American identity and attitudes toward English language policy initiatives.

Relatively little is known about what individual-level factors drive Americans' attitudes toward offering services to immigrants. Using national-level data and logistic regressions, we examine what factors co-vary with whether respondents agree or disagree with specific policy initiatives regarding support for English language use for immigrants. We then examine what factors are related to whether respondents agree that tax money should be used to fund English classes for immigrant children and adults. We find that age, race, and general warmth toward undocumented immigrants predict English-only attitudes, and that marital status, education, and warmth toward undocumented immigrants predict attitudes toward the use of public funds to teach English.

Keywords: language policy, immigrants, attitudes, English classes

Introduction

The foreign born population of the United States grew from 7.9 percent in 1990 to 11.1 percent, or 31.1 million residents by 2000 (Schmidley, 2001; U.S. Department of State, 2002). Since 2000 the United States has continued to welcome large numbers of immigrants admitting 1,063,732 in 2002 alone of which over 40 percent originated from Spanish-speaking countries (U.S. Department of Justice, Immigration and Naturalization Service 2001; U.S. Department of Homeland Security, 2003).

Language is at the core of the policy debate over immigrants' impact on American culture (Lapinski, Peltola, Shaw, and Yang 1997). However, when we examine the research done on American attitudes toward English usage in public schools and the use of public tax money to teach immigrants English, the literature is modest. Some research has isolated correlates related to anti-immigrant attitudes (see Cowan, Martinez, and Mendiola, 1997 and Esses, Dovidio, Jackson, and Armstrong, 2001), but these conclusions have been based on small samples of college students and may not be representative of the general population. Research on sentiment toward making English the official language, as Propositions 187 and 227 in California intend, indicates the importance of language in shaping attitudes toward illegal immigrants (Cowan, et al. 1997). Americans who believe that English should be the only language in schools increased from 40 to 48 percent from 1993 to 1995 (Lapinski et al., 1997). This English-only sentiment is an important indicator of openness toward immigrants, especially if this trend continues. Preliminary analysis of our survey data shows this trend has become more pronounced with 66 percent of 395 respondents in 2001 reporting that English should be the only language used in public schools.

Using data from the University of Oklahoma's 2001 Survey of American Attitudes (SAA) national telephone survey we examine the individual-level factors that may predict more altruistic and open attitudes toward English language policy initiatives, such as whether English should be the only language used in public schools, and whether the same types of individuals who agree that tax money should be used to teach English to immigrant children also agree that tax money should be used to teach English to immigrant adults.

Throughout the twentieth century, the general trend in public opinion has been a growing negativity toward immigrants (Simon, 1985; Jarret, 1999), possibly caused by the perception that these newcomers threaten existing American cultural identity, beliefs, and values (Espenshade and Calhoun, 1993; Esses et al., 2001). Because English language use is a salient component of American identity, the symbolic politics model is useful in framing this analysis. The symbolic politics model posits that cultural symbols, such as language choice, may signify what it means to be an American and can influence opinions on other related issues such as bilingual education or immigration policy in general (Citrin, Reingold, Walters, and Green, 1990a; Citrin, Haas, Muste, and Reingold, 1994). In this paper, we examine the importance of English as a cultural symbol. Using the symbolic politics model, we can predict that because speaking English is such an important part of American identity, Americans would be willing to support the use of public funds to teach English to immigrants, regardless of whether they are children or adults.

The labor market competition model has also been used to understand American public opinion toward immigrants and immigration policies. According to the labor market competition theory, persons with lower social and economic status are less likely to view increased levels of immigration as a good policy direction, because low-wage, low-skill workers compete with immigrants for jobs in the economy (Abowd and Freeman, 1991; Bean, Lowell, and Taylor, 1988; Borjas and Freeman, 1992; Oliver and Mendelberg, 2001). Indeed, it is reasonable that those with lower social and economic status in society would face greater competition and threat to their livelihood with an influx of less-educated immigrants, compared with those of higher status. However, as Smith and Edmonston (1997) note, this may be more perception than reality. Additionally, those with lower status and fewer skills are likely to resent that public money would be spent to provide English training exclusively for immigrants when they, too, are challenged with their own skill levels in the labor market but offered no publicly funded assistance or training.

