Shift Still Happens: Spanish Language Maintenance in the Face of Growth and Change in the Western United States

Abstract

The recent release of 2020 U.S. Census data reflects the continued growth of the Hispanic/Latino population over the last four decades. The Hispanic/Latino population has increased by a factor of 3.25 since 1980, with nearly one in five inhabitants of the United States identifying as Hispanic or Latino. With these demographic changes, language maintenance figures have shown significant change as well. In every state, a language shift is evident, as all of the western states have lower measures of Spanish language maintenance among the Hispanic population than they did a generation ago. So, while Spanish language use is growing with regard to overall numbers in most of the Western United States, language shift is still a reality among the Latino population. Social variables such as education, income, and employment also show a different relationship with the Spanish language than they did 40 years ago. While there were strong negative correlations between Spanish language use and these variables in the previous generation, many (but not all) of these correlations have weakened, many to the point of no statistical significance in the data from 2020.

1. Introduction

Over the last 40 years, the Hispanic/Latino population in the United States has increased significantly, both in overall numbers and as a percentage of the total United States population. A population that measured a total of 14.6 million in 1980 has effectively increased by that amount every decade since, now more than 62 million strong as of the 2020 Census, jumping from 6.4% (in 1980) to 18.7% (in 2020) of the nation’s population. Hispanics as a whole comprise the largest minority group in the United States, and the proportion relative to the total population continues to grow.

The Spanish language community in the United States has long felt the effects of intergenerational language shift, wherein the language shifts from Spanish to English over the course of three generations. Additionally, Hudson et al. (1995) found that speaking Spanish in the southwest United States has historically come at a cost, wherein “educational and economic success in the Spanish-origin population are purchased at the expense of Spanish language maintenance in the home” (p. 179). Their data for such a conclusion came from the 1980 United States Census. Since then, the rapid and sizeable growth of the Spanish-speaking population has encouraged further studies on Spanish language maintenance and the relationship between Spanish as a home language and socioeconomic variables such as income, education, and professional experience. McCullough and Jenkins (2005) found that, in Colorado, those relationships were not as strong in 2000 as they had been in 1980 and 1990. They also identified recent immigration in a community as a strong predictor of language maintenance. Jenkins (2009b) examined the same southwestern states as Hudson et al. (1995) and found that the relationships between Spanish language use and social variables had weakened almost categorically. Then, 2010 data showed that those relationships had weakened even further, with zero significance in the relationship in categories like employment and poverty (Jenkins 2013).

What was once a geographically discrete language community in the Southwestern United States has expanded not only numerically but also geographically. Spanish-speaking communities that were once predominantly confined to areas close to the Mexican border have expanded outward such that the Mexican community has strong representation throughout the United States. Every contiguous state outside of the Eastern Time Zone (and 40 total states) counts Mexicans as their largest Latino population, and the percentage of Hispanics in each state has increased at a corresponding pace. In 1980, five states west of the Mississippi (the four border states plus Colorado) had Spanish-speaking (or even Hispanic) counts greater than 10%, which merited a relatively closed focus on that area. That figure has doubled, with more states, like Kansas, approaching that benchmark. Illinois could also make a strong case for inclusion in studies on Southwest Spanish, even though it is nowhere near the Southwest. It boasts the fourth-largest population of Mexicans in the country; the Midwest as a whole merits further sociolinguistic study as it relates to Mexican-origin populations (cf. Potowski 2020). The current study will focus on the 10 contiguous western states with Hispanic populations greater than 10% as of 2020, serving as a comparison of “old” versus “new” as well as a contrast between “south” and “north”. We will collectively refer to Arizona, California, Colorado, New Mexico, and Texas as the “Southwest”, and Idaho, Nevada, Oregon, Utah, and Washington as the “Northwest”. We recognize that Nevada and Utah are not Northwestern states as traditionally recognized (i.e., the “Pacific Northwest”), but we use the term here out of convenience to distinguish those five states from the traditional, Spanish-speaking Southwest.

