ABSTRACT
Woodford County, open for six years (quoted from Staff 2002). Introduction For Kentucky, the 1990s marked the beginning of a new era in cultural and ethnic diversity. Between 1990-2000, the Hispanic population nearly tripled to become the second largest minority group in the state. This influx of migrants precipitated a gradual but widespread change in the cultural landscape of many counties as both urban and rural communities began to evolve towards a more diverse multi-ethnic character. Many of these changes stem from broader statewide migration trends. For example, between 1995-2000, ‘net gains from international migration’ to Kentucky exceeded net domestic migration gains and total net migration eclipsed natural increase (Price et al., 2004, p. 3). International migrants are primarily Asian or Hispanic and the majority of growth for each of these minority groups has occurred since 1995. By 2000, Kentucky ranked among the top ten states in the U.S. with rapidly growing Latino populations (U.S. Bureau of Census, 2000b) and between 1990-2000 it also had the 12th fastest growing nonmetro Hispanic population (Kandel and Cromartie, 2004). From a statewide perspective these are noteworthy trends, but equally important are the differential demographic changes occurring across smaller geographic areas of the state. Unlike all other states, the Latino population in Kentucky is geographically dispersed and generally lacks a significant level of concentration in any particular rural or urban locale. Grocery stores and restaurants owned by and catering to Hispanics have appeared almost overnight in rural communities and along welltraveled highways throughout the state, hinting at the social and economic importance of these newcomers to the Bluegrass state. Hispanic migrants to Kentucky are socially and economically diverse and represent a broad spectrum of
education and economic levels. Because of this diversity and the widespread settlement patterns of this group, this paper utilizes a three-tiered urban-rural typology to evaluate the Kentucky Hispanic population. This population is still quite small and dispersed so it was necessary to utilize the 2000, 5 per cent Public Use Microdata Sample (PUMS) because it provides greater demographic specificity for these small and dispersed populations. The PUMS files, however, aggregate data to areas of a minimum population of 100,000, which means that the analysis is based on groups of counties (PUMAs or Public Use Microdata Areas) covering larger geographic areas. For this analysis I categorized the PUMAs as Urban (all metropolitan (metro) counties), Rural (all counties are nonmetropolitan (nonmetro)), and Mixed Urban-Rural (both metro and nonmetro counties are represented) in order to compare the socio-demographic characteristics of Latinos based on their geographic location. The central premise of this chapter is that urban and rural areas of Kentucky likely attract different segments of the new and growing Latino population. This chapter focuses on the demographic characteristics and geographic distribution of the emerging Hispanic population in Kentucky. Following this introduction is a discussion of the increasing diffusion of the Hispanic population in the U.S. followed by a more specific overview of Latinos in Kentucky. The analysis section is divided into three parts. The first section cartographically examines changes in Latino settlement patterns in Kentucky at the county level using 2000 Census data. The demographic profiles in the second section are based on the three-tiered urban-rural categorization. Finally, multivariate analysis is used to evaluate the contributions of different county-level variables in explaining variation in Hispanic population distribution across the state. Major findings and conclusions are discussed in the last section. One of the most important findings is that the Hispanic population is dispersing away from urban areas and showing signs of residential stability in rural communities. The implications for Kentucky are significant: the attraction of migrants who are settling and remaining in place provides hope for invigorating local rural economies through population growth, entrepreneurial investment, and increased ethnic diversity. Growth and Diffusion of the Hispanic Population in the United States Geographic clustering generally characterized Hispanic settlement patterns in the U.S. until late in the 1990s. Even in 2000, with increased diffusion away from gateway cities, 44.7 per cent of Latinos lived in western states and 46.4 per cent overall lived in metro areas (Therrien and Ramirez, 2001). Macro-level assessments however mask a growing social and demographic diversity within this population as well as increasingly widespread distribution across the country. Previous research, mostly at an aggregate U.S. scale, suggests that Hispanic migration largely reinforces existing Latino concentrations. McHugh (1989) assessed the geographic distribution of Hispanics in the U.S. concluding, at that time, that they remained concentrated in nine gateway states although the level of
