Demographic Shifts and Segregation in Fargo and West Fargo, North Dakota Schools

In this article, we examine how demographic changes in Fargo and West Fargo, North Dakota between 2000 and 2017, including the resettlement of refugees, have impacted equitable educational arrangements in Fargo Public Schools (FPS) and West Fargo Public Schools (WFPS). Drawing on multiple data sources, including North Dakota’s Department of Public Instruction (DPI), Common Core of Data (CCD) available from the National Center for Educational Statistics (NCES) and block group data from the U.S. Census Bureau, we use Geographic Information Systems (GIS) to examine city and district level changes in the years 2000 and 2017. We also conduct descriptive statistics and a multivariate analysis of variance (MANOVA) to assess the relationships among Black student enrollment, performance on state tests and enrollment characteristics that include race and free and reduced lunch. Findings underscore the increasing isolation of students over time by race, socioeconomic background and language.

The increasing segregation of public schools is especially concerning given that public school enrollment in the United States has grown substantially in size and diversity. Public schools are becoming less White and the student bodies have become poorer, trends that are likely to continue (Fry, 2007;Orfield, 2009). As racial and socioeconomic demographics around the country continue to evolve, school districts that have served predominantly White student populations are now serving larger populations of New Americans, people of color and English language learners (ELLs). Coupled with increasing segregation, these enrollment shifts generate questions of educational equity within the public education system.
The midwestern cities of Fargo and West Fargo, North Dakota are less commonly associated with racial and ethnic diversity. However, the rapidly changing composition of these cities that comprise an urban metro that lies on the west side of the Red River opposite of Moorhead, Minnesota, is shifting from an ethnically homogenous, predominantly White community, to a more racially and ethnically diverse community. Located 250 miles west of Minneapolis, Minnesota, Fargo and West Fargo seem unlikely international border towns situated on the domestic margin of North Dakota and Minnesota. But an examination of public schools in this transnational region is necessary as recent demographic shifts bring questions about educational equity in areas al. (2016) find that the sharp increase of students of color students enrolled in public schools coincides with the decrease of White students and results in the exposure of White students to other races. A larger share of students of color also come from lower socioeconomic backgrounds and they enroll in schools located in lower income neighborhoods. So, as White students are exposed to racially and socioeconomically mixed enrollments, students of color are getting less exposure to White students and more exposure to lower-income peers. Fry (2007) found that in 1993-1994, one-third of all White students in the U.S. attended an "all-White" school, schools in which 95% or greater of the enrollment was White. By [2005][2006], one in five white students was attending a nearly all-White school (Fry, 2007). Conversely, 29% of Hispanic students and 31% of Black students attended schools in 2005-2006 that were "all-minority", schools in which fewer than 5% of the students were White (Fry, 2007). The number of nearly all-minority public schools virtually doubled from 5,498 in 19935,498 in -19945,498 in to 10,135 in 20055,498 in -20065,498 in (Fry, 2007. Moreover, Orfield et al. (2016) found that between 1993 and 2013, the proportions of students who qualified for a free or reduced lunch increased for all racial groups. However, Black students and Latino students attended schools in which two-thirds of the students qualified for a subsidized lunch. Therefore, students of all races are increasingly exposed to lower-income peers, but students of color students have the most.
Reduced oversight of school district student assignment at the state and federal levels has also resulted in increased associations between school enrollment characteristics and residential segregation patterns (Reardon, Grewal, Kalogrides, & Greenberg, 2012;Reardon & Yun, 2005). Findings suggest that court-ordered desegregation plans for school districts are effective in reducing racial segregation, but school districts may rapidly resegregate to levels reflected in residential segregation patterns in the absence of continued court oversight (Reardon et al., 2012).