Rivera-Mills (2010) noted a certain mystique among policymakers once round figures, like 10%, are achieved in a minority population. Ballot initiatives regarding language and educational policy are more common in those states that have reached that threshold. Kusnet (1992) also points out that “the five percent level is an important political and social threshold because a 1975 Voting Rights Act amendment states that jurisdictions must provide bilingual ballots and bilingual election materials when 5 percent of its voting age population belongs to a single-language minority” (p. 15). Five percent has also been used in previous population studies to delimit the Spanish-speaking Southwest (Nostrand 1970; Haverluk 1997).

The current study engages two questions as they relate to Spanish language use in the Western United States. First, has the large population growth and increased density of the Spanish-speaking population impacted rates of Spanish-language maintenance among Hispanics in the Western United States? Second, what is the current relationship between the use of Spanish as a home language and socioeconomic indicators? What is the cost of Spanish language loyalty in 2020?

This study will employ data from the United States Census for the last five decennial census periods, from 1980 through 2020. It bears mentioning that, due to the pandemic, the 2020 Census faced challenges of both practicality and policymaking. As a result, the census post-enumeration survey showed that “the Hispanic or Latino population had a statistically significant undercount rate of 4.99%. This is statistically different from a 1.54% undercount in 2010” (U.S. Census Bureau 2022b). This undercount is important, yet the scope of the Census is such that, even with this challenge, general trends can still be observed with the available data. Additionally, the annual American Community Survey helps to offset some of the Census limitations to provide a more accurate picture of the Hispanic and Spanish-speaking communities in the country. We will employ data from both of these surveys, with the American Community Survey featuring more prominently in more recent data.

2. Demographic Trends

Both the number of Hispanics and the proportion of Hispanics within the U.S. population have steadily increased since 1980, with the count increasing by 325% and the proportion of the population more than tripling in that time period. Though the proportions grew the most in the Hispanic immigrant boom between 1990 and 2000, the most recent decade from 2010 to 2020 showed the greatest total growth, increasing by 14.3 million in that ten-year period alone (

Table 1. Hispanic or Latino population, by total count and percentage of total population (U.S. Census Bureau 1983, 1992, 2000, 2010, 2020e).

	1980	1990	2000	2010	2020	Δ 1980–2020
USA	14,608,673 6.4%	22,354,059 9.0%	35,305,818 12.5%	47,727,533 15.7%	62,080,044 18.7%	325%

).

Growth rates have been greater in areas that have historically not had large Hispanic populations, which is true at both the county level and the state level. States with the largest percentage growth in the Hispanic population generally can be found outside of the Southwestern United States. Conversely, areas with negative growth in the Hispanic population are generally those where Hispanic densities are highest, in areas like New Mexico, Southern Colorado, and West Texas, as seen in Figure 1, which shows growth rates from 2010 to 2020. Note that, while high growth and negative growth areas are easily discernible, the map is incomplete in proportional reporting, as areas with fewer than 1000 Hispanics are left unshaded.

Among western states, we see a wide range of growth, as reflected in

Table 2. Hispanic or Latino population (raw count and percentage of the state total) for 10 western states, 1980–2020 (U.S. Census Bureau 1982, 1992, 2000, 2010, 2020e).

	1980	1990	2000	2010	2020	Δ 1980–2020
California	4,544,331 19.2%	7,687,938 25.8%	10,966,556 32.4%	14,013,719 37.6%	15,579,652 39.4%	243%
Texas	2,985,824 21.0%	4,339,905 25.5%	6,669,666 32.0%	9,460,921 37.6%	11,441,717 39.3%	283%
Arizona	440,701 16.2%	688,338 18.8%	1,295,617 25.3%	1,895,149 29.6%	2,192,253 30.7%	397%
Colorado	339,717 11.8%	424,302 12.9%	735,601 17.1%	1,038,687 20.7%	1,263,390 21.9%	272%
Washington	120,016 2.9%	214,570 4.4%	441,509 7.5%	755,790 11.2%	1,059,213 13.7%	783%
New Mexico	477,222 36.6%	579,224 38.2%	765,386 42.1%	953,403 46.3%	1,010,811 47.7%	112%
Nevada	53,879 6.7%	124,419 10.4%	393,970 19.7%	716,501 26.5%	890,257 28.7%	1552%
Oregon	65,847 2.5%	112,707 4.0%	275,314 8.0%	450,062 11.7%	588,757 13.9%	794%
Utah	60,302 4.1%	83,097 4.8%	201,559 9.0%	358,340 13.0%	492,912 15.1%	717%
Idaho	36,615 3.9%	52,927 5.3%	101,690 7.9%	175,901 11.2%	239,407 13.0%	554%
Total	9,124,454				34,758,369	281%

. At the high end of the growth spectrum, Nevada experienced 1552% growth in the Hispanic population in the 40-year period. Northwestern states with newer Hispanic populations, like Oregon, Utah, and Washington, saw growth rates greater than 700% in that period. States with more established Hispanic populations, like California, Colorado, and Texas, while seeing significant total growth, experienced more modest growth rates as a percentage of the total population, each ranging between 200 and 300% in that time period. By virtue of the size of their total populations, California and Texas account for a large percentage of the total growth in the West, and that same enormity of their total populations explains the less drastic growth percentages. New Mexico is the outlier on the low end of the spectrum, with little more than a doubling of the Hispanic population in the four-decade term.

3. Spanish Language Use

Spanish language growth in the West has effectively paralleled the growth of the Hispanic population, with a few notable differences in the data. In

Table 3. Spanish-speaking population (raw count and percentage of the state total) for 10 western states, 1980–2020 (Highest percentage for each state is in bold).

	1980	1990	2000	2010	2020	Δ 1980–2020
California	3,132,690 14.3%	5,475,367 20.0%	8,105,505 25.8%	9,706,949 28.5%	10,462,968 28.3%	234%
Texas	2,484,188 19.0%	3,442,418 22.1%	5,195,182 27.0%	6,547,178 29.2%	7,666,020 28.8%	209%
Arizona	331,038 13.2%	478,157 14.2%	927,395 19.5%	1,199,689 20.7%	1,358,980 20.2%	311%
Colorado	179,607 6.7%	203,877 6.7%	421,670 10.5%	542,257 11.9%	602,273 11.2%	235%
Washington	78,489 2.1%	143,594 3.2%	321,490 5.8%	477,752 7.8%	602,058 8.5%	667%
Nevada	34,813 4.7%	85,474 7.7%	299,947 16.2%	480,267 19.6%	593,610 20.9%	1605%
New Mexico	352,488 29.7%	388,123 27.9%	485,681 28.7%	532,568 28.5%	514,071 26.0%	46%
Oregon	41,806 1.7%	83,072 3.1%	217,614 6.8%	302,701 8.6%	349,549 8.9%	736%
Utah	36,056 2.8%	51,945 3.3%	150,244 7.4%	225,418 9.4%	297,926 10.3%	726%
Idaho	27,118 3.2%	37,078 4.0%	80,241 6.7%	106,792 7.6%	131,145 8.0%	384%
Total	6,698,293				22,578,600	237%

, we can see the general upward trend in Spanish language density for the Southwest. The 10-state region as a whole grew by a factor of 2.37 in the 40-year period, which is slightly lower than the 265% growth rate for the country as a whole. We can see four different groupings in these data, as reflected in Figure 2. First, we see a clear clustering among the newer Spanish-speaking states of Idaho, Oregon, Utah, and Washington, which all increased from relatively small densities in 1980 (1.7–3.2%) to much more sizeable densities in 2020 (8.0–10.3%). The total growth rate of these four states is remarkable, with a growth range of 667% to 736% in Oregon, Utah, and Washington and a more modest, but still noteworthy, rate of 384% in Idaho. Each of these states achieved its highest density to date in the 2020 Census.

Second, we see the traditional Spanish-speaking areas of Arizona, California, Colorado, and Texas growing rather significantly since 1980, peaking in 2010, then declining in density slightly in the 2020 data. While California and Texas started with much larger populations than any of the other states, and their gross numerical gains far outstrip the total growth for all of the other states combined, their relative growth is lower, with 234% and 209% growth, respectively. Arizona’s 311% growth rate is comparable to these, and Colorado’s growth rate, at 235%, is almost identical to that of California.

The two outliers in the data reflect their own unique trends. First, Nevada’s Spanish-speaking population grew by an astounding 1605% in the 40-year period. In 1980, Nevada counted fewer Spanish speakers than any western state outside of Idaho. By 1990, it had outgained Oregon and Utah, and by 2020, it had surpassed New Mexico in the total number of Spanish speakers and now rivals Colorado and Washington in its count, even though those two states have total populations that are nearly or more than double that of Nevada. It ranks fourth behind only California, Texas, and New Mexico in Spanish language density, with better than one in five inhabitants of the state claiming Spanish as a home language.

New Mexico represents the final trend, and certainly the least optimistic for Spanish language maintenance. Even though the total count of Hispanics in the state has risen every decade since 1980, for a total growth rate of 112%—considerably lower than that of any other state in the region—the language density growth rate is a relatively paltry 46%. In 1980, New Mexico had the 3rd-largest Spanish-language count of the 10 states; now, it ranks 7th. By comparing the Hispanic and Spanish-language growth figures, the conclusion is that Spanish language density in New Mexico has steadily declined since 1980. Bills and Vigil (2008, p. 240), using data from a generation ago, pointed out that “the Traditional New Mexican Spanish dialect [in the northern half of the state] is dying out,” due to contact with both English and Mexican Spanish, and the present data bear that assertion out. In Figure 3, we can see that, while the Hispanic/Latino population has increased as a percentage of the total population every decade, the density of Spanish speakers in the state has steadily declined, showing an expanding language maintenance gap among Latinos.

4. Language Maintenance and Shift

As Spanish language densities and counts change, an obvious question arises as to how such changes impact Spanish language maintenance across generations. Additionally, these changes call to question whether Spanish language use continues to have the same negative impact on socioeconomic variables as first reported by Hudson et al. (1995) and Bills et al. (1995). Using data from 1980, they demonstrated a “linguistic cost of social, political and economic integration into mainstream society” (Hudson et al. 1995, p. 182). Subsequent studies (McCullough and Jenkins 2005; Jenkins 2009a, 2009b) showed a weakening of those correlations that corresponded with the immigrant population boom of the 1990s and 2000s. As growth has slowed since then, the question comes up as to whether correlations between language use and social variables have had a corresponding shift or whether they have continued to weaken with the progression of time.

Hudson et al. (1995) used various quantifiable demographic measures to identify extralinguistic variables that affect, and are affected by, language maintenance and shift in the Southwest. We will employ their measures of “Count”, “Density”, and “Language Loyalty” on a per-county basis in the ten-state region.

Count refers to the raw, absolute number of Spanish language claimants in a county. This census measurement applies to the total population five years of age and older. This measurement does not take into account the proportion of the Spanish-speaking population relative to the total population but rather gives an indication of community size as a whole.

Density, on the other hand, is the percentage of the total population (five years of age or older) who claim Spanish as a home language. This relative frequency contrasts markedly with the absolute frequency of Count, often providing a better picture of those communities that are regarded as “Spanish speaking”, irrespective of their total size. Density can be high in small counties and likewise low in large counties. Density tends to be highest where Hispanic populations are more heavily concentrated.

Language Loyalty is the percentage of persons claiming to be Hispanic or Latino (ages 5+) who speak Spanish as a home language. Loyalty differs from Density in its denominator, with the latter being a percentage of the total population, while Loyalty measures its percentage relative to the population claiming Hispanic or Latino ethnicity. This term was first used by Weinreich (1953) with a decidedly different definition, “as a principle…in the name of which people will rally themselves and their fellow speakers consciously and explicitly to resist changes in either the functions of their language (as a result of a language shift) or in the structure or vocabulary (as a consequence of interference)” (99). By the time Hudson et al. (1995) employed this term, it was effectively stripped of any intentionality among Spanish speakers and rather was used as an objective indicator of Spanish language maintenance in the Latino population. Wilson and Jenkins (2020, p. 34) alternatively label the measurement “language maintenance among Hispanics” (LMH). We continue the wording of “Language Loyalty” with the definition provided by Hudson et al. for the sake of continuity, with an acknowledgment of the potentially problematic nature of the term.