Evidence suggests that those with more education, higher incomes, and high status jobs are more likely to hold more favorable attitudes toward increased immigration levels, compared with those of lower status (Day 1990; Hoskin and Mishler, 1983; Simon, 1987; Simon 1985; Simon and Alexander, 1993; Sorensen and Krahn, 1996; Starr and Roberts, 1982). Some studies suggest that there may be variation across these social and economic categories (Burns and Gimple, 2000; Morris, 1985; Peterson and Kozmetsky, 1982). Other research provides additional weight to the importance of education in driving social status, showing that negative attitudes toward immigrants decrease with more education (Day 1990; Hoskin and Mishler, 1983; Moore, 1986; Starr and Roberts, 1982). A picture of higher status translates into more open, favorable attitudes toward immigration and immigrants. Thus, we would expect that those with higher status in society would hold more open attitudes toward English usage in public schools and the use of public funds to pay for the teaching of English to immigrants. Likewise, an extension to the labor market thesis would hold that Non-Whites may be less supportive or less open to immigrants because they are more likely to be in competition for lower status jobs (Jarret, 1999; Smith and Edmonston, 1997).

Espenshade and Calhoun (1993) established the need to control for individual-level demographic variables, such as age, sex, marital status, and race, when examining American attitudes toward immigrants. We therefore include these variables in our models. Further, Lapinski and associates (1997) distinguish between legal immigration and undocumented immigration when assessing Americans' attitudes toward immigration policies, because at the same time that more balanced beliefs are held regarding legal immigrants, negative attitudes are held of illegal immigrants. Other research (Passel, 1986) shows that it is important to separate immigrants and undocumented immigrants when examining attitudes toward immigrants or immigration issues. Cowan and associates (1997) in their survey of 140 Los Angeles area college students further established that attitudes toward undocumented immigrants are uniquely understood. Frendreis and Tatalovich (1997), using data from the 1992 American National Election Study, found that respondents' attitudes toward undocumented immigrants helped to predict support for English-only policy initiatives. These studies point to the need to control for respondents' general warmth perceptions toward immigrants and undocumented immigrants when explaining open attitudes toward the specific English language policy initiatives that affect immigrants.

In this study, we improve on previous research in several ways. First, we use data gathered from a national sample. Second, we examine Americans' attitudes toward the public funding of teaching English to immigrant children and immigrant adults, and we consider how these attitudes may co-vary with individual characteristics. Third, we improve on earlier studies by using multivariate models to control simultaneously for many factors, like education and age, which have been previously established in bivariate analysis as related to open attitudes toward immigrants or immigration policy issues.

Drawing on the research that has been done on Americans' attitudes toward immigration, we expect that Americans will agree to English-only in public schools and support the use of public funds to teach English to immigrants, regardless of whether they are children or adults because speaking English is seen as an integral part of being American. Further we expect to find that those with less education are likely to compete directly with immigrants for jobs, and, hence, are less likely to be open to the use of public funds to provide English training to immigrants, which may make immigrants more competitive in the labor market. As an extension of the labor market thesis, we believe that because Non-Whites view immigrants as competition for low-level employment, they are less likely to hold open attitudes toward these English-language issues. Finally, we expect that those with warm feelings toward immigrants in general, and undocumented immigrants in particular, are likely to hold more open attitudes toward these English language policy initiatives, specifically English-only language use in schools and public funding to teach English to immigrant children and adults.

Data and Analytical Samples

The analysis in this article is based on data from a telephone survey, the Survey of American Attitudes (SAA), administered from August 27th through September 22, 2001 by the University of Oklahoma's Public Opinion Learning Laboratory.

Faculty and graduate students at the University of Oklahoma developed the SAA survey instrument. Trained interviewers collected data from 395 respondents, ages 18 years and older. Respondents were initially separated into pre-September 11th and post-September 11th groups, however preliminary analysis did not yield significant differences between the two groups on relevant variables, including baseline demographic and socioeconomic variables and the dependent variables. They were subsequently treated as one sample for this research paper.

Multivariate Methods

Overall, we specify 12 logistic regression models to examine three English language policy initiatives: 1) attitudes toward the use of English-only in public schools, 2) attitudes toward the use of tax money to teach English to immigrant children, and 3) attitudes toward the use of tax money to teach English to immigrant adults. Because the dependent variables of interest are dichotomous and their values fall between 0 and 1, ordinary least-squares (OLS) regression is likely to yield out-of-bound and therefore nonsensical predications (Aldrich and Nelson, 1984). We accordingly choose a logistic regression model to analyze these dependent variables. The statistical advantages of the logistic and multinomial logit specification over the linear probability model for binary and categorical variables are well known (e.g., King, 1989; Long, 1997). Models 1 through 4 examine the use of English only in public schools, and models 5 through 12 examine the use of tax money to teach English to immigrant children and adults, separately.