concentration varied across individual Hispanic sub-groups. Between 1985-1990, the general pattern of internal Hispanic migration continued to focus on key gateway cities; however, as early as the 1990 Census, a small amount of dispersal away from traditional settlement areas was occurring among U.S.-born Latinos (Frey and Liaw, 1999). In a cartographic depiction of net in-and out-migration for Latinos, three of the top five states gaining Latinos were located in the South, including Virginia, Georgia, and Florida (Frey and Liaw, 1999). Virginia and Florida received the most foreign-and U.S.-born Latinos while Georgia was a magnet for foreign-born Latinos. Southern states that lost U.S.-born Latinos included Louisiana and Mississippi while Alabama, Louisiana, and Mississippi lost foreign-born Latinos. These exchanges are noteworthy because they demonstrate that southern states, as early as 1990, were redistributing both U.S. and foreignborn Latinos, albeit in small numbers. The 1990s were marked by continued change as new migration patterns emerged and the internal distribution of Hispanics continued to evolve. Changes in the Immigration Reform and Control Act of 1986 (IRCA) initiated a shift in Mexican immigration away from the gateway states of Arizona, California, Illinois, New Mexico, and Texas (Durand et al., 2000). Nearly one-third of immigrants chose an alternative destination, as compared to only ten per cent prior to IRCA. The new legal status of many immigrants opened new labor markets and opportunities for migration. Interstate migration patterns also differed across subgroups of the Hispanic population with patterns strongly influenced by the dominant point of entry (Foulkes and Newbold, 2000). An evaluation of interstate migration patterns of Mexican, Cuban, and Puerto Rican Hispanic migrants revealed geographic concentrations of Cubans in Florida, and dispersion of Mexican and Puerto Rican Hispanics in the southwest and northeast respectively. The combination of the IRCA, with its focus on immigration control, coupled with a growing economy, created an environment in which recently legalized Hispanic immigrants could move freely across the U.S. By 2000, this dispersion away from key states was no longer limited to urban areas and indeed many rural areas were beginning to gain sizable Hispanic populations as well. Kandel and Cromartie (2004) note that although most of the Hispanic population resided in an urban place in 2000, Latinos ‘accounted for… 25 per cent of nonmetro population growth during the 1990s’ (p. 1), growing from 1.5 to 3.2 million in nonmetro counties. One of the key factors influencing movement away from traditional concentrations is the availability of employment. Frey and Liaw (1999) argue that out-migration from traditional points-of-entry occurs when too much competition for low-skilled jobs pushes migrants out of an area or when employment opportunities draw migrants to a new place. During the 1990s, work opportunities expanded in southern states and many migrants were attracted to agricultural and manufacturing employment in both urban and rural areas. In the South, Hispanic laborers can be found in carpet manufacturing (Engstrom, 2001), meat processing (Guthey, 2001), oil extraction (Donato et al., 2001), and on tobacco, fruit, poultry, hog, and dairy farms (Denton, 2002; Roka and Emerson, 2002; Studstill and Nieto-Studstill, 2001). Not just
temporary employees, many of these migrants have settled permanently into their new locations changing the face of southern communities. In many rural places, population decline is a constant threat to economic viability and the new influx of Hispanics contributes to reversing the rates of population loss, particularly in the South and Midwest (Kandel and Cromartie, 2004). Additionally, since this population tends to be younger than the average rural resident, it creates the potential for future population growth. Hispanic inmigrants also contribute to local economies often opening small businesses such as restaurants and grocery stores and further attracting family and friends to the area. Although the economic contribution of Hispanic migrants to rural communities is rarely quantified, a recent study of southern Michigan suggests that it could be quite significant. Rosenbaum (2002) views seasonal and migrant workers as a ‘community economic development event’ (p. 112) and estimates that per worker contribution to the economy can be as much as $18,000. Despite the potential economic development gains, many rural communities struggle to adapt to the special needs of this ethnically and culturally diverse group of newcomers who often need specialized social services. Community responses to migrant workers, in particular, often vary widely. Some communities are much more welcoming than others. For example, in Winchester, Kentucky, the proposed hiring of migrant workers by Georgia Vegetable Company led to large-scale opposition when the company proposed introducing Mexican farm workers (Denton, 2002) while other communities in Georgia have focused on the positive social and economic contributions of their new residents (Studstill and NietoStudstill, 2001). Finally, local schools often face the toughest challenges when the population composition changes, particularly if students and/or parents are not proficient English-speakers (Clark, 1998; Dale et al., 2001; Staff, 