Methodology
National trends raise significant questions concerning educational opportunity and exclusion by social class, race, and geography at the local level, including Fargo and West Fargo, North Dakota, cities that are experiencing rapid demographic shifts. We examine the racial, socioeconomic and linguistic shifts that have occurred in Fargo and West Fargo, North Dakota and how these profound changes have affected educational opportunity in Fargo Public Schools (FPS) and West Fargo Public Schools (WFPS). We incorporate Geographic Information System (GIS) maps and state and federal data to study what has happened in Fargo and West Fargo, North Dakota within 17 years of demographic shifts, refugee resettlement and population growth.

Data
Data sources include the Common Core of Data (CCD) available from the National Center for Educational Statistics (NCES) between the 1999-2000 and 2017-2018 school years. The CCD data contained longitude and latitude information for each school and enrollment characteristics, including race, the number of refugees and English language learners (ELL) enrolled in each school, and the percentage of students in each school who receive free or reduced-price lunches in FPS and WFPS. North Dakota's Department of Public Instruction (DPI) provided student characteristics and assessment results for each school between 2007-2008 and 2017-2018, including the percent of students at each school who scored proficient on the English language arts (ELA) and science portions of the North Dakota State Assessment (NDSA), the statewide assessment that measures student performance against achievement standards. Census block group data from the American Community Survey (ACS), 5-year estimates available from the U.S. Census Bureau were downloaded for the years 2000, 2013, and 2017. Data include median home values, race, and language spoken at home. These years were selected because they had census block group data for Fargo and West Fargo. Census block level data were unavailable for Fargo and West Fargo.

Sample
The sample included 35 public schools operating in FPS and WFPS during the time periods of 1999-2000 and 2016-2017 school years. Charter schools do not currently operate in North Dakota. The schools operating in FPS and WFPS during the 2016-2017 school year were grouped into quartiles based on the percentage of Black students enrolled in each school to determine if relationships exist among schools with higher proportions of Black students, academic performance and other school enrollment characteristics. Therefore, we investigate whether inequitable educational arrangements exist in FPS and WFPS by race, socioeconomic background and language. Table 1 identifies the number of each district's schools in each of the quartiles. Although we had district data for the school year 2018-2019, we used data from the 2016-2017 school year because the MANOVA, described below, incorporates census block group data from 2017, the most recent year available.

Statistical Analyses
We conducted a multivariate analysis of variance (MANOVA) to test for significant relationships among the quartiles based on average Black student enrollment and the percentage of students proficient on the ELA and science portions of the NDSA, the percentages of students receiving a free or reduced lunch, refugees, immigrants, LEP, White, Hispanic, Asian, and students of two or more races enrolled at each school in FPS and WFPS. Additionally, we tested for significant relationships among the quartiles and the census block group characteristics of median income, median home values and the percentages of Black households living in the block groups surrounding each school in the sample. The alpha level of significance was set at p < .05. We also conducted descriptive statistics to illustrate patterns among student performance, race, socioeconomic background and language among the quartiles.

Fargo And West Fargo: A Brief History of the School Districts
The Fargo Township dates back to 1874, the same year as the establishment of FPS, School District 1 (Still, 1957). Fargo's founding coincided with the development of the Northern Pacific Railroad. As the railroad expanded west, a stop was needed on the west side of the of the Red River with the highest elevation possible to avoid floods (NDSU Archives, 2019;Still, 1957 The city and district boundaries have not changed since the passing of House Bill 1180 in 1973. Today, nearly half of WFPS local tax revenue comes from property in Fargo (Knutson, 2003). In 2003, 47% of the property tax that benefitted West Fargo schools came from property located in Fargo due the configuration of boundaries (Cole, 2003). Although the common boundary of FPS and WFPS has not changed since the passing of House Bill 1180, even as both cities and districts have grown, there have been attempts to alter the boundaries. In 2003, Al Carlson, a Republican state representative from Fargo, introduced House Bill 1428 that would "require land annexed by a city to become part of that city's school district if the newly acquired land wasn't already in the district" (Knutson, 2003). About 90% of the land Fargo could annex is located in the WFPS, but this bill failed in the North Dakota House (Knutson, 2003).

The Migration of Refugees to North Dakota
The migration of refugees, or New Americans, into the United States, including the Great Plains, continues to transform the racial and ethnic landscape of public school districts. North Dakota is one state that is undergoing a demographic restructuring due to the influx of New Americans, individuals who are immigrants or refugees. This article uses the term 'refugee' rather than the more inclusive term 'New American' to distinguish between immigrants and refugees. A refugee is an individual who was forced to leave his or her own country due to war or violence or "has a well-founded fear of persecution for reasons of race, religion, nationality, political opinion or membership in a particular social group" (United Nations High Commissioner for Refugees, 2019). An immigrant is an individual who is not a U.S. citizen nor a U.S. national who was granted the right to work and live within the U.S by the United States Citizenship and Immigration Services and who may be able to live permanently in the U.S. (IRS, 2019). North Dakota is sixth in the nation and first in the Great Plains for per capita resettlement of refugees, a startling statistic given that North Dakota is ranked as 48 th in the nation for total state population (Gaber et al., 2004). Lutheran Social Services of North Dakota (LSS), the organization that provides resettlement services for refugees in Fargo, has settled nearly 5,000 refugees from 40 countries in Fargo since 1990 and 40% of those settled are children (McMullen, 2016).

Refugee Resettlement in Fargo
Refugees are the majority of the overall foreign-born population of Fargo. In the 1980s, 465 refugees were resettled in the Fargo-Moorhead area (Singer & Wilson, 2006). By 2004, a total of 4,345 refugees were resettled (Singer & Wilson, 2006). As a state, North Dakota has served over 3,500 English language learners (Hageman, 2018). Fargo and West Fargo have 47% of the state's English language learner (ELL) population and 10% live in Grand Forks, a city 80 miles north of Fargo (Hageman, 2018). Both FPS and WFPS enroll significant numbers of ELLs. West Fargo Schools face some struggles to be ready for New Americans and refugees due to poor communication with LSS (Aksamit, 2000). FPS and WFPS sometimes receive two weeks' notice that LSS will be sending new students (Aksamit, 2000). Other times, schools are notified by the sponsoring families or not until they arrive at school (Aksamit, 2000).

City and District Racial Enrollment Shifts
The student enrollment demographics in FPS have evolved over time. The 2017 map indicates that the large census block group south of I-94 and west of I-29 increased in the number of households who speak a language other than English at home, rising from 2.5-5% in 2013 to 5-10% in 2017. Only two block groups located in West Fargo in 2017 have less than 2.5% of non-English speaking households compared to 18 in Fargo. The number of non-English speaking households in Fargo declined in 2017. The majority of census block groups located on the border of Minnesota are shaded in white, representing the 2.5-5% level. Several of the block groups located on the east side of I-29 and north of I-94 decreased in the percentage of non-English speaking households. However, two additional block groups located south of I-94 increased and are now shaded in red, indicating the 10-20% level.
The percentage of refugees enrolled in FPS and WFPS has also shifted. In 2013, WFPS had five schools with refugee enrollments over 7.5%, one of which was south of I-94. Four of five of these schools were in or bordering census block groups with higher percentages of non-English speaking households. FPS had seven schools with refugee enrollments over 7.5, three of which were south of I-94 and zero were located in the northern area of the district. Additionally, five of these schools were located in census block groups with higher percentages of non-English speaking households.
In 2017, one school in WFPS had refugee enrollment between 15-20% and two schools had between 10-15%, all of which are located east of the corridor in the city of West Fargo. Also, two schools had refugee enrollments between 5-7.5%. Although the number of schools south of I-94 increased, none of them had a refugee enrollment over 2.5% of total enrollment. In FPS, one school has a refugee enrollment between 15-20% and two between 10-15%. Although the northern area of FPS does have census block groups in which there are larger percentages of non-English speaking households, no school in that area has a refugee enrollment higher than 2.5%. In fact, the schools with larger refugee enrollments are located in the center of the district. Table 2 presents the distribution of FPS racial enrollment for the school years 1999-2000, 2007-2008, and 2017-2018 by percent of students that receive a free or reduced lunch. This table indicates that, over time, Black, Hispanic, American Indian, Asian and students who receive free or reduced lunches are disproportionately attending schools in which 45% or more of the students receive free or reduced lunches. In other words, over the past 18 years, FPS have begun experiencing greater racial and socioeconomic isolation in which students who are either poor or are of color are more likely than White students to attend schools that enroll predominantly poor students and students of color.  Vol. 9, No. 2;2020 In the 1999-2000 school year, 2% of White students attended schools in which 45-60% of students were receiving free or reduced lunches. This figure doubled to 4% by the 2007-2008 school year and increased seven-fold to 15% in the 2017-2018 school year. In comparison, 9% of Black students in the 1999-2000 school year attended schools in which 45-60% of students were receiving free or reduced lunches. By the 2007-2008 and the 2017-2018 school years, 13% and 34% of Black students, respectively, attended schools with predominantly poor enrollments. Hispanic, American Indian and Asian students each show similar patterns in which they are more likely than White students to attend schools with higher proportions of poor enrollments. Over time, students who receive free or reduced lunches are likely to attend schools with higher enrollments of poor students, from 7% in the 1999-2000 school year, to 16% in 2007-2008, and 35% in 2017-2018. These figures correspond to the national trend discussed by Orfield et al. (2016) that students of all races are increasingly exposed to lower-income peers, but students of color have the most.