We use dichotomous variables to indicate whether or not an individual reported agreeing that English should be the only language used in public schools, that tax money should be used to teach English to immigrant children, and that tax money should be used to teach English to immigrant adults. Respondents were asked whether they agreed or disagreed with the following statements:

1) English should be the only language used in public schools.

2) Tax money should be used to teach English to children who immigrate to the United States.

3) Tax money should be used to teach English to adults who immigrate to the United States.

Across all three statements, we used one 'agree' category from those who reported that they strongly and somewhat agreed with the statements, and one 'disagree' category from those who reported that they strongly and somewhat disagreed with the statements. Each of these statements is a measure of openness toward English language policy initiatives for immigrants. For the first statement, we recoded the direction of English-only use, so that favorable or open attitudes toward all three of these policy issues can be interpreted across models in the same direction.

We include several socio-economic and demographic factors in the base model (see Table 1). Age is a continuous variable. Sex and marital status are two-level categorical variables: male or female, and married or not married. Education is a three-level categorical variable denoting high school degree or less, some college, and college degree or more. Race is a two-level categorical variable indicating White or Non-White.

We use a continuous variable that shows a respondent's overall warmth or coolness toward immigrants and undocumented immigrants, respectively, as a ranking from 1 indicating extremely cold, negative feelings toward immigrants to 10 indicating extremely warm, positive feelings toward immigrants. A response around 5 indicates neither warmth nor coolness toward a group. The warmth measure for immigrants has a mean of 5.62 and includes 387 valid responses. The warmth measure for undocumented immigrants has a mean of 3.71 and includes 382 valid responses. This higher level of warm feeling toward documented immigrants compared with undocumented immigrants indicates that respondents felt differently toward these two groups. Furthermore, it would suggest that attitudes toward English education and the use of public funds may vary based on the documentation status of immigrants.

Descriptive Results

Table 1 presents the bivariate relationships between the independent and dependent variables. While most respondents (66 percent) indicated that English should be the only language used in public schools, they also reported being strongly in favor of spending tax dollars to teach English to immigrant children (79 percent) and immigrant adults (74 percent). These findings indicate an overall support of the English language as a cultural symbol for Americans, and the willingness to pay for the cultural and social integration of immigrants. Findings such as these are consistent with previous research which shows that Americans are willing to support bilingual education and its associated costs because English language is seen as an integral part of American culture and identity (Huddie and Sears 1990; Citrin et al., 1990a; Citrin et al., 1994).

The bivariate relationships suggest that younger individuals and unmarried individuals hold more open attitudes across all three policy questions. In addition, women report higher rates of support for bilingual education and tax money to teach English to immigrant children, compared with men. Those with more education hold more open attitudes toward using tax money to teach English to immigrant children and immigrant adults, but this relationship between education and bilingual education does not appear to be in the hypothesized direction of each level of increased education translating into more open attitudes. Thirty-five percent of high school graduates or less, 29 percent of those with some college education, and 37 percent of those with a college degree or more in education agreed that English should not be the only language used in public schools. Non-White individuals have much higher levels of agreement that English should not be the only language used in public schools, which supports the cultural affinity hypothesis. The high levels of agreement across White and Non-White racial/ethnic categories indicates further support for the symbolic politics argument mentioned above, because English language is an American cultural symbol that they are willing to use tax money to support.

We next consider these variables in multivariate analyses to determine whether these relationships hold once we account for other inter-correlated variables.

Multivariate Results

Table 2 shows respondents odds of not agreeing that English should be the only language used in schools. Model 1 shows that age and race are significant variables. With each increasing year of age, a respondent is more likely to believe that English should be the only language used in public schools. In contrast, Non-White respondents show more openness toward language usage, being more than twice as likely to report that English should not be the only language used in schools. These effects hold throughout the analysis and support the cultural affinity hypothesis (Espenshade and Hempstead, 1996). In Model 2 we see that a respondent's warmth toward immigrants is also a significant predictor of agreeing that English should not be the only language used in public schools. A similar effect is observed when the independent effect of warmth toward undocumented immigrants is added in Model 3. However, when both warmth measures are included in Model 4, the measure for respondents' attitudes toward undocumented immigrants, specifically, is the dominant variable driving this warmth effect. Age, race, and the warmth measures are significant predictors of whether a respondent will agree that English should not be the only language used in public schools.