2003a). Across the U.S., growing Hispanic populations are contributing to increasing ethnic diversity, population growth, and economic productivity at both national and local levels. The American South, in particular, has become an increasingly popular destination for Latinos and indeed six of the top ten growth states for Hispanics are located in this region. The Hispanic Population in Kentucky In 1990, Latinos accounted for only 0.6 per cent of the total population in Kentucky, growing to 1.5 per cent in 2000. Despite its small size, this population represents a dramatic change in the cultural and ethnic diversity in Kentucky, which in 1990, was 92 per cent white (non-Hispanic) and seven per cent African American (non-Hispanic). The Hispanic population is also quite dispersed with small but growing Latino communities located around the state. As rates of natural increase continue to decline in Kentucky, domestic and international migration has become more important to population growth, accounting for the majority of growth between 1995 and 2000 (Price et al., 2004). International migrants comprise about 1 per cent of all Kentucky residents and almost 25 per cent of these
migrants are Hispanic or Latino. While these newcomers are an important source of growth for the state of Kentucky they also represent a fundamental shift in the ethnic and cultural diversity of the state. Local newspaper reports in Kentucky have paid much attention to the growth of the Hispanic population highlighting the struggles and triumphs of adapting to rural and urban environments. Most articles reflect positive contributions made by Hispanics to local economies including the growth of small businesses such as grocery stores (Staff, 2002) and goat farms (Morgan, 2003). Louisville, for example, is home to a growing Cuban community attracted by a lower cost of living and greater job opportunities (Garay, 2003a and 2003b) and contributing favorably to the revitalization of urban apartment complexes and rural small towns (Staff, 2003c). In rural communities and small towns, tobacco work is a big draw as well, with one article claiming that Latinos make up 80 per cent of tobacco workers in the state (Staff, 2003b). Others highlight growing pains in communities with large, new Hispanic populations. Struggles for health care (Cambell, 2003), discrimination and labor abuse (Staff, 2003b; Schreiner, 2003), and a pressing need for bilingual teachers and school officials (Staff, 2003a) are all indicative of the rapid influx of migrants and lack of community preparedness. In Kentucky, Hispanic population growth is not limited to population centers nor does it always correspond with areas of general population and economic growth; instead, small but growing Hispanic populations are found in both growing and declining nonmetro areas. A cartographic examination of this population suggests that by 2000, the distribution of Hispanics was changing through diffusion away from urban places into both periphery urban areas and nonmetro counties across the state (Figure 5.1). This expansion has been recent and widespread with clusters of rural Hispanic population growth found in both eastern and western Kentucky. A review of recent literature found few scholarly articles about Hispanic migrants in Kentucky. One exception is Denton’s (2002) analysis of the conflict in Winchester between the Georgia Vegetable Company and the community over the hiring of Mexican farm workers. The Kentucky State Data Center and the Center for Business and Economic Research each also published overviews of migration in Kentucky using different data sets (Perry and Clark, 2004; Price et al., 2004). Although both reports acknowledged growth of the Hispanic population, neither provided detailed analysis for this specific population. This chapter takes a first look at the distribution and characteristics of the Hispanic population in Kentucky. Data and Methods Analysis for this chapter utilizes the 1990 and 2000 U.S. Census data and the 2000, 5 per cent Public Use Microdata Sample (PUMS) for Kentucky. PUMS files are individual records of a sample of housing units and persons extracted from Census confidential microdata files (U.S. Bureau of Census, 2003, pp. 2-1). They contain most of the detailed records from the long-form questionnaires and allow estimates
to be produced that approximate other published data. For this analysis, the Hispanic origin variable in the PUMS data was used to identify the Hispanic population. This variable relies on self-identification and can be used to differentiate the ethnic background of the Hispanic population. Data also come from the 1990 and 2000 U.S. Decennial Census. While PUMS data are rich in demographic detail, the smallest geographic unit is a PUMA; an area comprised of several counties unless it is a moderately large city, in which case the city can be divided into several smaller PUMAs. Census data, in contrast, have less detailed demographic information for small populations but greater geographic specificity. The methodology used within this study can be divided into three parts. The first and second sections are primarily descriptive and cartographic analyses of migrant origins and socio-demographic characteristics and the third section uses multivariate analysis to evaluate the contribution of county-level characteristics in explaining the relative size of Hispanic populations across the state. An OLS regression model is fitted using 1990 and 2000 Census data. Variables were selected based on previous research and information about the Hispanic population in Kentucky. Demographic Analysis with Urban, Rural, and Mixed-UR Geographic Areas For the demographic analysis, Hispanics are divided into three groups based on their PUMA of residence in 2000. A PUMA is a geographic area, corresponding to the boundaries of a county or group of counties with a minimum population of 100,000 (U.S. Bureau of Census, 2003, pp. 1-1). A large metro county could also be divided into several PUMAs. In this chapter, each PUMA is categorized as Urban, Rural, or Mixed Urban-Rural. Because most PUMAs are comprised of several counties and not all counties are entirely urban or rural, three categories were developed to differentiate between PUMAs with all metro or all nonmetro counties. Metro and nonmetro designations come from the 2003 USDA Rural Urban Continuum Codes (ERS, 2003) in which each county is coded on a scale of one through nine based on the size of the urban population and proximity to urban places. For this study, counties identified as metro (codes 1-3) are Urban and those identified as nonmetro (codes 4-9) are Rural (Figure 5.2). The second step involved overlaying the PUMA boundaries with the county boundaries in a GIS. PUMA areas in which all counties or areas were metro are classified as Urban PUMAs and PUMA areas comprised of all nonmetro counties are classified as Rural PUMAs. PUMAs with both metro and nonmetro counties (mostly urban fringe areas) are categorized as Mixed Urban-Rural (Mixed-UR). There are 30 PUMAs of which nine are Rural, 12 are Urban, and nine are Mixed-UR. The 12 Urban PUMAs are primarily concentrated around the states’ large metro areas: Lexington (two PUMAs), Louisville and adjacent counties (seven PUMAs), Northern Kentucky-Cincinnati area (two PUMAs) and Owensboro (one PUMA). These areas have the largest population concentrations in the state and are linked to each other and other metro areas in adjacent states via major interstate highways. The majority of rural PUMAs are found in southern and eastern Kentucky (five
PUMAs) although several are located in the western and central parts of the state. These areas are the most spatially extensive and are characterized by small towns and very low population densities. In eastern Kentucky, the steep ridges and valleys characteristic of the Appalachian Mountains dominate much of the landscape while rolling hills are more common in the central and western landscapes. As in many rural places in the U.S., little or no public transportation exists in these rural areas and settlements are clustered in small towns with many outlying and widely dispersed housing units. The communities in these PUMAs lack proximity to urban areas and many also lack access to major roads or other transportation infrastructure. Mixed-UR PUMAs are the most diverse geographic areas. Five are adjacent to metro areas and form the transition zone between urban and rural areas with many characterized by suburban and exurban residential development. The remaining four PUMAs contain one or more small metro areas with populations less than 100,000. Although most of the counties in these four Mixed-UR areas are lightly settled, commuting patterns and residential settlements reinforce the economic influence of the small metro neighbor. For example, in eastern Kentucky, Ashland is an important, albeit small, urban area and in western Kentucky, Bowling Green and Hopkinsville form small urban nodes. Each of these PUMAs has a diverse mix of urban and rural landscapes. Dividing the PUMA areas into the three classes based on urban proximity provides a better context for evaluating differences within the Hispanic population than looking at state or county profiles. Rural PUMAs are particularly important since they are often places with the fewest resources and greatest social needs. Impoverished populations also face greater barriers in meeting everyday needs since low-density development requires access to individual transportation. Schools may have fewer resources to address the needs of non-English speaking students. These concerns decline somewhat as proximity to urban areas and resource availability increases. Results The Geography of Hispanics in Kentucky This first section of analysis utilizes county-level census data to evaluate rural growth areas. In the South, the movement of Latino populations into rural areas represents a new source of population growth for many communities and stems, at least to some extent, from increased job opportunities in rural places. As noted previously, growth of the Hispanic population, as measured by the decennial U.S. Census, was widespread (see figure 5.1). All counties, metro and nonmetro, in Kentucky grew except one: Meade County, a metro county in the northwest region of the state, in which the Hispanic population declined from 578 to 567. Of Kentucky’s 120 counties, 35 are categorized as metro counties and 85 are nonmetro. Thirty-four are considered completely rural.
