micro-politan statistical areas. OMB Bulletin No. 03-04. Washington, DC:US Office of Management and Budget.

van den Heuvel, F. P., L. Rivera, K. H. van Donselaar, A. de Jong, Y. Sheffi,P. W. de Langen, and J. C. Fransoo. 2014. “Relationship between freightaccessibility and logistics employment in US counties.” Transp. Res.Part A Policy Pract. 59 (Jan): 91–105. https://doi.org/10.1016/j.tra.2013.11.002.

Voss, P. R., and G. Chi. 2006. “Highways and population change.” RuralSociology 71 (1): 33–58. https://doi.org/10.1526/003601106777789837.

White House. 2017. “President Trump’s plan to rebuild America’sinfrastructure.” White House. Accessed February 14, 2018. https://www.whitehouse.gov/articles/president-trumps-plan-rebuild-americas-infrastructure/.

White House. 2018. “Fact sheet: 2018 budget, infrastructure initiative.”White House. Accessed February 14, 2018. https://www.whitehouse.gov/sites/whitehouse.gov/files/omb/budget/fy2018/fact_sheets/2018%20Budget%20Fact%20Sheet_Infrastructure%20Initiative.pdf.

White, K. J. C. 2008. “Sending or receiving stations? The dual influenceof railroads in early 20th-century Great Plains settlement.” Popul.Res. Policy Rev. 27 (1): 89–115. https://doi.org/10.1007/s11113-007-9056-3.

© ASCE 04018037-12 J. Urban Plann. Dev.

J. Urban Plann. Dev., 2018, 144(4): 04018037

Dow

nloa

ded

from

asc

elib

rary

.org

by

Gua

ngqi

ng C

hi o

n 09

/18/

18. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.