Table 3 shows the odds of respondents agreeing that tax money should be spent to teach immigrant children English, and Table 4 shows the odds that respondents agree tax money should be spent to teach immigrant adults English. Marital status, education, and the warmth measures are significant predictors of whether a respondent thinks that tax money should be spent to teach English to immigrants. Across both models, with each level of additional education, respondents" attitudes are more favorable to using public funds to teach immigrants English. Table 3 shows that female respondents are more likely to agree that tax money should not be spent to teach English to immigrant children, but there is a significant relationship between sex and agreeing that tax money should be spent to teach English to immigrant adults. Warmth toward immigrants in general and warmth toward undocumented immigrants, specifically, are significant predictors of whether a respondent will agree that tax money should be spent to teach English to immigrant children (Model 7 and Model 8 of Table 3) and adults (Model 11 and Model 12 of Table 4). These results remain significant when both warmth toward immigrants and warmth toward undocumented immigrants are included in the analysis, indicating that the independent effect of each immigration warmth measure is a significant predictor of respondents' attitudes toward the English language policy issues of using tax money to teach English both to immigrant children and adults.

One difference across the models presented for immigrant children and immigrant adults is that women are 1.7 times as likely as men to report that they agree that tax money should be spent to teach English to immigrant children in Table 3. In contrast, women are not significantly more likely than men to report that they agree that tax money should be spent to teach English to immigrant adults.

Discussion

The bivariate relationships between our dependent variables and age, education, and marital status did not always hold true with multivariate analyses. Moreover, unique sets of factors emerge as predicting opinions in the two distinct areas of inquiry relating to English-language issues. While we find that being young is a significant predictor of open attitudes toward non-English-only use in schools, and as previously noted by Espenshade and Calhoun (1993) to influence general opinions of immigrants, it is not a significant predictor of favorable attitudes toward publicly funding English classes for immigrant children or immigrant adults. Furthermore, being young and single are related to open attitudes regarding the use of public funds to teach English to immigrants, but we find that education and marital status do not co-vary with attitudes of English-only usage. These findings are consistent with previous research on the attitudes of married and unmarried respondents (Shapiro and Mahajan, 1986; Conver, 1998).

In contrast, Non-White respondents are more likely to support the use of languages other than English in public schools, yet they are not more likely to support the use of public funds to teach immigrants English. This suggests that individual-level factors shape opinions of immigrants differently even within the same issue area. Non-Whites are much more open than Whites to having languages other than English used in public schools, but no relationship exists between race and the public funding of teaching English to immigrant children or to immigrant adults. This suggests that the bivariate race association may actually be due to an education effect. Indeed, Whites are much more likely to fall in the highest education category, 41 percent, compared with just 18 percent of Non-Whites. Alternatively, this effect may be due to an increased desire to assimilate immigrants to American culture as suggested in the symbolic politics model.

Across our models, as found previously (Passel, 1986; Cowan et al., 1997; Frendreis and Tatalovich, 1997; Lapinski et al., 1997), the measures of general warmth toward immigrants and general warmth toward undocumented immigrants are significant predictors of attitudes toward English-only usage in school or the use of public funds to teach English to immigrant children and to immigrant adults. Further, the warmth toward undocumented immigrants' measure bears out as a major explanatory variable once both are included in our models, suggesting that the sentiment toward specific English language services offered to immigrants may be shaped through a filter of general warmth of the undocumented immigrant population. This finding supports earlier research on the centrality of sentiment toward undocumented immigration in framing opinions on a variety of issues associated with immigrants more generally (Espenshade and Calhoun, 1993; Citrin et al., 1990a). Alternatively, it is unclear in which direction these relationships exist. Is it that general warmth affects open attitudes toward English-language issues or is it that attitudes toward English-language issues shape general warmth?

One might expect that attitudes toward immigrant children would be overwhelmingly more altruistic than for adults, because social norms posit that childhood is a time of protection, education, and vulnerability (Corsaro, 2004). However, our results show for the most part that respondents hold very altruistic and similar attitudes toward the use of public funds to teach English both to children and adults (74 percent and 79 percent, respectively [Table 1]). Likewise, the factors relevant for predicting open attitudes toward using public funds to teach children English--marital status, education, and general warmth toward immigrants--are also salient in explaining public attitudes toward funding to adults. Altruism, however, may not be entirely at the core of attitudes such as these. An alternative explanation, and one that is consistent with the symbolic politics model, is that immigrants are expected to abandon cultural practices of their sending country for those of the core culture of the receiving country (Gordon, 1964).

Finally, while our results largely indicate little difference in what drives attitudes toward funding the teaching of English to immigrant children and immigrant adults, there does appear to be a gender difference. It is striking that women are 1.7 times as likely to agree that tax money should be used to teach English to immigrant children (Table 3, Model 8), all else being equal. Further, it is striking that women hold more altruistic attitudes than men when it comes to the use of public tax money to fund the teaching of English to immigrant children, but their attitudes do not diverge significantly from men when considering adults (Table 3, Model 12), all else being equal. This gender difference in more altruistic attitudes toward children for women may be reflective of their greater caretaking role of children in American society.

Conclusion

An important lesson culled from the debate over Proposition 187 in California is the importance of the English language to Americans. As the foreign-born population continues to grow and migrate to nontraditional locations this is an issue that will inevitably resurface (Saenz, 1996, Hernandez-Leon and Zuniga, 2000; Garcia, 2005). Previous scholarship highlights the importance of how English language policies are presented and framed, because this affects Americans' attitudes of these policies (Huddie and Sears 1990; Citrin et al., 1990a). Other studies emphasize the importance of English language use for American identity and culture (Citrin, Reingold, and Green, 1990b; Espenshade and Calhoun 1993).

This analysis offers additional support to the symbolic politics model as a tool to explain attitudes toward immigrants. Not only do most respondents believe that English should be the only language used in public schools, but they are also open to using tax money to teach English to immigrant children and immigrant adults. This indicates that respondents view English language as a cohesive force solidifying the United States citizenry, and that a strong willingness exists to meet immigrants half-way in their assimilation process by providing English language education. This in turn may help explain the higher level of openness expressed when documented immigrants are included in the model, as they may be perceived as following the rules to become a part of American culture. Overall, our findings suggest that English language is a cultural symbol that respondents are willing to support with money from public tax coffers.

When examining the willingness to use tax money to fund English training for immigrants, we also find some support for the labor market competition thesis. At each increment of more education, individuals express more open attitudes, agreeing that tax money should be spent to teach English to immigrants. It is reasonable that those with less education are less likely to support English training initiatives that would make immigrants yet more competitive for the low-status jobs they are likely to hold.

This research addresses an important piece of the debate over what factors distinguish Americans' views of English language usage in schools and the level of commitment to incorporate immigrants socially and culturally into American society. We find that while unique variables explain English-only preferences and attitudes toward public funding to teach English, the general warmth toward immigrants measures bear out as salient in patterning attitudes toward specific English language policy initiatives. While the directionality cannot be confirmed using cross-sectional data, our research suggests that attitudes toward specific services offered to immigrants go hand in hand with warmth regarding the immigrant population in general and the undocumented immigrant population in particular. These findings help explain public attitudes toward English-language issues as they concern immigrants.

References

Aldrich, J. H. & Nelson, F.D. (1984) Linear probability, logit, and probit models. Beverly Hills, CA: Sage.

Abowd, J. M. & Freeman, R.B. (1991) Immigration, trade, and the labor market. Chicago, IL: The University of Chicago Press.

Bean, F. D., Lowell, B.L., & Taylor. J.T. (1988) "Undocumented Mexican immigrants and the earning of other workers in the United States." Demography. 25, 35-52.

Borjas, G. J. & Freeman, R.B. (1992) Immigration and the work force: Economic consequences for the United States and source areas. Chicago: The University of Chicago Press.

Burns, P. & Gimpel J.G. (2000) "Economic insecurity, prejudicial stereotypes, and public opinion on immigration policy." Political Science Quarterly. 115, 201-225.

Citrin, J., Reingold, B., Walters, E., & Green, D.P. (1990a) "The 'Official English' movement and the symbolic politics of language in the United States." The Western Political Quarterly. 43, 535-559.

Citrin, J., Reingold, B. & Green, D.P. (1990b) "American identity and the politics of ethnic change." The Journal of Politics. 52, 1124-1154.

Citrin, J, Haas, E.B., Muste, C., & Reingold, B. (1994) "Is American nationalism changing? Implications for foreign policy." International Studies Quarterly. 38, 1-31.

Conover, P. J. (1988) "Feminists and the gender gap." The Journal of Politics. 50(4):985-1010.

Corsaro, W. (2004) The Sociology of Childhood (Sociology for a New Century Series) 2nd Edition. Thousand Oaks, CA: Pine Forge Press.

Cowan, G., Martinez, L. and Mendiola, S. (1997) "Predictors of attitudes toward illegal Latino immigrants." Hispanic Journal of Behavioral Sciences. 19, 403-15.

Day C. L. (1990) Ethnocentrism, economic competition, and attitudes toward U.S. immigration policy. Presented at Annual Meeting of the Midwest Political Science Association, Chicago Illinois, April 5-7.

Espenshade, T. J. & Calhoun, C. A. (1993) "An analysis of public opinion toward undocumented immigration." Population Research Policy Review. 12, 189-224.

Espenshade, T. J. & Hempstead, K. (1996) "Contemporary American attitudes toward U.S. immigration." International Migration Review. 30, 535-570.

Esses, V.M., Dovidio, J.F., Jackson, L.M., & Armstrong, T.L. (2001) "The immigration dilemma: The role of perceived group competition, ethnic prejudice, and national identity." Journal of Social Issues. 57, 389-412.

Frendreis, J. P. & Tatalovich, R. (1997) "Who supports English-only language laws? Evidence from the 1992 national election study. Social Science Quarterly. 78, 354-68.

Garcia, C. 2005. "Buscando trabajo: Social networking among immigrants from Mexico to the United States." Hispanic Journal of the Behavioral Sciences. 27, 3-22.

Gordon, M. M. (1964) Assimilation in American life: The role of race, religion, and national origins. (New York: Oxford University Press).

Hernandez-Leon, R. & Zuniga, V. (2000) "'Making carpet by the mile': The emergence of a Mexican immigrant community in an industrial region of the U.S. historic South." Social Science Quarterly. 18, 49-66.

Hoskin, M. & Mishler, W. (1983) "Public opinion toward new migrants: A comparative analysis." International Migration. 21, 440-462.

Huddie, L. & Sears, D. O. (1990) "Qualified public support for bilingual education: some policy implications." Annals of the American Academy of Political and Social Science. 508:119-134.

Jarret, C. (1999) "Troubled by new comers: Anti-immigrant attitudes during two eras of mass immigration to the United States." Journal of American Ethnic History. 18: 9-39.

King, G. (1989) Unifying political methodology. New York: Cambridge University Press.

Lapinski J. S., Peltola, P., Shaw, G., & and Yang, A. (1997) "Trends: Immigrants and immigration. Public Opinion Quarterly. 61, 356-383.

Long, J. S. (1997) Regression models for categorical and limited dependent variables. Thousand Oaks, California: Sage.

Moore, S. (1986) "Social scientists' views on immigrants and U.S. immigration policy: A postscript," The Annals of the American Academy of Political and Social Science. 487, 213-217.

Morris, M.D. (1985) Immigration: The beleaguered bureaucracy. Washington, DC: Brookings Institute.

Oliver, E.J. & Mendelberg, T. (2000) "Reconsidering the environmental determinants of White racial attitudes." The American Journal of Political Science. 44,574-589.

Passel J. S. (1986) "Undocumented immigration" The Annals of the American Academy of Political and Social Science. 487, 181-200.

Peterson, R. A. and Kozmetsky, G. (1982) "Public opinion regarding illegal aliens in Texas," Texas Business Review. 56, 118-121.

Saenz, R. 1996. "The demography of Mexicans in the Midwest." in R. Rochin (ed.) Immigration and Ethnic Communities: A Focus on Latinos. pp. 23-29 (East Lansing MI: Julian Samora Research Institute).

Schmidley, D. (2001) Profile of the foreign born population in the United States: 2000. U.S. Census Bureau, current population reports series P23-206. U.S. Washington, DC General Printing Office.

Shapiro, R. Y. & Mahajan, H. (1986) "Gender differences in policy preferences: A summary of trends from the 1960s to the 1980s" Public Opinion Quarterly. 50, 42-61.

Simon, R.J. (1985) Public opinion and the immigrant: print media coverage 1880-1980. Lexington, MA: Lexington.

Simon, R.J. (1987) