View in EDS HTML Full Text PDF Full Text

The Politics of Language Enrollment in US Higher Education

Saved in:
Bibliographic Details
Title: The Politics of Language Enrollment in US Higher Education
Language: English
Authors: Michael Gradoville (ORCID 0000-0002-8462-0009)
Source: Foreign Language Annals. 2025 58(2):367-391.
Availability: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
Peer Reviewed: Y
Page Count: 25
Publication Date: 2025
Document Type: Journal Articles
Reports - Evaluative
Education Level: Higher Education
Postsecondary Education
Descriptors: Language Enrollment, Second Language Learning, Second Language Instruction, Voting, Elections, Presidents, Political Attitudes, Predictor Variables, Correlation, College Second Language Programs, State Norms
DOI: 10.1111/flan.12798
ISSN: 0015-718X
1944-9720
Abstract: This article examines enrollments in languages other than English in United States higher education from the perspective of geographical distribution. While the overall decline in language enrollments is well known, enrollments are also very unequal across states when accounting for population. By cross-referencing MLA language enrollment data with demographic data, numerous predictors were tested for their efficacy in explaining inequality in Fall 2021 language enrollments. The only predictor found to affect language enrollments in a state was the state's political leanings as defined by voting in the 2020 U.S. presidential election: the more strongly a state voted for the Republican candidate, the lower its rate of language enrollment. An analysis of the changes in enrollment between 2016 and 2021 also showed a non-significant trend whereby states that voted more strongly for the Republican candidate tended to have more enrollment decline than those that voted for the Democratic candidate.
Abstractor: As Provided
Entry Date: 2025
Accession Number: EJ1477335
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwFD_VewHtnJxBT9iOmMUT-mAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDMfyFI-MD9crv0oPNgIBEICBm5z6NMY6_XGAd1ye425jwQ7pmmhVpXSHLQSZCAusSTK5qv62Gy20q4gxpqNg6z8ZQCw0nEwoPHKPTYa3NH6Tei0uR9tGMzInD8PPBAx3ZnZFDbbRSirrqHQ8pZ1DRMozpwYYeKgJL4QuOY6rmKzNF5MQlYQuti_mguHCwf1Vi_a4dZOqDUhgmpLEZkkBBOn2A_9N-gFaGVQ8K4VW
Text:
  Availability: 1
  Value: <anid>AN0186745414;fla01jul.25;2025Jul22.02:37;v2.2.500</anid> <title id="AN0186745414-1">The politics of language enrollment in US higher education </title> <p>This article examines enrollments in languages other than English in United States higher education from the perspective of geographical distribution. While the overall decline in language enrollments is well known, enrollments are also very unequal across states when accounting for population. By cross‐referencing MLA language enrollment data with demographic data, numerous predictors were tested for their efficacy in explaining inequality in Fall 2021 language enrollments. The only predictor found to affect language enrollments in a state was the state's political leanings as defined by voting in the 2020 U.S. presidential election: the more strongly a state voted for the Republican candidate, the lower its rate of language enrollment. An analysis of the changes in enrollment between 2016 and 2021 also showed a non‐significant trend whereby states that voted more strongly for the Republican candidate tended to have more enrollment decline than those that voted for the Democratic candidate.</p> <p>Keywords: language enrollments; languages other than English; political leanings; United States</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/FLA/01jul25/flan12798-gra-0001.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="flan12798-gra-0001.jpg" title="." /> </p> <p></p> <hd id="AN0186745414-3">INTRODUCTION</hd> <p>The present study investigated the distribution of enrollments in languages other than English (LOTEs) in higher education in the United States. Access to education in the United States is regarded as highly unequal (Mullen, [<reflink idref="bib38" id="ref1">38</reflink>]), and access to education in LOTEs is one area where such inequalities have been observed at the primary and secondary levels as well as within higher education (Glynn & Wassell, [<reflink idref="bib17" id="ref2">17</reflink>]). The present study advances this discussion by modeling state‐level 2021 enrollments in LOTEs in US higher education.</p> <p>The purpose of the present study was to examine enrollment in LOTEs at a state level using the <emph>MLA Language Enrollment Database, 1958–2021</emph> (Lusin et al., [<reflink idref="bib33" id="ref3">33</reflink>]; henceforth, MLA Data), specifically looking at enrollment numbers that were normalized based on the state's overall higher education enrollment. Since the report on Fall 2013, the absolute numbers of enrollments have continued to drop from 1,673,566 in Fall 2009 to 1,182,562 in Fall 2021 (Lusin et al., [<reflink idref="bib33" id="ref4">33</reflink>]), a drop of 29.3%. These statistics are alarming, especially because demographic trends are predicted to result in a contraction in higher education in the United States, possibly stretching into the 2040s (Harvey, [<reflink idref="bib20" id="ref5">20</reflink>]). That contraction can be expected to impact LOTE enrollments dramatically over the next two decades. However, it is necessary to examine the data from a more sophisticated perspective than absolute enrollments and to consider the particulars of different contexts within which LOTE enrollments occur (Tripiccione, [<reflink idref="bib55" id="ref6">55</reflink>]).</p> <p>MLA reports and other studies routinely examine regional differences in LOTE enrollments from the perspective of defined megaregions (e.g., South Atlantic: Alabama, District of Columbia, Florida, Georgia, Kentucky, Maryland, North Carolina, South Carolina, Tennessee, Virginia, West Virginia). However, because neighboring states can vary considerably in terms of demographic composition, wealth, and political stances, focusing on one state alone may provide a clearer picture. In particular, public institutions are usually controlled by a board that is appointed by the elected state government, which can affect the direction of the university. Given the role of state governments in public institutions, the partisan political leanings of the officials in said governments may play a role in which programs at universities receive administrative support and to what extent. Likewise, a student's own political leanings relate to their own world view and can influence their decision to enroll in LOTE coursework at the postsecondary level.</p> <p>To test these possibilities, I used regression analysis to determine which factors (i.e., states' demographic composition, financial support for higher education, and political leanings) best explained inequality in language enrollments that were normalized based on the states' overall higher education student populations. The present study also examined inequalities in the distribution of changes in normalized enrollment between Fall 2016 and Fall 2021.</p> <hd id="AN0186745414-4">BACKGROUND</hd> <p></p> <hd id="AN0186745414-5">Prior analyses of the MLA data</hd> <p>The most prominent analyses of the MLA Data are, naturally, the reports commissioned by the MLA (Furman et al., [<reflink idref="bib14" id="ref7">14</reflink>]; Goldberg et al., [<reflink idref="bib18" id="ref8">18</reflink>]; Looney & Lusin, [<reflink idref="bib29" id="ref9">29</reflink>]; Lusin et al., [<reflink idref="bib33" id="ref10">33</reflink>]). The reports reveal that enrollment in modern languages grew between 1958 and 1970 but suffered declines in the 1970s. From 1980 to 2009, enrollment growth was consistent with the exception of the 1995 census. After the 2009 census, LOTE enrollments declined in the 2013, 2016, and 2021 censuses. In terms of individual languages, most of the top 15 languages saw declines in enrollment from 2016 to 2020 with German seeing a 32.2% decline (Lusin, [<reflink idref="bib32" id="ref11">32</reflink>]). Of the top 15 languages, only Korean, Biblical Hebrew, and American Sign Language experienced increases in enrollment over the same period. The languages outside of the top 15, however, showed an increase in enrollment of 12.4%. Regarding institution level, less than 20% of enrollments in the census were recorded by 2‐year institutions, but from 2013 to 2016, 2‐year institutions saw greater declines (15.8%) than 4‐year institutions, a disparity that mostly disappeared between 2016 and 2020 (Lusin, [<reflink idref="bib32" id="ref12">32</reflink>]).</p> <p>Other analyses of the MLA Data include works by Thompson ([<reflink idref="bib53" id="ref13">53</reflink>]), Lusin ([<reflink idref="bib32" id="ref14">32</reflink>]), Tripiccione ([<reflink idref="bib55" id="ref15">55</reflink>]), Karmanov ([<reflink idref="bib24" id="ref16">24</reflink>]), and Nagano ([<reflink idref="bib40" id="ref17">40</reflink>]) with different foci. Thompson ([<reflink idref="bib53" id="ref18">53</reflink>]) presented an overview of the benefits of language learning, trends of language learning, and learners' motivations in the context of the United States. In her seven‐state sample from the MLA Data, Thompson ([<reflink idref="bib53" id="ref19">53</reflink>]) found a somewhat uneven distribution of LOTE enrollments, with New York having more than double the rate of enrollments of Florida, despite having relatively similar overall enrollments in higher education. While acknowledging the likely existence of a large number of variables that might contribute to this disparity, Thompson ([<reflink idref="bib53" id="ref20">53</reflink>]) pointed to funding models that penalize students for the increased coursework associated with double‐majoring in a language and another area of study. Thompson ([<reflink idref="bib53" id="ref21">53</reflink>]) also highlighted the existence of attitudes toward particular languages that may affect enrollments.</p> <p>The findings of Thompson's ([<reflink idref="bib53" id="ref22">53</reflink>]) study highlight the importance of going beyond raw numbers when analyzing the MLA Data, as state‐level enrollments are a function of the state's population, and all enrollments are a function of the number of students in higher education at any given time. Tripiccione ([<reflink idref="bib55" id="ref23">55</reflink>]) corrected for the number of students enrolled in higher education, finding that the share of higher education students enrolled in LOTEs actually peaked in 2006 (not 2009) but has since been in decline. Spanish had a disproportionate effect on the overall picture due to it having about half of LOTE enrollments for the last 40 years. Similarly, very large institutions had a disproportionate effect on the overall enrollment trends because they contributed so much to the overall enrollment numbers, in some cases obscuring trends at smaller institutions, such as 2‐year institutions. Regional differences represented another important variable, with the Far West region's post‐2009 decline in enrollment having a disproportionate effect on the overall trend due to the combination of its large population and more substantial declines (9% in 2013, 11% in 2016) compared to other regions.</p> <p>Trends in the variety of LOTEs studied have also changed over time (Karmanov, [<reflink idref="bib24" id="ref24">24</reflink>]). Whereas between 1958 and 1968, six "core" European languages (i.e., French, Spanish, German, Italian, Russian, and Latin) dominated the study of LOTEs in the United States, these languages (except Spanish) have given way to "emerging" languages, notably three East Asian languages (i.e., Chinese, Japanese, and Korean), Arabic, Portuguese, and American Sign Language.</p> <p>A more nuanced analysis of the data suggests that trends for community colleges might differ from those for 4‐year institutions. Whereas the overall trend in LOTE enrollment involved peaks and valleys, as previously described, Nagano ([<reflink idref="bib40" id="ref25">40</reflink>]) found that 2‐year institutions showed relatively steady increases between 1959 and 2009 that outpaced the growth rate at 4‐year institutions, recording substantial losses thereafter. LOTE study at community colleges in some years outpaced the overall growth of the community colleges themselves, a trend not shared by LOTE study in 4‐year institutions.</p> <p>Of particular importance for the present study is how these reports treated regional variation in the statistics on language enrollment, as well as how the data presentation allowed for meaningful comparisons. While recent MLA reports included totals and changes at a state level, they typically used six regions that appeared to correspond to the domains of the six regional affiliates of the MLA, the constituent states of which were often diverse. For example, although the states of the South Atlantic had geographical proximity in common as well as a historical association with slavery and the subsequent sociopolitical history, states such as Florida, Georgia, North Carolina, and Virginia had much larger populations, according to the US Census. The District of Columbia and its two surrounding states (Maryland and Virginia) were richer than the other states in the South Atlantic in terms of median household income, also according to the US Census. Finally, census data showed that Florida outpaced the other South Atlantic states as far as the population of foreign origin was concerned. The differing characteristics of these states might yield diverging enrollment trends relative to their neighbors. Consequently, the discussion of regional differences in the MLA reports was mainly informative for regional MLAs. Another issue was that the MLA reports used raw numbers of enrollments. It was thus not particularly surprising that California had the largest enrollment in LOTEs since it also had, by far, the largest population in the United States. A more informative view of state‐level enrollments would be to normalize enrollment statistics, accounting for the relative sizes of the states in question.</p> <p>One approach to this issue is to normalize higher education enrollment by state population. Thompson ([<reflink idref="bib53" id="ref26">53</reflink>]) applied this approach in her discussion of enrollments in seven states by dividing the total 2016 LOTE enrollments by states' populations. Although Thompson ([<reflink idref="bib53" id="ref27">53</reflink>]) did not use overall enrollment in higher education to directly normalize her enrollment data, such an approach is an alternative to using the state's population. The MLA reports have taken this approach to examine shifts in LOTE enrollment trends at the national level, as did Lusin ([<reflink idref="bib32" id="ref28">32</reflink>]) and Tripiccione ([<reflink idref="bib55" id="ref29">55</reflink>]). However, to my knowledge, no prior study has examined state LOTE enrollments, adjusting for the overall enrollment in higher education.</p> <hd id="AN0186745414-6">Rationalizing the decline in LOTE enrollments</hd> <p>The decline in LOTE enrollments since the late 2000s has been cause for concern within the field and, consequently, has motivated studies aiming to understand different aspects of this problem. One important ecological factor is the overall requirements that students are expected to fulfill with respect to LOTEs. Broadly speaking, requirements may take the form of (<reflink idref="bib1" id="ref30">1</reflink>) state‐level high school graduation requirements, (<reflink idref="bib2" id="ref31">2</reflink>) university entry requirements, and (<reflink idref="bib3" id="ref32">3</reflink>) university general education coursework requirements.</p> <p>Regarding state‐level high school graduation requirements, the analysis carried out by O'Rourke et al. ([<reflink idref="bib42" id="ref33">42</reflink>]) found that at the time, only seven states and the District of Columbia required coursework or proficiency in a LOTE for graduation, although an additional 22 states counted LOTE study toward a graduation requirement that could also be fulfilled by coursework in other areas. This means that nearly half of states (21/50 = 42%) had no LOTE requirement of any kind for graduation. Although a bill proposed in one more state (Utah) at the time would have made coursework or proficiency in an LOTE a requirement for high school graduation, most bills that O'Rourke et al. ([<reflink idref="bib42" id="ref34">42</reflink>]) reported either would have eliminated the LOTE requirement (Michigan) or would have allowed computer science courses to fulfill a requirement previously reserved for LOTE study (Florida, Washington), which could be expected to reduce high school LOTE enrollment. The impact of high school LOTE requirements on LOTE enrollments in higher education can be expected to vary. On the one hand, a larger number of entering students with LOTE proficiency can result in more students testing into higher levels of their respective languages. On the other hand, students with LOTE proficiency may be exempted from their university's LOTE requirement, if it exists.</p> <p>Some universities require coursework in a LOTE as a condition of admission. For example, the University of Washington requires students to have taken 2 years of an LOTE in high school or two quarters at another institution to be admitted (University of Washington, [<reflink idref="bib61" id="ref35">61</reflink>].). This approach to LOTE general education requirements can be expected to have a net effect of shifting LOTE enrollment from the institution to the high schools that feed into it, especially considering that the University of Washington has no university‐wide LOTE coursework requirement, although students' LOTE proficiency may contribute to them seeing value in said proficiency (Murphy et al., [<reflink idref="bib39" id="ref36">39</reflink>]; Van Gorp et al., [<reflink idref="bib62" id="ref37">62</reflink>]). The MLA periodically surveys institutions regarding their admissions and degree requirements in LOTE study (Lusin, [<reflink idref="bib31" id="ref38">31</reflink>]); however, the latest survey is 15 years old. At that time, 24.7% of institutions responding required LOTE study for entrance.</p> <p>Some universities' general education programs require graduates to demonstrate proficiency in an LOTE or to satisfy an LOTE coursework requirement. Spelman College presents an example of the former category, where all students graduating with a bachelor's degree must demonstrate proficiency in an LOTE equivalent to the fourth semester (Spelman College, [<reflink idref="bib50" id="ref39">50</reflink>]). The University of New Mexico's Core Curriculum, which represents an example of the latter, requires all bachelor's candidates to complete at least one course, at any level appropriate to their proficiency, in an LOTE (University of New Mexico, [<reflink idref="bib60" id="ref40">60</reflink>]). These types of requirements likely increase higher education LOTE enrollment. The MLA survey reported that 50.7% of institutions in 2009‐2010 required LOTE study for graduation (Lusin, [<reflink idref="bib31" id="ref41">31</reflink>]).</p> <p>While discussing the neoliberalization of higher education, Brown ([<reflink idref="bib7" id="ref42">7</reflink>], p. 23) notes the "rapid compression of general education requirements and time to degree" in response to the proliferation of what she refers to as "'best bang for your buck' rankings." The removal or reduction of a requirement in LOTEs reduces enrollment both at the lower and upper division levels, the former because of the now absent requirement and the latter because of the corresponding decrease in opportunities to recruit students into the major or minor. For example, the University of New Mexico College of Arts and Sciences removed its requirement that majors in the College complete coursework to the fourth‐semester level (University of New Mexico, [<reflink idref="bib58" id="ref43">58</reflink>], [<reflink idref="bib59" id="ref44">59</reflink>]). As a result, enrollments in LOTEs decreased from 6307 in Fall 2013 to 2720 in Fall 2021. From just Fall of 2016 to 2021, LOTE enrollment dropped by 47.4%,[<reflink idref="bib1" id="ref45">1</reflink>] only a small portion of which can be explained by a decrease in students at the university. The University of New Mexico example suggests that the recent declines in LOTE enrollments should not be attributed solely to declining student interest, as changes to requirements can have a dramatic effect.</p> <p>Aside from high school and university requirements, learners of LOTEs are often actively discouraged from language study by those around them. Thompson ([<reflink idref="bib52" id="ref46">52</reflink>]) examined how motivation for learning a language related to the language selected by students in Florida. She found that first‐language English speakers learning LOTEs, in many cases, defied the expectations of others (e.g., family, advisors, etc.). Many had people discouraging them from language study, especially when studying languages other than Spanish. It is likely that many students capitulate to such pressure, resulting in lower LOTE enrollments. The findings of Thompson ([<reflink idref="bib52" id="ref47">52</reflink>]) were mostly replicated by Thompson and Morgan ([<reflink idref="bib54" id="ref48">54</reflink>]) in West Virginia.</p> <p>While the preceding studies provided critical insights, they were largely limited to students who were already taking LOTE coursework. However, in a recent study, Murphy et al. ([<reflink idref="bib39" id="ref49">39</reflink>]) found that students at the University of Wisconsin‐Madison who had never enrolled in LOTE courses placed less value on proficiency in LOTEs than those enrolled at the time and previously enrolled. They also found that Hispanic/Latine students placed more value on proficiency than other students, suggesting that states with larger Hispanic/Latine populations might exhibit higher LOTE enrollment rates. Finally, students reported that they would be more likely to enroll in LOTEs if such study aligned better with their career plans, academic major, and personal interests. A follow‐up study at Michigan State University found similar results in many respects, although the overall value placed on the LOTE study was lower, which Van Gorp et al. ([<reflink idref="bib62" id="ref50">62</reflink>]) reasoned to be a consequence of fewer requirements for LOTE study at Michigan State compared to the University of Wisconsin‐Madison. They also found effects for home language, international student, and first‐generation student variables, which were either non‐significant in or had to be removed from the Murphy et al. ([<reflink idref="bib39" id="ref51">39</reflink>]) study. While these studies have many implications, it is clear that the value placed on LOTEs by institutions, as communicated to students with, among other things, requirements, plays a significant role in why students enroll in LOTEs at a decreased rate in recent years.</p> <hd id="AN0186745414-7">Ideologies and political affiliation</hd> <p>Belief systems and ideologies influence why students choose to or not to pursue the study of LOTEs, as well as why administrators choose to or not to support the study of LOTEs. Ideology "refers to a set of idea‐elements that are bound together, that belong to one another in a nonrandom fashion" (Gerring, [<reflink idref="bib15" id="ref52">15</reflink>], p. 980). While space precludes a complete treatment of all of the different language and political ideologies that might influence LOTE enrollments in higher education, the following discussion focuses on three clusters of ideologies: the neoliberal political ideology (Saunders, [<reflink idref="bib45" id="ref53">45</reflink>]), the one‐nation‐one‐language and related language ideologies (Fuller & Leeman, [<reflink idref="bib13" id="ref54">13</reflink>]), and the instrumentalism and commodification of language that combines political and language ideologies (Fuller & Leeman, [<reflink idref="bib13" id="ref55">13</reflink>]).</p> <p>Neoliberalism and neoliberal ideology are factors that affect the environment in which LOTEs are taught (Bernstein et al., [<reflink idref="bib3" id="ref56">3</reflink>]). Neoliberalism can be conceptualized in terms of an ideology, a mode of governance, and a policy package (Steger & Roy, [<reflink idref="bib51" id="ref57">51</reflink>]). As an ideology and mode of governance, neoliberalism champions free‐market capitalism and globalization with minimal governmental interference except to protect the interests of capital. Neoliberal political ideology may contribute in various ways to shifts in higher education in general, and LOTE education in higher education in particular (Steger & Roy, [<reflink idref="bib51" id="ref58">51</reflink>]). Economic policies in line with neoliberal ideology have led to declines in public spending overall on many services, including higher education, resulting in increased costs to students through tuition increases (Saunders, [<reflink idref="bib45" id="ref59">45</reflink>]). This has fueled the administrative desire for economic efficiency within the university, leading to, for example, the increased reliance on contingent faculty (Saunders, [<reflink idref="bib45" id="ref60">45</reflink>]).</p> <p>One notable example of neoliberal ideologies in education is school voucher programs. School voucher programs take taxpayer money out of public schools for parents to apply to private schools (Hayek, [<reflink idref="bib21" id="ref61">21</reflink>] [1960]). Although both major political parties have adopted neoliberal policies over the past several decades (Gerstle, [<reflink idref="bib16" id="ref62">16</reflink>]), nine of the 16 states that had a school voucher program in 2021 voted for the Republican candidate for president in 2020 (Education Commission of the States, [<reflink idref="bib11" id="ref63">11</reflink>]).</p> <p>The one‐nation‐one‐language ideology also likely plays a role in the decisions of students to pursue the study of LOTEs and of administrators to support programs in LOTEs. The one‐nation‐one‐language ideology assumes a one‐to‐one relationship between a language and a nation, with English being the language associated with the United States (Fuller & Leeman, [<reflink idref="bib13" id="ref64">13</reflink>]). This ideology is often cited to support policies banning bilingual education. Although this ideology likely exists to some extent across the political spectrum, a study of the language ideologies of Arizona voters showed that Republicans were far more likely to support monolingualist ideologies (Fitzsimmons‐Doolan, [<reflink idref="bib12" id="ref65">12</reflink>]). While a geographically more comprehensive treatment of the relationship between political preferences and language ideologies is called for, one might postulate that in areas that more heavily support Republican politicians, there will be a greater prevalence of the one‐nation‐one‐language and related ideologies, such as English hegemony, normative monolingualism, and zero‐sum ideologies (Fuller & Leeman, [<reflink idref="bib13" id="ref66">13</reflink>]), which may result in reduced enrollment in LOTE courses.</p> <p>A third ideology, the instrumentalism and commodification of language, or the marketing of language skills as a means to attain better jobs and, therefore, a higher paycheck, might influence enrollment in LOTEs if students do not see a financial incentive. This is particularly the case when one considers the increased expense of U.S. higher education, which itself is a result of economic policies informed by neoliberalism (Saunders, [<reflink idref="bib45" id="ref67">45</reflink>]). In the case that knowledge of a language is viewed as a commodity, if said commodity's value is deemed to be low, it can be expected that the likelihood of enrollment in a language will be correspondingly low. The previously mentioned study of Arizona voters found that Republicans tended to see less value in language as a skill than Democrats (Fitzsimmons‐Doolan, [<reflink idref="bib12" id="ref68">12</reflink>]). Given the increased likelihood of Republicans holding such a view, we would expect lower LOTE enrollments in areas with a larger number of Republicans, due to a combination of policy decisions that reflect this instrumentalism and students' enrollment decisions that are influenced by their own and others' instrumentalism.</p> <p>Based on the preceding discussion of language ideologies, supporters of the Republican Party are more likely to exhibit ideological stances that would disfavor the pursuit of LOTE study. As such, we might expect states with larger concentrations of such supporters to exhibit lower rates of LOTE enrollment.</p> <hd id="AN0186745414-8">Research questions</hd> <p>With the notable exception of Thompson ([<reflink idref="bib53" id="ref69">53</reflink>]), studies of the MLA Data have focused on issues other than geographical distribution. Consequently, relatively little is known about the geographical distribution of enrollments in LOTEs in higher education in the United States. Thus, the present study is guided by three research questions:</p> <p></p> <ulist> <item> 1. At a state level, how were higher education LOTE enrollments distributed in Fall 2021?</item> <p></p> <item> 2. What attributes of states best explained their rates of LOTE enrollments? To what extent does the political environment of the state relate to LOTE enrollments?</item> <p></p> <item> 3. How do the identified attributes of states contribute to the loss of LOTE enrollments?</item> </ulist> <hd id="AN0186745414-9">METHODS</hd> <p>The present study used openly available data from five sources: the <emph>MLA Language Enrollment Database, 1958–2021</emph> (Lusin et al., [<reflink idref="bib33" id="ref70">33</reflink>]; MLA Data), the U.S. Census Bureau's online tables for 2021 (U.S. Census Bureau, [<reflink idref="bib56" id="ref71">56</reflink>]; henceforth, Census Data), the National Center for Education Statistics' <emph>Digest of Educational Statistics, 2021</emph> (de Brey et al., [<reflink idref="bib8" id="ref72">8</reflink>]; henceforth, NCES Data), the Tax Policy Center's compilation of data on per capita state and local general expenditures (U.S. Census Bureau, [<reflink idref="bib57" id="ref73">57</reflink>]; henceforth, Tax Data), and the Federal Election Commission's <emph>Official 2020 Presidential General Election Results</emph> (Public Records Branch, [<reflink idref="bib43" id="ref74">43</reflink>]; henceforth, Election Results). The unit of analysis used within this study was the US state, as policy over at least public universities is mostly controlled at the state level. The following subsections address the response variable, the predictors, and the method of analysis.</p> <hd id="AN0186745414-10">Response variable</hd> <p>The response variable for this study was generated from a combination of data from the MLA Data and the NCES Data. For each state, the number of enrollments in LOTE courses in Fall 2021 was summed, including both undergraduate and graduate enrollment statistics from the MLA Data. These state LOTE enrollment totals were then normalized to compensate for different‐sized states. The response variable used the MLA total state enrollment divided by the NCES higher education enrollment and then multiplied by 100,000, yielding the state's enrollments per 100,000 higher education students (henceforth, EnrollHEd).[<reflink idref="bib2" id="ref75">2</reflink>]</p> <p>The MLA Data is not without statistical noise. While it can generally be considered a census, the response rate is not 100%. In addition, the enrollment data are assigned to specific universities, which themselves are assigned to specific states. In the case of online LOTE courses, individual students could be in different states (or possibly countries) than the university they attend. Given the growth of online education, it would also probably be ideal for the modality of enrollments to factor into the gathering of enrollment data. Likewise, the data are in terms of number of enrollments, not number of students. Consequently, if during Fall 2021, a given student were enrolled in two different LOTE courses, that student would be counted twice. In addition, a student enrolled in a short two‐course sequence in Fall 2021 would be counted twice, while a student enrolled in a single accelerated course that might cover the same material as two courses would be counted once. Furthermore, the data are subject to the whims of the individuals from each university who report the data. If, for example, enrollments in all course sections of a language's prefix are reported to the MLA, it is possible that some of the enrollments reported include courses not taught in the target language, potentially overreporting enrollments. These caveats do not prevent us from making important observations about these data, as the counts calculated from the MLA Data still reveal overall trends, and the aforementioned issues affect all states equally.</p> <hd id="AN0186745414-11">Predictor variables</hd> <p>Several predictor variables were included in this analysis, based on available data, as well as factors that might be expected to impact LOTE enrollment. These predictors are summarized in Table 1. First, per capita higher education spending in 2021 (henceforth, HEdSpend) was included as a predictor, using Tax Data. We would expect greater HEdSpend to result in greater support for language programs, whose generally smaller class sizes are potentially more costly to staff. Thus, higher EnrollHEd should result in states with higher HEdSpend.</p> <p>1 TABLE Predictor variables used in this study.</p> <p> <ephtml> <table><thead valign="bottom"><tr valign="bottom"><th>Predictor</th><th>Description</th><th>Hypothesized relationship</th></tr></thead><tbody valign="top"><tr><td><sc>HEdSpend</sc></td><td>Per capita public higher education spending in 2021</td><td>Positive</td></tr><tr><td><sc>ForeignBn</sc></td><td>Proportion of the population born outside of the United States</td><td>Positive</td></tr><tr><td><sc>HomeLang</sc></td><td>Proportion of households that speak a language other than English at home</td><td>Positive</td></tr><tr><td><sc>LimEnglish</sc></td><td>Proportion of households defined as having limited English proficiency</td><td>Positive</td></tr><tr><td><sc>PolSlant</sc></td><td>Difference in the percentage of voting for the Republican versus the Democratic presidential candidates in 2020</td><td>Negative</td></tr></tbody></table> </ephtml> </p> <p>Three predictors were included to address issues relating to the presence of people of foreign origin in the states as well as the prevalence of LOTEs, issues that are partially related. We might expect a larger presence of a foreign‐born population or a population with other languages to foment more study of LOTEs, due to a greater need for knowledge of said languages to provide services. We must note that the definitions of these variables based on the American Community Survey have been considered problematic (see Leeman, [<reflink idref="bib25" id="ref76">25</reflink>], for discussion and further references); however, no better data exists. The first variable, the proportion of the population born outside of the United States (henceforth, ForeignBn), was taken from Census Data's table S0501. We would expect greater ForeignBn to result in greater EnrollHEd for cultural and market‐based reasons. Regarding cultural reasons, a foreign‐born population may serve to normalize bi‐/multilingualism within the community in question relative to the country as a whole, which may encourage some students to enroll in LOTEs. With respect to market reasons, a foreign‐born population may increase the number of potential interlocutors in an LOTE to the point that knowledge of an LOTE is marketable for use in a job within the community. The second variable, the proportion of households that speak an LOTE at home (henceforth, HomeLang), was taken from Census Data's table S1601. This variable gets at the issues addressed by ForeignBn far more directly. For the aforementioned reasons, we would expect greater HomeLang to result in greater EnrollHEd. Naturally, HomeLang is a rather simplistic representation of what, for some families, may be a much more complex question as the use of a particular language in the home is often not exclusive. Finally, the proportion of households defined as having limited English proficiency (henceforth, LimEnglish) was included as a variable from Census Data's table S1602. Higher LimEnglish would be expected to result in greater EnrollHEd, since households with limited English proficiency will need services in their LOTEs. The efficacy of this variable is reduced by the fact that, within a given household, different individuals can be expected to have differing proficiency in English, in addition to the issues associated with reporting proficiency without a proficiency test.</p> <p>The final predictor addressed the overall political situation of the state in question. The predictor, PolSlant, was defined using data from Election Results as the percentage of votes for the Republican candidate in the 2020 election minus the percentage of votes for the Democratic candidate, yielding a positive number in states that the Republican won and a negative number in states that the Democrat won. While other votes could have been used, a notable advantage of the presidential election over other options was that voters in every state were presented with the same choice. In the case of US Senate elections, in addition to there being a different candidate for each seat, only approximately two‐thirds of the states have elections in any one cycle. In the case of the US House of Representatives, some districts do not have candidates from both the Democratic and Republican parties. A variable of this nature is motivated by the assumption made here that language ideologies are related to political ideologies and that political ideologies associated with the right are likely to result in less enrollment in LOTEs. Consequently, increased PolSlant (more votes for the Republican) was expected to yield less EnrollHEd.</p> <hd id="AN0186745414-12">Analysis</hd> <p>The analysis of the data in this study was conducted using the statistical programming language R (R Core Team, [<reflink idref="bib44" id="ref77">44</reflink>]).[<reflink idref="bib3" id="ref78">3</reflink>] The analysis occurred in three stages. In the first stage, the distributions of each variable were examined using the ggplot2 package (Wickham, [<reflink idref="bib65" id="ref79">65</reflink>]) while the correlations between different predictors were tested using the GGally package (Schloerke et al., [<reflink idref="bib47" id="ref80">47</reflink>]). Steps were taken to mitigate identified collinearities in the subsequent analysis.</p> <p>In the second stage, regression analysis was carried out. As the nature of the response variable is that of a count variable, a negative binomial model was used from the MASS package (Venables & Ripley, [<reflink idref="bib63" id="ref81">63</reflink>]), as models fit using Poisson or quasi‐Poisson regression were found to be overdispersed. Since regressions for count variables assume that the response variable will be a positive integer, the response variables were rounded to the nearest integer. All predictors were proportions or were scalar and were <emph>z</emph>‐scored to put them on the same scale and, therefore, make the estimates more directly comparable. Model selection followed a manual step‐wise procedure, stepping up. Interactions were not considered. Predictor variables were selected for inclusion in the final model iteratively according to improvements in the Akaike information criterion (AIC), as well as the existence of a statistically significant difference between the simpler and more complex models. At each iteration, large changes in predictor estimates were noted due to the potential for uncontrolled collinearities.</p> <p>In the third stage, to address the third research question, an additional analysis was performed, examining as a response variable the changes in EnrollHEd as predicted by the variables found to predict 2021 EnrollHEd earlier in the analysis. As this portion of the analysis depends more heavily on the results earlier in the analysis, additional steps will be described where this analysis is described in Section 4.4. The following section discusses the results of the procedure described herein.</p> <hd id="AN0186745414-13">RESULTS</hd> <p></p> <hd id="AN0186745414-14">Research Question 1</hd> <p>The MLA Data reported 1,182,562 Fall 2021 enrollments in LOTE courses in a country with a total enrollment in higher education of 18,991,798 in 2020, making an LOTE enrollment rate of 6231.797 enrollments per 100,000 higher education students. However, this enrollment was distributed throughout states in a highly unequal manner, as we can see in Figure 1. The density curve in this plot represents the distribution of states' values of EnrollHEd (<emph>x</emph>‐axis) conceived in a continuous manner while the histogram portion breaks EnrollHEd into bins, which are represented as bars. The y‐axis represents the frequency of particular values in the data set, or number of states within a particular range of values. The numbers on the <emph>y</emph>‐axis are in scientific notation; if all of the values below the density curve were summed, the result would be equal to 1. A table of the raw data behind this figure is provided in Appendix A. When normalizing based on enrollment in higher education (EnrollHEd), at the low end was New Hampshire at 2101.355 enrollments per 100,000 students, while the District of Columbia was at the high end with 18,013.767 enrollments per 100,000 students. The second‐place contender was Hawaii at 13,366.726 enrollments per 100,000 students, which was much lower than the District of Columbia. It is worth noting that New Hampshire was at the bottom of this distribution because it is a small state that hosts Southern New Hampshire University, a private liberal arts college that has seen explosive growth in its online program (denominator). The university itself has a very small language program (numerator) that has not participated in the online growth, with only 139 enrollments in Fall 2021.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/FLA/01jul25/flan12798-fig-0001.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="flan12798-fig-0001.jpg" title="1 Density plot and histogram of higher education LOTE enrollments in US states. LOTE, languages other than English." /> </p> <p></p> <p>The District of Columbia is unlike other states in ways other than its high LOTE enrollment. For example, regarding the predictors, the District of Columbia provided the Democratic candidate with an 86.8% margin of victory over his Republican rival in the 2020 presidential election. The next contender in this regard was Vermont, which gave the Democrat only a 35.4% margin of victory, less than half of the margin in the District of Columbia. As the District of Columbia is not a state, it also spent less than half per capita on higher education ($251) than the state with the next lowest spending ($553). In fact, its higher education is dominated by private institutions in a way that higher education in most states is not. In other words, the District of Columbia is outside of the normal range of states for many predictors, which is not entirely surprising, considering that it is also the nation's capital and, consequently, can be expected to differ from the states in unique ways. With this in mind, the District of Columbia was omitted from the statistical analysis, as it might skew the results relative to the broader trends in the data.</p> <p>Regarding the distributions of the predictor variables, Table 2 provides the minima, maxima, medians, and means for each one. PolSlant ranged from +43.4% in Wyoming to −35.4% in Vermont, reflecting the political polarization that currently affects the United States. Three variables in Table 2 have the same states as minima and maxima with some similarity in medians, namely the proportions of LimEnglish, HomeLang, and ForeignBn. In all three cases, West Virginia had the lowest proportion while California had the highest proportion, suggesting that these three variables were related. Finally, we can see that HEdSpend varied widely, from a minimum of $553 in Nevada to a maximum of $1718 in neighboring Utah.</p> <p>2 TABLE Descriptive statistics of predictor variables with the identity of states for minima, maxima, and medians.</p> <p> <ephtml> <table><thead valign="bottom"><tr valign="bottom"><th>Predictor</th><th>Min.</th><th>Median</th><th>Mean</th><th>Max</th></tr></thead><tbody valign="top"><tr><td><sc>PolSlant</sc></td><td>−35.4%(VT)</td><td>+0.6%(GA, NC)</td><td>+2.3%</td><td>+43.4%(WY)</td></tr><tr><td><sc>LimEnglish</sc></td><td>0.3%(WV)</td><td>2.2%(AK, DE, MN, NC, UT)</td><td>2.8%</td><td>8.3%(CA)</td></tr><tr><td><sc>HomeLang</sc></td><td>2.3%(WV)</td><td>11.7%(KS, NE, PA)</td><td>14.9%</td><td>43.9%(CA)</td></tr><tr><td><sc>ForeignBn</sc></td><td>1.6%(WV)</td><td>7.3%(PA, NE)</td><td>9.4%</td><td>26.6%(CA)</td></tr><tr><td><sc>HEdSpend</sc></td><td>$553(NV)</td><td>$965(AR, IN, OK)</td><td>$970.30</td><td>$1718(UT)</td></tr></tbody></table> </ephtml> </p> <hd id="AN0186745414-16">Multicollinearity in the predictors</hd> <p>Moving into a discussion of the correlations among the variables, Figure 2 is a Kendall's tau correlation matrix of the predictors included in the study. Each variable is assigned to a column and a row. Where the variable meets itself in the diagonal going from the top left to the lower right, a density plot is included to show the distribution of that variable. In the upper right portion of the plot are Kendall's tau correlations between each pair of variables, including the standard symbols to indicate the significance level. In the lower left portion of the plot are scatter plots with loess lines to show the relationship between each pair of variables.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/FLA/01jul25/flan12798-fig-0002.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="flan12798-fig-0002.jpg" title="2 Kendall's tau correlation matrix of the predictor variables included in the study." /> </p> <p></p> <p>More than half of the pairwise Kendall's tau correlations in Figure 2 were significant. The strongest correlations shown in Figure 2 were between HomeLang and ForeignBn (0.867), between HomeLang and LimEnglish (0.837), and between LimEnglish and ForeignBn (0.831). The positive correlation between, for example, HomeLang and ForeignBn (0.867) meant that as the percentage of LOTE home languages increased, so did the percentage of the population that was born outside of the United States. Given that these three variables can be expected to be related, it was unsurprising that they correlated highly. PolSlant also correlated relatively strongly with ForeignBn (−0.532), and slightly less so with HomeLang (−0.456) and LimEnglish (−0.457). The negative correlation between PolSlant and ForeignBn (−0.532) meant that as the vote share for the Republican candidate increased, the percentage of the population born outside of the United States decreased. No correlation between HESpend and any other variable was significant.</p> <p>There are two main approaches to addressing collinearity in regression modeling, namely dimensionality reduction with principal components analysis and combining variables. While principal components analysis is very effective at identifying common variances among variables, it can be difficult to determine how to interpret the influence of individual predictors since their associated variance is assigned to multiple Principal Components. Consequently, the approach of combining variables was selected instead. To this end, HomeLang, ForeignBn, and LimEnglish were combined into a single variable by first <emph>z</emph>‐scaling them so that they were all expressed in terms of standard deviations from the mean and then averaging them. Henceforth, this combined variable will be referred to as NonEngInf for non‐English influence.</p> <p>Regarding the correlations involving PolSlant, the researcher decided to leave them as is because the correlations were not as strong as the other aforementioned correlations and they were more clearly measuring different things; however, these correlations were considered both in terms of model building and in theoretical interpretation of the results. The following subsection addresses how well these variables predict higher education LOTE enrollment at the state level.</p> <hd id="AN0186745414-18">Research Question 2</hd> <p>Table 3 presents the best negative binomial regression model of predictors of EnrollHEd. The table contains a column for the estimate, columns for the lower bound and upper bound of the 95% confidence interval, the <emph>z</emph> value, and the <emph>p</emph>‐value. Only PolSlant is a significant predictor of EnrollHEd.[<reflink idref="bib4" id="ref82">4</reflink>] Appendix B shows the coefficients and model fits of all models fit during the manual stepwise process.</p> <p>3 TABLE Negative binomial regression model of predictors of states' EnrollHEd.</p> <p> <ephtml> <table><thead valign="bottom"><tr valign="bottom"><th /><th>Estimate</th><th>CI 2.5%</th><th>CI 97.5%</th><th><italic>z</italic> value</th><th><italic>p</italic></th><th /></tr></thead><tbody valign="top"><tr><td>Intercept</td><td>8.73234</td><td>8.7283715</td><td>8.7354665</td><td>215.21</td><td><0.001</td><td>***</td></tr><tr><td><sc>PolSlant</sc></td><td>−0.17255</td><td>−0.1806393</td><td>−0.1735355</td><td>−4.21</td><td><0.001</td><td>***</td></tr></tbody></table> </ephtml> </p> <p>1 <emph>Note</emph>: Null deviance: 68.807 on 49 degrees of freedom.</p> <ulist> <item>2 Residual deviance: 50.687 on 48 degrees of freedom.</item> <item>3 AIC: 893.48.</item> </ulist> <p>The Intercept of 8.73234 indicated that, when PolSlant was zero,[<reflink idref="bib5" id="ref83">5</reflink>] the model predicted that a state's EnrollHEd would be 6200.22. With PolSlant's negative estimate (−0.17255), increased voting for the Republican candidate saw a reduction in EnrollHEd. A state that voted for the Republican one standard deviation above the mean would be predicted to have 84.2% of the enrollment, or 5217.586 enrollments per 100,000 higher education students, of the variable mean of 6200.22. On the other hand, if a state voted for the Republican one standard deviation below the mean, it was predicted to have 115.8% of EnrollHEd of the mean, or 7182.854. This effect can be visualized in the scatter plot in Figure 3. The <emph>x</emph>‐axis of this scatter plot represents PolSlant, while the <emph>y</emph>‐axis represents EnrollHEd. The scatter plot also includes a regression line that demonstrates the overall trend, which was observably negative in line with the estimate from Table 3. While a few states in Figure 3 strayed far from the line, overall the trend appeared to be consistent.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/FLA/01jul25/flan12798-fig-0003.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="flan12798-fig-0003.jpg" title="3 Scatter plot of the relationship of PolSlant to EnrollHEd." /> </p> <p></p> <hd id="AN0186745414-20">Research Question 3</hd> <p>As discussed above, Lusin et al. ([<reflink idref="bib33" id="ref84">33</reflink>]) reported a decline of 16.6% in LOTE enrollments between Fall 2016 and Fall 2021, noting increased rates of university non‐responses in Fall 2021 possibly due to the circumstances of the time, namely the COVID‐19 pandemic and resulting turmoil. Considering only universities that responded in both years, they reported a decline of 15.6%. Some of the non‐responses, nevertheless, could have been due to program elimination.</p> <p>The shifts in enrollment reported by Lusin et al. ([<reflink idref="bib33" id="ref85">33</reflink>]) were based on raw numbers and did not account for shifts in overall higher education enrollment. As such, enrollment rates were calculated for Fall 2016 and 2021, taking into consideration the population reflected (see Appendix A). Using NCES data from Snyder et al. ([<reflink idref="bib49" id="ref86">49</reflink>]), the national enrollment rate in 2016 was 7158.686 per 100,000 students (7094.196 without DC). The comparable number in 2021, on the other hand, was 6231.797 per 100,000 students (6170.961 without DC). When examined in this manner, considering the overall higher education enrollment, the decrease in enrollment rate was 12.9% (13.0% without DC), a smaller percentage drop than when examining only raw numbers. This indicates that some portion of the decline in raw LOTE enrollment was due to a decrease in the number of students in higher education.</p> <p>As reported by Lusin et al. ([<reflink idref="bib33" id="ref87">33</reflink>]), the decline in enrollments was unequal. When normalizing based on higher education enrollment, changes in enrollment varied from an 11.5% increase in Vermont to a 43.8% decrease in neighboring New Hampshire, the latter of which can be explained by the aforementioned explosive growth at Southern New Hampshire University coupled with a very small language program. Given the results from the previous sections, it made sense to see whether enrollments were declining at a faster rate in states that voted more heavily for the Republican candidate. Figure 4 is a scatter plot that relates PolSlant to the ratio of the 2021 to the 2016 EnrollHEd. The slightly negative slope provided some support for this notion with greater decreases in LOTE enrollments in states that voted more Republican; however, a linear model that was fit to relate PolSlant to the change in EnrollHEd did not find a significant relationship, which was not surprising when considering the diffuse relationship in Figure 4.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/FLA/01jul25/flan12798-fig-0004.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="flan12798-fig-0004.jpg" title="4 Scatter plot of 2020 Election Result as a predictor of enrollment change." /> </p> <p></p> <hd id="AN0186745414-22">DISCUSSION</hd> <p>Returning to the first research question, regarding the distribution of state‐level enrollments in higher education LOTE courses, LOTE enrollment was very unevenly distributed. When considering the number of students in higher education, students in Hawai'i enrolled in LOTE courses at more than six times the rate of students in New Hampshire. If we were to consider the District of Columbia, which had even higher enrollment than Hawai'i, likely due to diplomatic and defensive applications of LOTEs that are readily apparent in a city that hosts almost 200 embassies of foreign governments, this imbalance would be even greater. While these enormous disparities may be due to some combination of differences in access to, support for, and interest in the LOTE study, the potential causes will become clearer as I continue interpreting the data.</p> <p>Regarding the second research question, some variables clearly did not factor into state‐level LOTE enrollments. The inability of NonEngInf to improve the statistical models in any way, by and large, suggested that universities were not responsive to the presence of populations of LOTE speakers in their communities. This result was surprising when considering the existence of programs in less commonly taught languages in areas with large populations of speakers of that language, such as Portuguese in Massachusetts, Vietnamese at several institutions in the Los Angeles area, Navajo at several institutions in Arizona and New Mexico, and Czech at the University of Nebraska, Lincoln. However, it must be noted that, as these are all less commonly taught languages, the size of the programs in question was very small, even when compared to Spanish programs at smaller institutions. In other words, the existence of language programs responsive to local needs may, in fact, be exceptional. Given the lack of a relationship between HEdSpend and EnrollHEd, although increased spending on higher education may result in increased higher education enrollment, LOTE enrollment was not affected in a disproportionate manner.</p> <p>Regarding the relationship between PolSlant and LOTE enrollment, as previously discussed, the voting patterns in the 2020 United States presidential election serve as a natural index for the constellation of ideologies that likely exist within a particular subunit of the United States. While neoliberal ideology certainly influences the behavior of students, neoliberal policies increasingly affect higher education as well as particular areas of study, such as LOTEs. Within the neoliberal order, all actors are viewed following the example of the contemporary firm, and as such must maximize their present and future value "through practices of entrepreneurialism, self‐investment, and/or attracting investors" (Brown, [<reflink idref="bib7" id="ref88">7</reflink>], p. 22). The increased reliance on tuition to fund higher education, resulting from decreased government investment (Saunders, [<reflink idref="bib45" id="ref89">45</reflink>]), motivates students as firms to choose universities and areas of study on the basis of economic metrics such as return on investment, job placement, and future income (Brown, [<reflink idref="bib7" id="ref90">7</reflink>]). The shift toward the view of individuals as firms means that</p> <p>institutions of higher education cannot now recruit students with the promise of discovering one's passion through a liberal arts education. Indeed, no capital, save a suicidal one, can freely choose its activities and life course or be indifferent to the innovations of its competitors or parameters of success in a world of scarcity and inequality (Brown, [<reflink idref="bib7" id="ref91">7</reflink>], p. 41).</p> <p>As a firm itself, an academic department must similarly engage in entrepreneurialism, self‐investment, and attracting investors, the latter of which comes in the form of students and parents along with external funding bodies (grants) and philanthropists (donations); however, while academic departments can be conceived of as firms, so too can their potential investors, who would be motivated by similar return on investment issues. Thus, academic departments unable to convince potential investors of their potential return on investment are unlikely to receive said investment.</p> <p>A recent example that received national attention where a language program fell victim to political and administrative decisions made due to neoliberal ideology is the program at West Virginia University, the flagship institution of a state that voted Republican more than any state except Wyoming. Following the neoliberal desire for continued growth (Steger & Roy, [<reflink idref="bib51" id="ref92">51</reflink>]), the current university president pledged in 2014 to expand the university to 40,000 students (Mays, [<reflink idref="bib35" id="ref93">35</reflink>]). Enrollment instead declined in the ensuing 9‐year period, particularly after the start of the COVID‐19 pandemic (Slade, [<reflink idref="bib48" id="ref94">48</reflink>]), but also due to a demographic shift that resulted in fewer people of typical college age as well as a decreased proportion of college enrollments (Dean, [<reflink idref="bib9" id="ref95">9</reflink>]). Consistent with other universities before it (Mintz, [<reflink idref="bib36" id="ref96">36</reflink>]), according to one faculty member, the university administration invested considerable capital to create space for growth that never materialized (Jones, [<reflink idref="bib23" id="ref97">23</reflink>]). One analysis, however, assigned the causation of the budget crisis faced by the university to the withdrawal of funding from the state government (Allen, [<reflink idref="bib1" id="ref98">1</reflink>]), such austerity approaches being a policy promoted by neoliberal ideology. To resolve the resulting budgetary shortfall, the university opted to cut programs, with the hardest hit area being the university's language department, where all majors in modern languages were eliminated. The rationale cited, declining numbers of majors, did not take into account the fact that language programs often serve students majoring in other areas.[<reflink idref="bib6" id="ref99">6</reflink>] Given the fact that West Virginia already had one of the lowest rates of language enrollment in the entire country in 2021 (EnrollHEd = 3290.935), with the bulk of that at West Virginia University (54.2%), it is likely that West Virginia will drop to the bottom of the list due to the resulting lack of access to LOTE education for the population of the state.</p> <p>As previously discussed, the one‐nation‐one‐language ideology likely played a role in the reduced enrollment in LOTEs in states that voted more heavily for the Republican candidate. Previous research has shown that Republicans are more supportive of monolingualist ideologies (Fitzsimmons‐Doolan, [<reflink idref="bib12" id="ref100">12</reflink>]), a position that can be expected to have increased with the more strongly anti‐immigrant stance on the political right since that study was carried out. Assuming this to be the case on a broader level, even if students do not subscribe to the one‐nation‐one‐language ideology, research has shown that students learning LOTEs are, in many cases, defying the expectations of family members (Morgan & Thompson, [<reflink idref="bib37" id="ref101">37</reflink>]; Thompson, [<reflink idref="bib52" id="ref102">52</reflink>]; Thompson & Morgan, [<reflink idref="bib54" id="ref103">54</reflink>]). It is important to remember that this finding comes from studies of students mostly enrolled in LOTE coursework. Largely absent from the discussion are students who do not make it this far, students who may wish to learn an LOTE, but whose personalities or abilities may not allow them to defy lack of support (or possibly even hostility) within their network. The pattern uncovered relating PolSlant to LOTE enrollment likely represents the cumulative macro‐effect of many little decisions that individual students make that ultimately produce this outcome.</p> <p>A similar analysis applies to the ideology surrounding instrumentalism and the commodification of language. As noted before, enrollment in LOTEs may be dispreferred by students if there is not a more immediate financial incentive associated with the study of LOTEs, especially when considering the growing expense of higher education in the United States. In cases where the knowledge of LOTEs is viewed mainly in commodity terms, said commodity is not valued equally within the population. As previously mentioned, Republicans tend to see less value in language as a skill (Fitzsimmons‐Doolan, [<reflink idref="bib12" id="ref104">12</reflink>]). While instrumentalism may influence students not to pursue LOTE study, those students who do not share this view may be subject to a lack of support, especially from their parents (Morgan & Thompson, [<reflink idref="bib37" id="ref105">37</reflink>]; Thompson, [<reflink idref="bib52" id="ref106">52</reflink>]; Thompson & Morgan, [<reflink idref="bib54" id="ref107">54</reflink>]), who in many cases are funding their study in higher education and may have considerable sway. All of this is to say that part of the relationship between PolSlant and LOTE enrollment is likely a consequence of the relative instrumental value assigned to LOTE study by students and/or their parents.</p> <p>Examining the third research question, which addresses whether the attributes identified as predicting LOTE enrollments in different states also contribute to the ongoing loss of enrollments in said states, although there is certainly a trend where more Republican‐leaning states saw a greater decline in enrollments than more Democratic‐leaning states, the relationship was not significant. It is possible that the period between 2016 and 2021 is too shallow of a time depth to detect such an effect; however, overall, the lack of effect suggests that the relationship between PolSlant and enrollment has existed for some time.</p> <hd id="AN0186745414-23">FUTURE RESEARCH</hd> <p>The preceding discussion points to the need for further research in various areas related to LOTE enrollments. The assumption that Republicans are more likely to have the one‐nation‐one‐language ideology or to view language primarily instrumentally, assigning it a low value, is primarily based on a single study of Arizona voters (Fitzsimmons‐Doolan, [<reflink idref="bib12" id="ref108">12</reflink>]). More work in this area is needed to determine whether this continues to be the case and whether it is true across a broader portion of the U.S. population. In addition, while studies have found that students of LOTEs are likely to be defying the expectations of others (Morgan & Thompson, [<reflink idref="bib37" id="ref109">37</reflink>]; Thompson, [<reflink idref="bib52" id="ref110">52</reflink>]; Thompson & Morgan, [<reflink idref="bib54" id="ref111">54</reflink>]), more work is needed to determine the root cause of these expectations. Are these expectations motivated by the ideologies highlighted here, or are there other factors at play? Furthermore, given the large number of students who do not enroll in LOTEs, it behooves the field to build upon the work by Murphy and colleagues (Murphy et al., [<reflink idref="bib39" id="ref112">39</reflink>]; Van Gorp et al., [<reflink idref="bib62" id="ref113">62</reflink>]) to study the characteristics of the students who do not enroll as compared to those who do enroll. Understanding the motivations of such students could be instructive as to how to reach them, even if there is no change in general education requirements.</p> <hd id="AN0186745414-24">A CALL TO ACTION</hd> <p>Aside from further research, the present study, as well as related studies (e.g., Lusin, [<reflink idref="bib31" id="ref114">31</reflink>]; Morgan & Thompson, [<reflink idref="bib37" id="ref115">37</reflink>]; Van Gorp et al., [<reflink idref="bib62" id="ref116">62</reflink>]), point to a need for action at a variety of points for the continuation of LOTE study in the United States. I here divide these into three groupings: within system, policy goals, and systemic change.</p> <p>Within‐system action items acknowledge the neoliberal reality, influenced by ideologies unfavorable to the LOTE study, that we currently inhabit. Considering the reality that LOTE study must compete with other fields for students' "business," it behooves departments with LOTE programs to understand their (potential) students and universities (Thompson & Morgan, [<reflink idref="bib54" id="ref117">54</reflink>]) and to carry out effective marketing campaigns with the appropriate constituencies (Bezerra et al., [<reflink idref="bib6" id="ref118">6</reflink>]). Aside from the overall need to attract students, LOTE programs must look for ways to educate stakeholders (e.g., advisors, administrators, parents, and students) about what learning a language requires to manage expectations about what is possible over the course of a semester, as one or two courses in a LOTE are insufficient for a student to attain professional‐level proficiency.</p> <p>Another important attribute that departments must develop within their LOTE programs is the structure of their curricula, which must take into consideration the institutional context in which they exist. Among other things, such curriculum revision must truly reflect the substance of the American Council for the Teaching of Foreign Languages (ACTFL) standards (i.e., ACTFL, [<reflink idref="bib2" id="ref119">2</reflink>]; National Standards Collaborative Board, [<reflink idref="bib41" id="ref120">41</reflink>]). To empower LOTE instructors and departmental administrators to do this, broad training in the ACTFL standards is a sine qua non. While, of course, this should form a part of the preparation of teaching assistants, all department personnel require regular training on how the latest ACTFL standards translate into classroom teaching practices. This would help mitigate the disconnection with real‐world applications that people tend to perceive in language instruction. As part of revising the curriculum to reflect ACTFL standards, a smoother transition between lower and upper‐division courses might also help with the retention of students. In addition, third‐year composition and conversation courses should be re‐oriented toward professional concerns, focusing on skills in the target language such as writing cover letters for jobs, participating in job interviews, preparing a resume, reading news articles about employment opportunities in countries where the language is spoken, etc.</p> <p>As far as policy goals are concerned, faculty in LOTE programs must advocate for the importance of LOTE study in the development of well‐rounded citizens. Within the institution, a college‐ or university‐wide proficiency‐based language requirement, such as the aforementioned requirement at Spelman College, should be (re‐)established in cases where one does not currently exist to highlight the importance of language study. Considering the findings of Murphy and colleagues (Murphy et al., [<reflink idref="bib39" id="ref121">39</reflink>]; Van Gorp et al., [<reflink idref="bib62" id="ref122">62</reflink>]), the language requirement does more than to simply increase lower division enrollment in LOTEs. The requirement relays the values of the institution: institutions that have a language requirement communicate to their students that LOTE proficiency is important, while institutions lacking such a requirement effectively devalue LOTE study in the eyes of their students. Aside from the need for institutions to communicate the value of LOTE study, administrators must allocate the necessary resources to adequately support this endeavor. This includes reducing barriers within the university that stifle LOTE programs' marketing efforts. Outside of the university, policy‐makers must be informed of the importance of LOTE study with the included goal of the provision of financial resources specifically to promote LOTE study.</p> <p>Finally, with respect to systemic change action items, departments with LOTE programs should consider creating a space for discussion of the underlying ideologies that both result in the current state of LOTE study in the United States and in other negative consequences elsewhere within society. Study of the scholarship on the neoliberal aspects of contemporary literature and culture (K. Bezerra, [<reflink idref="bib4" id="ref123">4</reflink>]; L. Bezerra, [<reflink idref="bib5" id="ref124">5</reflink>]; Deckard & Shapiro, [<reflink idref="bib10" id="ref125">10</reflink>]; Lehnen, [<reflink idref="bib27" id="ref126">27</reflink>]; inter alia) can help foment a critical consciousness of the effects of neoliberalism on contemporary society. Discussion of language ideologies has long formed a part of sociolinguistics (Lippi‐Green, [<reflink idref="bib28" id="ref127">28</reflink>]; Schieffelin et al., [<reflink idref="bib46" id="ref128">46</reflink>]) and, more recently, of heritage language programs (Martínez, [<reflink idref="bib34" id="ref129">34</reflink>]; Villa, [<reflink idref="bib64" id="ref130">64</reflink>]; inter alia), leading to the adoption of critical language awareness pedagogy (Gutiérrez, [<reflink idref="bib19" id="ref131">19</reflink>]; Holguín Mendoza, [<reflink idref="bib22" id="ref132">22</reflink>]; Leeman, [<reflink idref="bib26" id="ref133">26</reflink>]; Loza & Beaudrie, [<reflink idref="bib30" id="ref134">30</reflink>]; inter alia); however, it is relevant to language departments beyond these particular spheres. While these particular issues can be addressed with course offerings in the target language, courses offered in English are important to help non‐LOTE students understand the cultural context that we currently inhabit. Beyond these particular issues, departments with LOTE programs must continue to inform potential students and other stakeholders about the benefits of multilingualism.</p> <hd id="AN0186745414-25">CONCLUSION</hd> <p>The present article has examined the factors that predict LOTE enrollment in higher education at a state level. From state to state, enrollment was shown to be highly unequal with some states enrolling LOTEs at rates several times others, when factoring in the size of the higher education student population in each state. The only significant predictor of this variation is the political slant of the states: Republican‐leaning states had less LOTE enrollment than Democratic‐leaning states. Although Republican‐leaning states also had a higher rate of enrollment decline than Democratic‐leaning states, the difference in rate of decline between 2016 and 2021 was not significant. Nevertheless, the findings of this study point to crucial issues in the context of LOTE learning in higher education in the United States, namely neoliberalism and language ideologies, which differentially affected LOTE programs depending on their geographical location within the country. The findings of this study call for future research in this area as well as action on the part of LOTE programs in support of the long‐term vitality of LOTE study within the context of higher education in the United States.</p> <hd id="AN0186745414-26">ACKNOWLEDGMENTS</hd> <p>The author would like to acknowledge the feedback of Kristin Davin and three anonymous referees, who provided crucial comments that resulted in substantial improvements in the article, as well as of Lígia Bezerra, who discussed the content of this manuscript with the author and provided feedback on complete reads of it. The author assumes responsibility for any remaining deficiencies.</p> <hd id="AN0186745414-27">A APPENDIX Enrollment figures for individual states.</hd> <p></p> <p> <ephtml> <table><thead valign="bottom"><tr valign="bottom"><th /><th>2021</th><th>2016</th><th align="left">% Change</th></tr><tr valign="bottom"><th>State</th><th>LOTE</th><th>Higher Ed</th><th><sc>EnrHEd</sc></th><th>LOTE</th><th>Higher Ed</th><th><sc>EnrHEd</sc></th><th>Raw</th><th><sc>EnrHEd</sc></th></tr></thead><tbody valign="top"><tr><td>AL</td><td>11,910</td><td>292,181</td><td>4076.240</td><td>14,618</td><td>304,052</td><td>4807.730</td><td>−18.53%</td><td>−15.21%</td></tr><tr><td>AK</td><td>1585</td><td>22,106</td><td>7169.999</td><td>2056</td><td>28,436</td><td>7230.271</td><td>−22.91%</td><td>−0.83%</td></tr><tr><td>AZ</td><td>28,516</td><td>600,103</td><td>4751.851</td><td>30,053</td><td>608,086</td><td>4942.229</td><td>−5.11%</td><td>−3.85%</td></tr><tr><td>AR</td><td>7228</td><td>149,294</td><td>4841.454</td><td>10,063</td><td>167,235</td><td>6017.281</td><td>−28.17%</td><td>−19.54%</td></tr><tr><td>CA</td><td>157,017</td><td>2,579,991</td><td>6085.951</td><td>177,262</td><td>2,700,445</td><td>6564.177</td><td>−11.42%</td><td>−7.29%</td></tr><tr><td>CO</td><td>18,792</td><td>362,267</td><td>5187.334</td><td>20,585</td><td>352,255</td><td>5,843.778</td><td>−8.71%</td><td>−11.23%</td></tr><tr><td>CT</td><td>13,914</td><td>186,717</td><td>7451.919</td><td>16,818</td><td>198,010</td><td>8493.510</td><td>−17.27%</td><td>−12.26%</td></tr><tr><td>DE</td><td>5668</td><td>58,678</td><td>9659.498</td><td>5507</td><td>61,139</td><td>9007.344</td><td>+2.92%</td><td>+7.24%</td></tr><tr><td>DC</td><td>17,560</td><td>97,481</td><td>18,013.767</td><td>19,390</td><td>93,040</td><td>20,840.499</td><td>−9.44%</td><td>−13.56%</td></tr><tr><td>FL</td><td>44,476</td><td>1,027,331</td><td>4,329.277</td><td>51,940</td><td>1,075,527</td><td>4829.260</td><td>−14.37%</td><td>−10.35%</td></tr><tr><td>GA</td><td>36,913</td><td>547,389</td><td>6743.468</td><td>45,603</td><td>533,073</td><td>8554.738</td><td>−19.06%</td><td>−21.17%</td></tr><tr><td>HI</td><td>7979</td><td>59,693</td><td>13,366.726</td><td>8198</td><td>65,843</td><td>12,450.830</td><td>−2.67%</td><td>+7.36%</td></tr><tr><td>ID</td><td>6989</td><td>122,997</td><td>5682.252</td><td>6435</td><td>123,796</td><td>5198.068</td><td>+8.61%</td><td>+9.31%</td></tr><tr><td>IL</td><td>33,819</td><td>681,988</td><td>4958.885</td><td>38,950</td><td>777,720</td><td>5008.229</td><td>−13.17%</td><td>−0.99%</td></tr><tr><td>IN</td><td>30,280</td><td>410,949</td><td>7368.311</td><td>41,829</td><td>419,284</td><td>9976.293</td><td>−27.61%</td><td>−26.14%</td></tr><tr><td>IA</td><td>13,242</td><td>208,220</td><td>6359.620</td><td>15,707</td><td>266,513</td><td>5893.521</td><td>−15.69%</td><td>+7.91%</td></tr><tr><td>KS</td><td>8751</td><td>193,119</td><td>4531.403</td><td>10,077</td><td>215,832</td><td>4668.909</td><td>−13.16%</td><td>−2.95%</td></tr><tr><td>KY</td><td>13,259</td><td>261,413</td><td>5072.051</td><td>16,860</td><td>255,062</td><td>6610.158</td><td>−21.36%</td><td>−23.27%</td></tr><tr><td>LA</td><td>15,392</td><td>243,507</td><td>6320.968</td><td>16,528</td><td>239,278</td><td>6907.447</td><td>−6.87%</td><td>−8.49%</td></tr><tr><td>ME</td><td>3411</td><td>69,838</td><td>4884.160</td><td>3994</td><td>72,116</td><td>5538.299</td><td>−14.60%</td><td>−11.81%</td></tr><tr><td>MD</td><td>21,506</td><td>348,054</td><td>6,178.926</td><td>24,827</td><td>366,809</td><td>6768.373</td><td>−13.38%</td><td>−8.71%</td></tr><tr><td>MA</td><td>40,462</td><td>473,731</td><td>8541.134</td><td>41,652</td><td>505,722</td><td>8236.146</td><td>−2.86%</td><td>+3.70%</td></tr><tr><td>MI</td><td>33,385</td><td>490,081</td><td>6812.139</td><td>38,890</td><td>583,034</td><td>6670.280</td><td>−14.16%</td><td>+2.13%</td></tr><tr><td>MN</td><td>22,378</td><td>395,705</td><td>5655.223</td><td>28,139</td><td>422,793</td><td>6655.503</td><td>−20.47%</td><td>−15.03%</td></tr><tr><td>MS</td><td>8853</td><td>163,054</td><td>5429.490</td><td>12,413</td><td>172,588</td><td>7192.273</td><td>−28.68%</td><td>−24.51%</td></tr><tr><td>MO</td><td>19,569</td><td>342,618</td><td>5711.609</td><td>32,081</td><td>401,098</td><td>7998.295</td><td>−39.00%</td><td>−28.59%</td></tr><tr><td>MT</td><td>3045</td><td>45,461</td><td>6698.049</td><td>3337</td><td>50,918</td><td>6553.675</td><td>−8.75%</td><td>+2.20%</td></tr><tr><td>NE</td><td>7350</td><td>135,026</td><td>5443.396</td><td>6997</td><td>136,098</td><td>5141.148</td><td>+5.05%</td><td>+5.88%</td></tr><tr><td>NV</td><td>6443</td><td>116,660</td><td>5522.887</td><td>8832</td><td>116,030</td><td>7611.825</td><td>−27.05%</td><td>−27.44%</td></tr><tr><td>NH</td><td>3962</td><td>188,545</td><td>3101.355</td><td>4978</td><td>133,159</td><td>3738.388</td><td>−20.41%</td><td>−43.79%</td></tr><tr><td>NJ</td><td>26,118</td><td>392,343</td><td>6656.930</td><td>33,398</td><td>421,386</td><td>7925.750</td><td>−21.80%</td><td>−16.01%</td></tr><tr><td>NM</td><td>7015</td><td>110,095</td><td>6371.770</td><td>11,547</td><td>134,607</td><td>8578.306</td><td>−39.25%</td><td>−25.72%</td></tr><tr><td>NY</td><td>108,767</td><td>1,182,412</td><td>9,198.740</td><td>135,767</td><td>1,273,634</td><td>10,659.813</td><td>−19.89%</td><td>−13.71%</td></tr><tr><td>NC</td><td>52,319</td><td>557,673</td><td>9381.663</td><td>59,101</td><td>561,415</td><td>10,527.150</td><td>−11.48%</td><td>−10.88%</td></tr><tr><td>ND</td><td>1530</td><td>51,308</td><td>2981.991</td><td>1827</td><td>54,203</td><td>3370.662</td><td>−16.26%</td><td>−11.53%</td></tr><tr><td>OH</td><td>43,332</td><td>654,555</td><td>6620.070</td><td>54,493</td><td>658,043</td><td>8281.070</td><td>−20.48%</td><td>−20.06%</td></tr><tr><td>OK</td><td>11,036</td><td>189,214</td><td>5832.549</td><td>13,253</td><td>208,333</td><td>6361.450</td><td>−16.73%</td><td>−8.31%</td></tr><tr><td>OR</td><td>15,997</td><td>203,759</td><td>7850.942</td><td>20,861</td><td>236,851</td><td>8807.647</td><td>−23.32%</td><td>−10.86%</td></tr><tr><td>PA</td><td>50,077</td><td>667,515</td><td>7502.004</td><td>65,778</td><td>725,682</td><td>9064.301</td><td>−23.87%</td><td>−17.24%</td></tr><tr><td>RI</td><td>6683</td><td>77,087</td><td>8669.425</td><td>9274</td><td>83,348</td><td>11,126.842</td><td>−27.94%</td><td>−22.09%</td></tr><tr><td>SC</td><td>24,238</td><td>235,009</td><td>10,313.648</td><td>29,008</td><td>246,563</td><td>11,764.944</td><td>−16.44%</td><td>−12.34%</td></tr><tr><td>SD</td><td>1593</td><td>50,849</td><td>3132.805</td><td>2330</td><td>53,683</td><td>4340.294</td><td>−31.63%</td><td>−27.82%</td></tr><tr><td>TN</td><td>21,751</td><td>316,270</td><td>6877.352</td><td>24,578</td><td>321,752</td><td>7638.803</td><td>−11.50%</td><td>−9.97%</td></tr><tr><td>TX</td><td>70,987</td><td>1,601,399</td><td>4432.812</td><td>84,615</td><td>1,605,498</td><td>5270.327</td><td>−16.11%</td><td>−15.89%</td></tr><tr><td>UT</td><td>18,078</td><td>395,572</td><td>4570.091</td><td>17,140</td><td>311,450</td><td>5503.291</td><td>+5.47%</td><td>−16.96%</td></tr><tr><td>VT</td><td>4753</td><td>39,646</td><td>11,988.599</td><td>4809</td><td>44,719</td><td>10,753.818</td><td>−1.16%</td><td>+11.48%</td></tr><tr><td>VA</td><td>33,049</td><td>555,755</td><td>5946.685</td><td>40,610</td><td>557,444</td><td>7285.037</td><td>−18.62%</td><td>−18.37%</td></tr><tr><td>WA</td><td>16,068</td><td>334,059</td><td>4809.929</td><td>22,696</td><td>366,547</td><td>6191.839</td><td>−29.20%</td><td>−22.32%</td></tr><tr><td>WV</td><td>4533</td><td>137,742</td><td>3290.935</td><td>6233</td><td>146,608</td><td>4251.473</td><td>−27.27%</td><td>−22.59%</td></tr><tr><td>WI</td><td>19,364</td><td>318,858</td><td>6072.923</td><td>25,457</td><td>341,717</td><td>7449.732</td><td>−23.93%</td><td>−18.48%</td></tr><tr><td>WY</td><td>1620</td><td>30,943</td><td>5235.433</td><td>2305</td><td>33,365</td><td>6908.437</td><td>−29.72%</td><td>−24.22%</td></tr></tbody></table> </ephtml> </p> <hd id="AN0186745414-28">B APPENDIX Predictor coefficients and fits of models in the stepping‐up process in predicting...</hd> <p></p> <p> <ephtml> <table><thead valign="bottom"><tr valign="bottom"><th>Model specification (coefficient)</th><th>AIC</th></tr></thead><tbody valign="top"><tr><td><sc>PolSlant</sc> (−0.17255***)</td><td>894.00</td></tr><tr><td><sc>NonEngInf</sc> (0.06963)</td><td>907.41</td></tr><tr><td><sc>HEdSpend</sc> (−0.01371)</td><td>909.45</td></tr><tr><td><sc>PolSlant</sc> (−0.21484***) + <sc>NonEngInf</sc> (−0.06819!)</td><td>894.77</td></tr><tr><td><sc>PolSlant</sc> (−0.172149***) + <sc>HEdSpend</sc> (−0.005971!)</td><td>896.35</td></tr></tbody></table> </ephtml> </p> <p>4 <bold>Key</bold>: ***: <emph>p</emph> ≤ .001; **: <emph>p</emph> ≤ .01; *: <emph>p</emph> ≤ .05; <emph>p</emph> ≤ .1; !: sign reversed compared to previous round.</p> <p>GRAPH: Supplementary information.</p> <ref id="AN0186745414-29"> <title> Footnotes </title> <blist> <bibl id="bib1" idref="ref30" type="bt">1</bibl> <bibtext> The 47.4% enrollment drop in New Mexico's flagship public university was influential in making it the state that lost the largest percentage of enrollment between 2016 and 2021 (39.2%), according to Lusin et al. ([33]). The other New Mexico institutions reporting in both 2016 and 2021 exhibited a loss of 34.2%, which would still be one of the largest percentage losses of any state during that period.</bibtext> </blist> <blist> <bibl id="bib2" idref="ref31" type="bt">2</bibl> <bibtext> Models were also fit where the response variable was LOTE enrollment normalized by the state's population from the Census Data, as Thompson ([53]) did. These models were basically the same as those with EnrollHEd as the response, except that the population‐normalized LOTE enrollment is predicted by the proportion of the population of typical college age, which is merely an artifact of the fact that states with more people in a certain age range are likely to have more college students and therefore more higher education LOTE enrollment. To simplify the prose, only the EnrollHEd response is addressed in this article.</bibtext> </blist> <blist> <bibl id="bib3" idref="ref32" type="bt">3</bibl> <bibtext> The R code used in to carry out this analysis can be freely downloaded on the IRIS Database; iris‐database.org.</bibtext> </blist> <blist> <bibl id="bib4" idref="ref82" type="bt">4</bibl> <bibtext> A set of models was fit without considering PolSlant; however, neither of the other two predictors was significant, with or without PolSlant.</bibtext> </blist> <blist> <bibl id="bib5" idref="ref83" type="bt">5</bibl> <bibtext> As PolSlant was <emph>z</emph>‐scaled, this means when it was +2.3% in favor of the Republican candidate.</bibtext> </blist> <blist> <bibl id="bib6" idref="ref99" type="bt">6</bibl> <bibtext> For example, at the author's university, while majors in LOTEs have, in general, declined recently, LOTEs are among the most frequently selected minors in the entire university. The unit thereby provides substantial support for students in other fields.</bibtext> </blist> </ref> <ref id="AN0186745414-30"> <title> REFERENCES </title> <blist> <bibtext> Allen, K. (2023). Erosion of state funding for higher education explains most of WVU's budget crisis. West Virginia Center on Budget & Policy https://wvpolicy.org/erosion-of-state-funding-for-higher-education-explains-most-of-wvus-budget-crisis/</bibtext> </blist> <blist> <bibtext> American Council for the Teaching of Foreign Languages (ACTFL). (2024). ACTFL Proficiency Guidelines. https://<ulink href="http://www.actfl.org/uploads/files/general/Resources-Publications/ACTFL%5fProficiency%5fGuidelines%5f2024.pdf">www.actfl.org/uploads/files/general/Resources-Publications/ACTFL%5fProficiency%5fGuidelines%5f2024.pdf</ulink></bibtext> </blist> <blist> <bibtext> Bernstein, K. A., Hellmich, E. A., Katznelson, N., Shin, J., & Vinall, K. (2015). Critical perspectives on neoliberalism in second/foreign language education. Introduction. L2 Journal, 7 (3), 3 – 14.</bibtext> </blist> <blist> <bibtext> Bezerra, K. (2017). Postcards from Rio. Favelas and the contested geographies of citizenship. Fordham University Press.</bibtext> </blist> <blist> <bibtext> Bezerra, L. (2022). Everyday consumption in twenty‐first century Brazilian fiction. Purdue University Press.</bibtext> </blist> <blist> <bibtext> Bezerra, L., Rodrigues, C., Teixeira, A. C., & Wood, N. (2016). Marketing your program. In M. Milleret, & M. Risner (eds.), A handbook for Portuguese instructors in the U.S. (pp. 64 – 95). Boa Vista Press.</bibtext> </blist> <blist> <bibl id="bib7" idref="ref42" type="bt">7</bibl> <bibtext> Brown, W. (2015). Undoing the demos. Neoliberalism's stealth revolution. Zone Books.</bibtext> </blist> <blist> <bibl id="bib8" idref="ref72" type="bt">8</bibl> <bibtext> de Brey, C., Zhang, A., & Duffy, S. (2023). Digest of Educational Statistics, 2021. National Center for Education Statistics. https://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2023009</bibtext> </blist> <blist> <bibl id="bib9" idref="ref95" type="bt">9</bibl> <bibtext> Dean, W. (2023, March 27). Gee addresses West Virginia University budget shortage in State of the University address. The Morgantown News https://<ulink href="http://www.wvnews.com/morgantownnews/news/gee-addresses-west-virginia-university-budget-shortage-in-state-of-the-university-address/article%5fb0d525e2-ccce-11ed-b2ca-d3bdbeb05146.html">www.wvnews.com/morgantownnews/news/gee-addresses-west-virginia-university-budget-shortage-in-state-of-the-university-address/article%5fb0d525e2-ccce-11ed-b2ca-d3bdbeb05146.html</ulink></bibtext> </blist> <blist> <bibtext> Deckard, S., & Shapiro, S. (eds.). (2019). World literature, neoliberalism, and the culture of discontent. Palgrave Macmillan.</bibtext> </blist> <blist> <bibtext> Education Commission of the States. (2021). Vouchers. https://reports.ecs.org/comparisons/vouchers-01</bibtext> </blist> <blist> <bibtext> Fitzsimmons‐Doolan, S. (2011). Language ideology dimensions of politically active Arizona voters: An exploratory study. Language Awareness, 20 (4), 295 – 314.</bibtext> </blist> <blist> <bibtext> Fuller, J. M., & Leeman, J. (2020). Speaking Spanish in the US. The sociopolitics of language (2nd ed.). Multilingual Matters.</bibtext> </blist> <blist> <bibtext> Furman, N., Goldberg, D., & Lusin, N. (2010). Enrollments in languages other than English in United States institutions of higher education, Fall 2009. Modern Language Association. https://<ulink href="http://www.mla.org/content/download/2872/file/2009%5fenrollment%5fsurvey.pdf">www.mla.org/content/download/2872/file/2009%5fenrollment%5fsurvey.pdf</ulink></bibtext> </blist> <blist> <bibtext> Gerring, J. (1997). Ideology: A definitional analysis. Political Research Quarterly, 50 (4), 957 – 994. https://doi.org/10.2307/448995</bibtext> </blist> <blist> <bibtext> Gerstle, G. (2018). The rise and fall(?) of America's neoliberal order. Transactions of the Royal Historical Society, 28, 241 – 264. https://doi.org/10.1017/S0080440118000129</bibtext> </blist> <blist> <bibtext> Glynn, C., & Wassell, B. (2018). Who gets to play? Issues of access and social justice in world language study in the U.S. In P. Garrett‐Rucks & S. M. Johnson (eds.), Dimension 2018 (pp. 18 – 32). Southern Conference on Language Teaching.</bibtext> </blist> <blist> <bibtext> Goldberg, D., Looney, D., & Lusin, N. (2015). Enrollments in languages other than English in United States institutions of higher education, Fall 2013. Modern Language Association. https://<ulink href="http://www.mla.org/content/download/31180/file/EMB%5fenrllmnts%5fnonEngl%5f2013.pdf">www.mla.org/content/download/31180/file/EMB%5fenrllmnts%5fnonEngl%5f2013.pdf</ulink></bibtext> </blist> <blist> <bibtext> Gutiérrez, J. R. (1997). Teaching Spanish as a heritage language. A case for language awareness. ADFL Bulletin, 29 (1), 33 – 36.</bibtext> </blist> <blist> <bibtext> Harvey, B. C. (2021). Teetering on the demographic cliff, part 1: Preparing now for the challenging times ahead. Planning for Higher Education, 49 (4), 1 – 13.</bibtext> </blist> <blist> <bibtext> Hayek, F. A. (2011 [1960]). The constitution of liberty. The definitive edition. The University of Chicago Press.</bibtext> </blist> <blist> <bibtext> Holguín Mendoza, C. (2018). Critical language awareness (CLA) for Spanish heritage language programs. implementing a complete curriculum. International Multilingual Research Journal, 12 (2), 65 – 79.</bibtext> </blist> <blist> <bibtext> Jones, S. (2023). What is a university without liberal arts? Intelligencer https://nymag.com/intelligencer/2023/08/what-is-a-university-without-liberal-arts.html</bibtext> </blist> <blist> <bibtext> Karmanov, F. (2023). The increasing diversity of world language study in the United States, 1958‐2016. In E. Heidrich Uebel, F. A. Kronenberg, & S. Sterling (eds.), Language program vitality in the United States (pp. 33 – 39). Springer. https://doi.org/10.1007/978-3-031-43654-3_4</bibtext> </blist> <blist> <bibtext> Leeman, J. (2018a). It's all about English: The interplay of monolingual ideologies, language policies and the US Census Bureau's statistics on multilingualism. The International Journal of the Sociology of Language, 252, 21 – 43.</bibtext> </blist> <blist> <bibtext> Leeman, J. (2018b). Critical Language Awareness and Spanish as a heritage language. Challenging the linguistic subordination of US Latinxs. In K. Potowski (ed.), The Routledge Handbook of Spanish as a heritage language (pp. 345 – 358). Routledge.</bibtext> </blist> <blist> <bibtext> Lehnen, L. (2013). Citizenship and crises in contemporary Brazilian literature. Palgrave Macmillan.</bibtext> </blist> <blist> <bibtext> Lippi‐Green, R. (1997). English with an accent. Language, ideology and discrimination in the United States. Routledge.</bibtext> </blist> <blist> <bibtext> Looney, D., & Lusin, N. (2019). Enrollments in languages other than English in United States institutions of higher education, Summer 2016 and Fall 2016: Final report. Modern Language Association. https://<ulink href="http://www.mla.org/content/download/110154/file/2016-Enrollments-Final-Report.pdf">www.mla.org/content/download/110154/file/2016-Enrollments-Final-Report.pdf</ulink></bibtext> </blist> <blist> <bibtext> Loza, S., & Beaudrie, S. (eds.). (2021). Heritage language teaching. Critical language awareness perspectives for research and pedagogy. Routledge.</bibtext> </blist> <blist> <bibtext> Lusin, N. (2012). The MLA survey of postsecondary entrance and degree requirements for languages other than English, 2009–10. Modern Language Association. https://<ulink href="http://www.mla.org/content/download/3316/file/requirements%5fsurvey%5f200910.pdf">www.mla.org/content/download/3316/file/requirements%5fsurvey%5f200910.pdf</ulink></bibtext> </blist> <blist> <bibtext> Lusin, N. (2023). MLA language enrollment trends. In E. Heidrich Uebel, F. A. Kronenberg, & S. Sterling (eds.), Language program vitality in the United States (pp. 9 – 17). Springer. https://doi.org/10.1007/978-3-031-43654-3_2</bibtext> </blist> <blist> <bibtext> Lusin, N., Peterson, T., Sulewski, C., & Zafer, R. (2023). Enrollments in languages other than English in US institutions of higher education, Fall 2021. Modern Language Association. https://<ulink href="http://www.mla.org/content/download/191324/file/Enrollments-in-Languages-Other-Than-English-in-US-Institutions-of-Higher-Education-Fall-2021.pdf">www.mla.org/content/download/191324/file/Enrollments-in-Languages-Other-Than-English-in-US-Institutions-of-Higher-Education-Fall-2021.pdf</ulink></bibtext> </blist> <blist> <bibtext> Martínez, G. (2003). Classroom based dialect awareness in heritage language instruction: A critical applied linguistic approach. Heritage Language Journal, 1 (1), 44 – 57.</bibtext> </blist> <blist> <bibtext> Mays, M. (2014, October 6). WVU's Gee sets goal to enroll 40,000. Charleston Gazette‐Mail https://<ulink href="http://www.wvgazettemail.com/news/wvu-s-gee-sets-goal-to-enroll-40-000/article%5f85f0a118-740c-5fe8-ae9f-7b21c00e28e8.html">www.wvgazettemail.com/news/wvu-s-gee-sets-goal-to-enroll-40-000/article%5f85f0a118-740c-5fe8-ae9f-7b21c00e28e8.html</ulink></bibtext> </blist> <blist> <bibtext> Mintz, B. (2021). Neoliberalism and the crisis in higher education: The cost of ideology. The American Journal of Economics and Sociology, 80, 79 – 112. https://doi.org/10.1111/ajes.12370</bibtext> </blist> <blist> <bibtext> Morgan, W. J., & Thompson, A. S. (2023). "My friend Travis says". A look at enrollment trends in language classes in the United States. Foreign Language Annals, 56 (2), 259 – 279. https://doi.org/10.1111/flan.12683</bibtext> </blist> <blist> <bibtext> Mullen, A. L. (2010). Degrees of inequality. Culture, class, and gender in American higher education. The Johns Hopkins University Press.</bibtext> </blist> <blist> <bibtext> Murphy, D., Sarac, M., & Sedivy, S. (2022). Why U.S. undergraduate students are (not) studying languages other than English. Second Language Research and Practice, 3 (1), 1 – 33.</bibtext> </blist> <blist> <bibtext> Nagano, T. (2023). World language enrollment at community colleges in the United States between 1960 and 2010. In E., Heidrich Uebel, F. A. Kronenberg, & S. Sterling (eds.), Language program vitality in the United States (pp. 41 – 55). Springer. https://doi.org/10.1007/978-3-031-43654-3_5</bibtext> </blist> <blist> <bibtext> National Standards Collaborative Board. (2014). World‐readiness standards for learning languages. American Council for the Teaching of Foreign Languages.</bibtext> </blist> <blist> <bibtext> O'Rourke, P., Zhou, Q., & Rottman, I. (2016). Prioritization of K‐12 world language education in the United States. Sate requirements for high school graduation. Foreign Language Annals, 49 (4), 789 – 800. https://doi.org/10.1111/flan.12232</bibtext> </blist> <blist> <bibtext> Public Records Branch. (2021). Official 2020 presidential general election results. Federal Elections Commission. https://<ulink href="http://www.fec.gov/resources/cms-content/documents/2020presgeresults.pdf">www.fec.gov/resources/cms-content/documents/2020presgeresults.pdf</ulink></bibtext> </blist> <blist> <bibtext> R Core Team. (2023). R. A language and environment for statistical computing. R Foundation for Statistical Computing. https://<ulink href="http://www.R-project.org/">www.R-project.org/</ulink></bibtext> </blist> <blist> <bibtext> Saunders, D. B. (2010). Neoliberal ideology and public higher education in the United States. Journal for Critical Education Policy Studies, 8 (1), 41 – 77.</bibtext> </blist> <blist> <bibtext> Schieffelin, B. B., Woolard, K. A., & Kroskrity, P. V. (eds.) (1998). Language ideologies. Practice and theory. Oxford University Press.</bibtext> </blist> <blist> <bibtext> Schloerke, B., Cook, D., Larmarange, J., Briatte, F., Marbach, M., Thoen, E., Elberg, A., & Crowley, J. (2023). GGally: Extension to 'ggplot2'. R package version 2.2.0. https://CRAN.R-project.org/package=GGally</bibtext> </blist> <blist> <bibtext> Slade, D. (2023). How West Virginia lawmakers have indirectly influenced budget cuts at WVU. Mountain State Spotlight (11 September 2023). Accessed on January 26, 2024 https://mountainstatespotlight.org/2023/09/11/wvu-budget-state-funding-law/</bibtext> </blist> <blist> <bibtext> Snyder, T., de Brey, C., & Dillow, S. (2019). Digest of educational statistics, 2017. National Center for Education Statistics. https://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2018070</bibtext> </blist> <blist> <bibtext> Spelman College. (2024). Course Sequence Booklet 2024. https://<ulink href="http://www.spelman.edu/%5f1%5fDocs-and-Files/spelman-course-sequence-booklet-2024.pdf">www.spelman.edu/%5f1%5fDocs-and-Files/spelman-course-sequence-booklet-2024.pdf</ulink></bibtext> </blist> <blist> <bibtext> Steger, M. B., & Roy, R. K. (2010). Neoliberalism. A very short introduction. Oxford University Press.</bibtext> </blist> <blist> <bibtext> Thompson, A. S. (2017). Language learning motivation in the United States. An examination of language choice and multilingualism. The Modern Language Journal, 101 (3), 483 – 500.</bibtext> </blist> <blist> <bibtext> Thompson, A. S. (2021). LOTEs in U.S. universities. Benefits, trends, motivations, and opportunities. In U. Lanvers, A. S. Thompson, & M. East (eds.), Language learning in Anglophone countries. Palgrave Macmillan. https://doi.org/10.1007/978-3-030-56654-8_10</bibtext> </blist> <blist> <bibtext> Thompson, A. S., & Morgan, W. J. (2022). "If I own a business one day". The value of language study for future careers. ADFL Bulletin, 47 (2), 71 – 81.</bibtext> </blist> <blist> <bibtext> Tripiccione, L. (2023). Beyond the crisis: Tools for analyzing historical enrollments in languages other than English. In E. Heidrich Uebel, F. A. Kronenberg, & S. Sterling (eds.), Language program vitality in the United States (pp. 19 – 31). Springer. https://doi.org/10.1007/978-3-031-43654-3_3</bibtext> </blist> <blist> <bibtext> U.S. Census Bureau. (n.d. a). Census Bureau tables. U.S. Department of Commerce. https://data.census.gov/table</bibtext> </blist> <blist> <bibtext> U.S. Census Bureau. (n.d. b). Annual survey of state and local government finances. Compiled by the Urban‐Brookings Tax Policy Center. Urban‐Brookings Tax Policy Center. https://<ulink href="http://www.taxpolicycenter.org/statistics/state-and-local-general-expenditures-capita">www.taxpolicycenter.org/statistics/state-and-local-general-expenditures-capita</ulink></bibtext> </blist> <blist> <bibtext> University of New Mexico. (2011). College of Arts and Sciences. In 2011–2012 Catalog. https://catalog.unm.edu/catalogs/2011-2012/colleges/arts-sciences/index.html</bibtext> </blist> <blist> <bibtext> University of New Mexico. (2012). College of Arts and Sciences. In 2012–2013 Catalog. https://catalog.unm.edu/catalogs/2012-2013/colleges/arts-sciences/index.html</bibtext> </blist> <blist> <bibtext> University of New Mexico. (2023). General education requirements. In 2023–2024 Catalog. https://catalog.unm.edu/catalogs/2023-2024/#/programs/Sk0wRqKHs?bc=true&bcCurrent=General%20Education%20Requirements&bcGroup=General%20Education&bcItemType=programs</bibtext> </blist> <blist> <bibtext> University of Washington. (n.d.). Requirements by college and school. https://advising.uw.edu/degree-overview/general-education/requirements-by-college-and-school/</bibtext> </blist> <blist> <bibtext> Van Gorp, K., Uebel, E. H., Kronenberg, F. A., & Murphy, D. (2024). How important is studying languages for undergraduate students, and why (not) study languages? Foreign Language Annals, 57 (4), 900 – 920.</bibtext> </blist> <blist> <bibtext> Venables, W. N., & Ripley, B. D. (2002). Modern applied statistics with S (4th ed.). Springer.</bibtext> </blist> <blist> <bibtext> Villa, D. J. (2002). The sanitizing of U.S. Spanish in academia. Foreign Language Annals, 35, 222 – 230.</bibtext> </blist> <blist> <bibtext> Wickham, H. (2016). ggplot2: Elegant graphics for data analysis. Springer‐Verlag.</bibtext> </blist> </ref> <aug> <p>By Michael Gradoville</p> <p>Reported by Author</p> </aug> <nolink nlid="nl1" bibid="bib38" firstref="ref1"></nolink> <nolink nlid="nl2" bibid="bib17" firstref="ref2"></nolink> <nolink nlid="nl3" bibid="bib33" firstref="ref3"></nolink> <nolink nlid="nl4" bibid="bib20" firstref="ref5"></nolink> <nolink nlid="nl5" bibid="bib55" firstref="ref6"></nolink> <nolink nlid="nl6" bibid="bib14" firstref="ref7"></nolink> <nolink nlid="nl7" bibid="bib18" firstref="ref8"></nolink> <nolink nlid="nl8" bibid="bib29" firstref="ref9"></nolink> <nolink nlid="nl9" bibid="bib32" firstref="ref11"></nolink> <nolink nlid="nl10" bibid="bib53" firstref="ref13"></nolink> <nolink nlid="nl11" bibid="bib24" firstref="ref16"></nolink> <nolink nlid="nl12" bibid="bib40" firstref="ref17"></nolink> <nolink nlid="nl13" bibid="bib42" firstref="ref33"></nolink> <nolink nlid="nl14" bibid="bib61" firstref="ref35"></nolink> <nolink nlid="nl15" bibid="bib39" firstref="ref36"></nolink> <nolink nlid="nl16" bibid="bib62" firstref="ref37"></nolink> <nolink nlid="nl17" bibid="bib31" firstref="ref38"></nolink> <nolink nlid="nl18" bibid="bib50" firstref="ref39"></nolink> <nolink nlid="nl19" bibid="bib60" firstref="ref40"></nolink> <nolink nlid="nl20" bibid="bib58" firstref="ref43"></nolink> <nolink nlid="nl21" bibid="bib59" firstref="ref44"></nolink> <nolink nlid="nl22" bibid="bib52" firstref="ref46"></nolink> <nolink nlid="nl23" bibid="bib54" firstref="ref48"></nolink> <nolink nlid="nl24" bibid="bib15" firstref="ref52"></nolink> <nolink nlid="nl25" bibid="bib45" firstref="ref53"></nolink> <nolink nlid="nl26" bibid="bib13" firstref="ref54"></nolink> <nolink nlid="nl27" bibid="bib51" firstref="ref57"></nolink> <nolink nlid="nl28" bibid="bib21" firstref="ref61"></nolink> <nolink nlid="nl29" bibid="bib16" firstref="ref62"></nolink> <nolink nlid="nl30" bibid="bib11" firstref="ref63"></nolink> <nolink nlid="nl31" bibid="bib12" firstref="ref65"></nolink> <nolink nlid="nl32" bibid="bib56" firstref="ref71"></nolink> <nolink nlid="nl33" bibid="bib57" firstref="ref73"></nolink> <nolink nlid="nl34" bibid="bib43" firstref="ref74"></nolink> <nolink nlid="nl35" bibid="bib25" firstref="ref76"></nolink> <nolink nlid="nl36" bibid="bib44" firstref="ref77"></nolink> <nolink nlid="nl37" bibid="bib65" firstref="ref79"></nolink> <nolink nlid="nl38" bibid="bib47" firstref="ref80"></nolink> <nolink nlid="nl39" bibid="bib63" firstref="ref81"></nolink> <nolink nlid="nl40" bibid="bib49" firstref="ref86"></nolink> <nolink nlid="nl41" bibid="bib35" firstref="ref93"></nolink> <nolink nlid="nl42" bibid="bib48" firstref="ref94"></nolink> <nolink nlid="nl43" bibid="bib36" firstref="ref96"></nolink> <nolink nlid="nl44" bibid="bib23" firstref="ref97"></nolink> <nolink nlid="nl45" bibid="bib37" firstref="ref101"></nolink> <nolink nlid="nl46" bibid="bib41" firstref="ref120"></nolink> <nolink nlid="nl47" bibid="bib10" firstref="ref125"></nolink> <nolink nlid="nl48" bibid="bib27" firstref="ref126"></nolink> <nolink nlid="nl49" bibid="bib28" firstref="ref127"></nolink> <nolink nlid="nl50" bibid="bib46" firstref="ref128"></nolink> <nolink nlid="nl51" bibid="bib34" firstref="ref129"></nolink> <nolink nlid="nl52" bibid="bib64" firstref="ref130"></nolink> <nolink nlid="nl53" bibid="bib19" firstref="ref131"></nolink> <nolink nlid="nl54" bibid="bib22" firstref="ref132"></nolink> <nolink nlid="nl55" bibid="bib26" firstref="ref133"></nolink> <nolink nlid="nl56" bibid="bib30" firstref="ref134"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1477335
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: The Politics of Language Enrollment in US Higher Education
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Michael+Gradoville%22">Michael Gradoville</searchLink> (ORCID <externalLink term="http://orcid.org/0000-0002-8462-0009">0000-0002-8462-0009</externalLink>)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Foreign+Language+Annals%22"><i>Foreign Language Annals</i></searchLink>. 2025 58(2):367-391.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 25
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2025
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Evaluative
– Name: Audience
  Label: Education Level
  Group: Audnce
  Data: <searchLink fieldCode="EL" term="%22Higher+Education%22">Higher Education</searchLink><br /><searchLink fieldCode="EL" term="%22Postsecondary+Education%22">Postsecondary Education</searchLink>
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Language+Enrollment%22">Language Enrollment</searchLink><br /><searchLink fieldCode="DE" term="%22Second+Language+Learning%22">Second Language Learning</searchLink><br /><searchLink fieldCode="DE" term="%22Second+Language+Instruction%22">Second Language Instruction</searchLink><br /><searchLink fieldCode="DE" term="%22Voting%22">Voting</searchLink><br /><searchLink fieldCode="DE" term="%22Elections%22">Elections</searchLink><br /><searchLink fieldCode="DE" term="%22Presidents%22">Presidents</searchLink><br /><searchLink fieldCode="DE" term="%22Political+Attitudes%22">Political Attitudes</searchLink><br /><searchLink fieldCode="DE" term="%22Predictor+Variables%22">Predictor Variables</searchLink><br /><searchLink fieldCode="DE" term="%22Correlation%22">Correlation</searchLink><br /><searchLink fieldCode="DE" term="%22College+Second+Language+Programs%22">College Second Language Programs</searchLink><br /><searchLink fieldCode="DE" term="%22State+Norms%22">State Norms</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1111/flan.12798
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 0015-718X<br />1944-9720
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: This article examines enrollments in languages other than English in United States higher education from the perspective of geographical distribution. While the overall decline in language enrollments is well known, enrollments are also very unequal across states when accounting for population. By cross-referencing MLA language enrollment data with demographic data, numerous predictors were tested for their efficacy in explaining inequality in Fall 2021 language enrollments. The only predictor found to affect language enrollments in a state was the state's political leanings as defined by voting in the 2020 U.S. presidential election: the more strongly a state voted for the Republican candidate, the lower its rate of language enrollment. An analysis of the changes in enrollment between 2016 and 2021 also showed a non-significant trend whereby states that voted more strongly for the Republican candidate tended to have more enrollment decline than those that voted for the Democratic candidate.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2025
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1477335
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1477335
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1111/flan.12798
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 25
        StartPage: 367
    Subjects:
      – SubjectFull: Language Enrollment
        Type: general
      – SubjectFull: Second Language Learning
        Type: general
      – SubjectFull: Second Language Instruction
        Type: general
      – SubjectFull: Voting
        Type: general
      – SubjectFull: Elections
        Type: general
      – SubjectFull: Presidents
        Type: general
      – SubjectFull: Political Attitudes
        Type: general
      – SubjectFull: Predictor Variables
        Type: general
      – SubjectFull: Correlation
        Type: general
      – SubjectFull: College Second Language Programs
        Type: general
      – SubjectFull: State Norms
        Type: general
    Titles:
      – TitleFull: The Politics of Language Enrollment in US Higher Education
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Michael Gradoville
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 07
              Type: published
              Y: 2025
          Identifiers:
            – Type: issn-print
              Value: 0015-718X
            – Type: issn-electronic
              Value: 1944-9720
          Numbering:
            – Type: volume
              Value: 58
            – Type: issue
              Value: 2
          Titles:
            – TitleFull: Foreign Language Annals
              Type: main
ResultId 1