As a predictor, Count corresponds to larger overall populations. The darkly shaded counties in Figure 4, as metropolitan areas and large population centers, can be discerned clearly. The heavy population counts along Interstate 5 from Seattle to Portland, Interstate 15 along the Wasatch Front in Utah, Interstate 25 along Colorado’s Front Range, and Interstate 35 between Dallas/Fort Worth and San Antonio are all readily identifiable on the map. Other metropolitan centers such as Houston, Phoenix, Las Vegas, Reno, Boise, Amarillo, and Lubbock are likewise easy to spot. California, for its part, boasts large total populations in every county between the Bay Area and the Mexican border. Each of these stands out on the map. Perhaps more notable are counties with considerably smaller overall populations (such as Southern Arizona or Northern and Southeastern New Mexico), but whose Spanish-language counts still place them in the top quintiles.

Count does not give a clear indication of those counties with high concentrations of Hispanics or Spanish speakers. The San Luis Valley of Southern Colorado, for example, has high Hispanic densities but does not stand out on a map depicting Count, given the fact that the total populations of these counties would not place them above the third quintile, even with a 100% Hispanic concentration. A much better indicator of Spanish-speaking communities is the Density measure. Figure 5 paints a much different picture than that of Count, as the size of the county is irrelevant to the concentration of Spanish speakers. This map clearly shows the traditional Spanish-speaking communities of the Southwest, especially along the Mexican border, as well as the entirety of New Mexico and Southern Colorado. These are the areas where most studies on Spanish in the Southwest originate. Perhaps the most noteworthy areas of high Density on this map can be seen in Central Washington, Southern Idaho, and even Northern California, as these areas are relatively newer among Spanish-speaking communities. An interesting visual comparison can be seen between Figure 5 and the 10 corresponding states in Figure 1, as they appear to be almost negative images of one another. That is, areas of greatest Spanish-language density are inverse to areas of Hispanic growth. Negative growth areas on map 1 are the areas of highest concentrations of Spanish speakers on 3. This holds true not only for the West but for the country as a whole. The Spanish-speaking population, especially that of Mexican and Central American origins, is expanding throughout the country; states like North Carolina and Georgia, whose Spanish language communities are relative newcomers to the region, lead the way nationally in Spanish-language growth rates. Florida, which has long been a center of Cuban and Puerto Rican populations, now boasts the sixth-largest Mexican-origin population in the country.

Language Loyalty, as a measure, shows what percentage of the Hispanic/Latino population claims Spanish as a home language. The United States as a whole has a Loyalty rate of 69.8%. For the ten-state West, that figure is similar, if slightly lower: 68.7%. These figures have declined for at least the past 20 years, as in 2000, the national Loyalty rate was 78.0%, followed by 76.1% in 2010. If we observe current Loyalty rates for each of the ten western states, then certain patterns emerge (

Table 4. Count, Density, and Language Loyalty in 10 Western States, 2020 (U.S. Census Bureau 2020c, 2020d, 2020e).

	Count	Density	Loyalty
Texas	7,666,020	28.8%	71.4%
California	10,462,968	28.3%	71.3%
Nevada	593,610	20.9%	70.9%
Washington	602,058	8.5%	63.4%
Oregon	349,549	8.9%	62.8%
Utah	297,926	10.3%	62.4%
Arizona	1,358,980	20.2%	62.2%
Idaho	131,145	8.0%	57.6%
New Mexico	514,071	26.0%	51.3%
Colorado	602,273	11.2%	49.0%
10-state total	6,689,293	23.5%	68.7%

). Leading the way are Texas and California, which both count Loyalty at 71%. These are the two largest Spanish-speaking states in the United States, accounting collectively for nearly half of the nation’s Spanish speakers. They not only have high counts but also the highest densities (~28%) of any western state. Nevada’s Loyalty rate also nears 71%. This is likely due to the continual and recent flow of immigrants into the state, which also plays an important role in California and Texas. McCullough and Jenkins (2005) found a strong correlation between Language Loyalty and recent immigration in Colorado, and various scholars (Fishman 1972; Silva-Corvalán 1994, among many others) have identified continuous immigration as the most significant indicator of language maintenance in a community.

After the top three states, Loyalty rates drop off dramatically, with a four-state cluster in the 62–63% range. Three of these four are the northwestern states of Washington, Oregon, and Utah, which have similar density trends. Idaho, in that same trend group and the smallest and least dense of any of the ten states, counts Loyalty at 58%. New Mexico (51%) and Colorado (49%) merit special attention in any discussion on language maintenance. Each of these latter states can identify two separate Spanish-speaking populations with very different histories, with recent Mexican immigrants forming the Latino majority in Southern New Mexico and Northern/Central Colorado, and the traditional Spanish-speaking communities in Northern New Mexico and Southern Colorado, which have roots in the area that date back for more than three centuries. The New Mexico territory was a majority Spanish-speaking area until the 1910 Census, with statehood coming two years thereafter. Southern Colorado draws much of its Hispanic heritage from northern New Mexico, with land grant settlers populating the area beginning in 1851. The San Luis Valley in that part of the state has been a majority Hispanic community since then. However, these areas have both seen precipitous declines in Spanish language density and loyalty over the past four decades. As a representative example, in 1980, the Loyalty rate for the San Luis Valley was 91.4%, and 77% of the population self-identified as “Spanish Origin” (the U.S. Census began using “Hispanic or Latino” in 1990). By 2020, that figure had dropped to 50% (

Table 5. Language Loyalty comparison of San Luis Valley and Denver metropolitan area, Colorado (Wilson and Jenkins 2020; U.S. Census Bureau 2020d).

	1980	1990	2000	2010	2020
San Luis Valley	91.4%	72.7%	70.8%	65.0%	50.0%
Denver Metro	53.9%	50.8%	66.3%	65.3%	53.0%

). The Denver metropolitan area, in the northern half of the state, showed an opposite trend between 1980 and 2000, plateauing for the next decade, then dropping at roughly the same rate as the San Luis Valley in the last 10 years, to pre-2000 levels. What was a once a high-density Spanish-language area has declined in its language maintenance to levels lower than those of the Denver metropolitan area. As suggested above, immigration appears to play a major role in this, as the San Luis Valley is an isolated region that attracts virtually no immigration, while Denver is a major Western city with a continual influx of both immigrants (from other countries) and in-migrants (from other states). The state as a whole does not have as high a density as the border states, the recent immigration rates are lower, and the result is a lower level of Language Loyalty than any other state in the West. While New Mexico has relatively high density rates, those numbers are declining more rapidly than in any other state and are not being replaced by immigration at a commensurate rate, and intergenerational language shift is the result.

5. Social Factors

One of the main research questions for this paper is whether population growth and change have impacted the relationship between Spanish language use and the social factors that surround language use. Considering the assertion by Hudson et al. (1995) that Spanish language use comes at a price in the Southwest, we revisited their original social variables with data from the 2020 Census, 40 years after that of the original analysis. Hudson et al. found strong or moderate negative correlations between language loyalty and virtually every socioeconomic variable within the Spanish origin population. Education (r = −0.68), per capita income (−0.59), poverty level (0.50), managerial and professional occupation (−0.45), and persons per household (0.56) all showed significant Pearson’s correlations greater than 0.40. They also found that Density correlated somewhat pessimistically with socioeconomic variables in the total population, with the conclusion that “these correlations…demonstrate only indirectly the effects of social factors on density. Instead, what they appear to reflect is the asymmetric distribution of the Spanish origin population across the various socioeconomic strata of U.S. society” (p. 177).

In order for valid comparisons across time, only the original five states from Hudson et al. (1995) were included in the 2020 correlations (see below for comparative data on the other five western states). The current data show some notable comparisons and contrasts with the 1980 data. Looking at the correlations in

Table 6. Measures of Spanish language maintenance in five Southwestern states, correlated with social variables, by county, 2020 (U.S. Census Bureau 2020a, 2020b, 2020c, 2020d, 2020e, 2020f, 2020g, 2020h, 2022a).

		Count	Density	Loyalty
	Count		0.15	0.15
	Density	0.15		0.59 *
	Loyalty	0.15	0.59 *
Total Population	Total population, N	0.96 *	0.08	0.12
	Hispanic/Latino population, N	0.996 *	0.15	0.14
	High School graduate or higher	−0.21	0.00	0.02
	Bachelor’s degree or higher	0.15	−0.33 *	−0.21
	Per capita income	0.11	−0.41 *	−0.18
	Poverty level	−0.01	0.40 *	0.16
	Unemployment rate	0.06	0.15	0.11
	Household size	0.18	0.56 *	0.38 *
	Managerial/professional %	0.11	−0.44 *	−0.33 *
Hispanic or Latino	Persons per household	0.13	0.14	0.40 *
	High School graduate or higher	0.00	−0.25 *	−0.63 *
	Bachelor’s degree or higher	0.09	−0.19	−0.34 *
	Per Capita income	0.04	−0.14	−0.25 *
	Poverty level	−0.04	0.12	0.08
	Unemployment rate	0.03	0.12	0.01
	Managerial/professional %	0.04	−0.08	−0.30 *

Boldface = 0.40 or greater * p < 0.005.

between Density and socioeconomic variables, we see significant correlations for per capita income, poverty, household size, and managerial and professional positions. Keeping in mind that these data reflect the total, and not just Hispanic, population, Hudson et al.’s (1995) assertion that these correlations are reflective of the “asymmetric distribution of the Spanish origin population” appears to continue to be reflected in these relationships. The fact that a negative correlation exists between Density and managerial or professional positions for the total population appears to show a scarcity of availability of such positions for any inhabitant of those counties where Density is higher.

When comparing the 2020 Language Loyalty data with the 1980 data, we see a somewhat different outcome. We still see a weak correlation with managerial/professional positions, a moderate correlation with persons per household, and a strong correlation with levels of education. The educational variable is likely one that will not change in the face of continued immigration, as a significant percentage of Spanish speakers in the United States are working-class products of a foreign educational system from which they did not graduate, a fact that will not be altered by social changes in the United States. The other variables show a departure from the relationships reflected in the 1980 data, as the relationship between Loyalty and per capita income is considerably weaker in the current data (−0.25 in 2020 vs. −0.59 in 1980), and no significant correlation exists between Loyalty and either poverty or unemployment.

Jenkins (2009b) conducted a similar comparison between the same variables for 1980 and 2000, which serves here as a midpoint in the data. That study also showed a weakening of the relationships between these variables. In fact, in that study, there were even fewer significant relationships between Linguistic Loyalty and socioeconomic variables, with significance only showing weak correlations with educational attainment and per capita income. The cases of persistent weakening are generally positive as they relate to language use within a community, but the reversing trend lines between 2000 and 2020 may be suggestive of a demographic blip that corresponded with the surge in immigration between 1990 and 2006. We know that Loyalty in the region decreased from 74% in 2000 to 69% in 2020. Data from 1980 are more difficult to obtain, but there is no indication of a large jump in Language Loyalty between 1980 and 2000. More study is needed to track these trend lines and determine if an up-and-down relationship between Loyalty and social variables corresponds to ebbs and flows in immigration over time.

Finally, as we look at sociodemographic data exclusively from the five Northwestern states not included in the data from

Table 7. Correlations across three census periods between Linguistic Loyalty and socioeconomic variables within the Hispanic/Latino population in five Southwestern states (U.S. Census Bureau 2020a, 2020b, 2020d, 2020f, 2020g, 2020h).

	1980	2000	2020
Persons per household	0.43 *	0.13	0.40 *
% High School Diploma †	−0.68 *	−0.26 *	−0.63 *
% Bachelor’s or Higher †	−0.68 *	−0.17	−0.34 *
Per capita income	−0.59 *	−0.25 *	−0.25 *
%Poverty level	0.45 *	0.04	0.08
Unemployment rate	−0.22	−0.05	0.01
% Managerial/professional	−0.45 *	−0.09	−0.30 *

Boldface correlations > 0.40; * p < 0.005; ^† 1980 figures are for “Median years of education”.

and

Table 8. Correlations across three census periods between Linguistic Loyalty and socioeconomic variables within the Hispanic/Latino population in five Northwestern states (U.S. Census Bureau 2020a, 2020b, 2020c, 2020d, 2020e, 2020f, 2020g, 2020h).

	Count	Density	Loyalty
Count		0.23 *	0.17
Density	0.23		0.69 *
Loyalty	0.17	0.69 *
Persons per household	0.08	0.33 *	0.44 *
High School graduate or higher	0.00	−0.49 *	−0.55 *
Bachelor’s degree or higher	0.04	−0.34 *	−0.26 *
Per capita income	0.10	−0.19	−0.18
Poverty level	−0.04	0.09	−0.01
Unemployment rate	0.03	0.01	−0.07
Managerial/professional %	0.08	−0.24 *	−0.19

Boldface correlations > 0.40; * p < 0.005.

, some important distinctions can be made with the southwestern states. While the population consists of individuals of the same national origins in roughly the same distributions, the Northwest as a Spanish-speaking region is newer and considerably smaller than the Southwest. The Count in the Southwest numbers 20.6 million speakers, compared to just under 2 million speakers in the Northwest. Density is also more than double in the Southwest (26.5%) what it is in the Northwest (10.7%). Language Loyalty also favors the Southwest, at 69%, contrasted with 65% in the Northwest.

Remarkably, the correlations between Language Loyalty and socioeconomic factors in the Northwest are essentially the same as those same comparisons in the Southwest. Persons per household shows a moderate correlation with Loyalty, and educational attainment continues to correlate negatively with Spanish language maintenance. No other social variable correlates with Language Loyalty in the Northwest, suggesting that Spanish language use and maintenance in a social context is subject to the same social processes in each of the two regions, in spite of the differences in Count, Density, and Loyalty.

6. Conclusions

The Spanish-speaking population in the Western United States continues to grow at impressive rates, strengthening its position as the largest minority population in the country. Growth is happening in areas that are not traditionally Spanish-speaking but in regions where the Spanish language is relatively new, both in the West as well as throughout the country. Among western states, northwestern states are growing at a much higher rate than southwestern states, though Hispanics in the Southwest still outnumber those in the Northwest by a more than 10:1 ratio. Density in the Southwest peaked with the 2010 Census, and rates are declining slightly, whereas in the Northwest, densities continue to increase. Community language loss is highest in areas like New Mexico and Southern Colorado, where language maintenance figures among Hispanics are around 50%. The Southwest United States, and especially New Mexico and Southern Colorado, have traditional Spanish-language communities that pre-date English, sometimes by centuries, but the shift to English in those areas over the last century has mirrored the language shift in immigrant communities. Additionally, New Mexico and Southern Colorado do not benefit from large measures of continued immigration as a maintenance factor. No community is immune to the Spanish language shift, but positive population growth and continuous immigration correlate strongly with areas where Language Density and Loyalty continue to increase.

The cost of Language Loyalty appears to be less than it was a generation ago, as correlations between Spanish language use and social variables have decreased in significance and in strength with each successive decade. While there is certainly no indication of an economic benefit to Spanish language use, the price of Spanish as a home language has decreased. Educational variables continue to show a strong negative correlation with language use, but that appears to be more a product of the educational status of immigrants than that of native U.S. Spanish speakers.

In spite of the indications of language loss in the area, the continued growth and strength of the Spanish-speaking population in the Western United States will continue to be a vital and fertile source of inquiry for those who study the language in its social context in the West.