Racial and Socioeconomic Isolation in FPS and WFPS
WFPS reflect trends parallel to FPS. Table 3 indicates that Black, Hispanic, American Indian, Asian and students who receive free or reduced lunches are disproportionately attending schools in which 30% or more of the students receive free or reduced lunches. During the 1999-2000 and 2007-2008 school years in WFPS, there were no enrollments in which 45%-65% of the enrollments were poor students. But, in the 1999-2000 school year, 4% of White students attended schools in which 30-45% of students were receiving free or reduced lunches, a figure that tripled to 13% by the 2007-2008 school year. In the 2017-2018 school year, 46% of White students attend schools in which 30-60% of the enrollment were poor students. Conversely, by 2017-2018, 68% of Black students, 61% of Hispanic, 59% of American Indian, 52% of Asian and 70% of poor students attended schools in which 30-45% of students were receiving free or reduced lunches. These figures correspond to the national trend discussed by Orfield et al. (2016) that students of all races are increasingly exposed to lower-income peers, but students of color have the most exposure. The tables indicate that students of all racial categories and students with low socioeconomic backgrounds enrolled at WFPS are attending schools that are more racially and economically segregated than those who attend FPS.

MANOVA Findings
Multicollinearity tests conducted before the MANOVA were performed revealed that the following independent variables had correlation coefficients of .80 and above that indicated higher degrees of correlation and were subsequently omitted from the MANOVA (Tabachnik & Fidell, 2013): percentage of immigrant and LEP enrollments, Hispanic students, White students, and students receiving a free or reduced lunch.
A non-significant Box's M test (p = .600) indicates homogeneity of covariance matrices of the dependent variables across the four quartiles. The Levene's Test of Equality of Error Variances was significant for the percentage of Black households living in the block groups surrounding each school (p < .001), the percentage of refugees enrolled (p <.001), and the percentage of Asian students enrolled (p = .012), indicating that the assumption of homogeneity of variances is violated. Given this, the level of significance for these variables was adjusted to a more conservative alpha of p < .01 (Tabachnik & Fidell, 2013). As the Levene's test was violated, the Pillai's Trace statistic was used to determine whether significant differences existed among the quartiles in terms of the independent variables (Tabachnick & Fidell, 2013).
The multivariate effect was significant, Pillai's Trace = 1.37, F(24,75) = 2.62, p < .001 . A multivariate partial η2 = .50 indicates that approximately 50% of multivariate variance of the dependent variables is associated with the group factor of the percentage of Black students enrolled. Table 4 displays the results of the Univariate tests indicated that there were significant differences across the four quartiles on the percentage of students proficient in science, F(3,30) = 6.09, p = .002 partial η2 = .38; the percentage of students proficient in ELA, F(3,30) = 4.21, p = .013, partial η2 = .30; the percentage of enrolled students who are two or more races F(3,30) = 3.56, p = .03, partial η2 = .26; and median home value, F(3,30) = 3.25, p = .035, partial η2 = .25. The percentage of enrolled refugees was significant at the alpha level of p <.01, F(3,30) = 13.83, p < .001, partial η2 = .58. The median household income of the block groups surrounding each school were not statistically significant at the alpha level of p <.05. The percentage of African American families living in the census block groups surrounding each school and the percentage of enrolled Asian students were not statistically significant at the alpha level of p < .01.
Tukey's post hoc tests indicated that the first quartile, the quartile with the fewest enrolled Black students had the highest science and ELA proficiency scores, and the lowest percentages of enrolled students who are refugees or who are two or more races. Although nonsignificant, the first quartile also had the highest median household incomes and the lowest percentage of enrolled Asian students.
The fourth quartile, the quartile with the highest percentage of Black students, had the lowest science and ELA proficiency scores, the highest percentages of enrolled refugees or students who are two or more races, and the lowest median incomes. Although nonsignificant, this quartile also had the lowest median home values and the highest percentage of African American households living in the block groups surrounding each school.

Descriptive Statistics Findings
Although immigrant enrollments, EL enrollments, and the percentage of students receiving a free or reduced lunch

Discussion
The maps and statistics we present underscore the isolation of students by race, socioeconomic background and language in FPS and WFPS. Over time, isolation by race and socioeconomic background has become more severe. The maps illustrate economic, racial and linguistic changes that have taken place within the cities of Fargo and West Fargo and within their school districts between 2000 and 2017. The statistical analyses indicate that students who are Black are likely to attend schools with higher percentages of refugees, ELs, immigrants, students of color and students who receive a free or reduced lunch. Additionally, schools that enroll higher shares of students of color, immigrants and refugees, and students who receive a free or reduced lunch tend to have lower science and ELA proficiency scores. Schools who serve higher shares of White students and lower enrollments of refugees, ELs, immigrants and students who receive a free or reduced lunch, or that are located in census block groups with higher median incomes, tend to have higher science and ELA proficiency scores.
Schools with less diverse student enrollments tend to be located on the northern section of FPS and in census block groups closer to the Minnesota border. Those areas are the areas with higher home values, fewer Black households and predominantly English-speaking households. Generally, WFPS serves a more racially and socioeconomically diverse enrollment than FPS. Although the census block groups in West Fargo are predominately English speaking, all except for two, individuals who live in the corridor located in Fargo attend schools in WFPS. As a result, WFPS have higher shares of refugee enrollments. Schools with higher shares of refugees tend to also serve higher Black, immigrant, and ELL enrollments and larger shares of students who receive a free or reduced lunch. The schools in WFPS that are located south of I-94 tend to have higher Black enrollments, larger enrollments of students who receive a free or reduced lunch, but smaller shares of refugees.
While the results indicate substantial inequitable arrangements for students based on race, socioeconomic background and language within FPS and WFPS, we have a limited understanding of how these arrangements occurred. The racial diversification of schools in West Fargo may be partially attributed to the resettlement of larger percentages of refugees in Fargo who live in the corridor. But, what is less understood is how enrollment boundaries have shaped enrollments throughout FPS and WFPS when the overall student enrollment is racially diversifying.
Enrollment boundary changes frequently occur when cities experience population growth. As cities grow and their respective school enrollments begin to stretch the capacity of existing buildings, new schools are built and enrollment boundaries are redrawn. Residential segregation is linked to school segregation because many school districts draw enrollment boundaries around immediate neighborhoods (Sohoni & Sapporito, 2009). When enrollment boundaries are drawn in this manner, schools become segregated because segregated neighborhoods produce segregated schools. Therefore, school attendance zones essentially determine the racial composition of school enrollment (Saporito & Sohoni, 2006). Oftentimes, racially segregated schools are also socioeconomically segregated, arrangements that are often linked to decreased student performance, teacher quality lower graduation rates, and higher dropout rates (Balfanz & Legters, 2004;Linn & Welner, 2007;Orfield & Lee, 2005;Swanson, 2004). Orfield and Eaton (1997) and Clark (1987) argue that school districts have drawn attendance boundaries to deliberately intensify segregation and undermine integration efforts. Moreover, evidence suggests that districts continue to draw racially inequitable boundaries (Orfield & Luce, 2009;Siegel-Hawley, 2013 (Reuer, 2017). The controversy arose due to the division of neighborhoods. For example, individuals who lived on either side of the same street were sometimes assigned to different schools (Pool, 2017). Also, students who attended one school the previous year were assigned to a different school the following year (Pool, 2017). This boundary change was perceived as a "one-year fix" by the superintendent who remarked that West Fargo would undergo "a steady diet of boundary changes in the next few years" as the district experiences population growth (Reuer, 2017).
In 2019, Fargo faced challenges in evaluating how the current boundaries could be altered to reduce overcrowded buildings (Amundson, 2019). Schools located in FPS have higher enrollments and crowded southside schools while northside schools, areas with predominantly White Households, are under-enrolled. A 40-member task force was charged with developing suggestions to draw new enrollment boundaries in FPS, with the final recommendation determined by the Board of Education (Amundson, 2019). The top five priorities for guiding the new boundaries were to: 1) Provide equitable space, resources and staff, 2) Strive for equitable number of students per classroom, 3) Strive for balance with a socioeconomic mix, 4) Set attendance areas to make best use of tax dollars, and 5) Work with area planners to consider stability (Amundson, 2019 p. A1). As of June 2019, the task force had not reached at least an 80% consensus on a school enrollment boundary recommendation, the percentage required for the recommendation to be made to the Board of Education (Fargo Public Schools, 2019).
School reassignment processes present significant opportunities for school districts to thoughtfully consider their evolving community demographics (Siegel-Hawley, 2013). But, when districts redraw enrollment boundaries in a way that exacerbates racial integration, educational gerrymandering ensues (Orfield & Luce, 2010). Richards (2014) argues that gerrymandered enrollment boundaries "provide an additional layer of stratification that exacerbates existing patterns of residential segregation" (p. 1121). However, attendance zones can be gerrymandered for the purpose of promoting integration. Without a strong commitment to diversity and integration, the opportunity to develop district wide diverse learning environments that promote positive educational outcomes and social benefits for all students may pass and isolation linked to redrawn boundaries can occur. For this reason, further research that investigates the role that enrollment boundaries have played in FPS and WFPS as mechanisms that perpetuate segregation and promote inequitable educational arrangements is necessary.

Limitations
Data for this study were taken from secondary sources so the authors had no control over the operationalized definitions and collection methods of the variables. For example, by definition, a census block group encompass an area that contains between 600 and 3,000 people (U.S. Census Bureau, 2019b). Some areas in FPS and WFPS are sparsely populated so the block groups will be larger than the typical block groups. As a result, some larger block groups overlap both school districts and this may affect the calculation of the community demographics.

Conclusion
Segregation is still an issue in public schools, a fact that is especially problematic given that enrollment diversifying across the nation. The rapidly changing compositions of Fargo and West Fargo, North Dakota, a once ethnically homogenous, predominantly White metro area, is transforming public school enrollment. But, as FPS and WFPS are changing in terms of race, class and language, these districts are producing unequal school arrangements that have lifelong impacts on students. Although FPS and WFPS are not regarded as the typical urban school district that grapples with segregation, predominantly White districts across the nation are becoming more diverse and they will need to determine how to best manage these changes. Becoming aware that inequitable school arrangements exist based on race, class or language is a first step towards progress. Determining how those arrangements have occurred may be a critical second step towards promoting an equitable educational landscape.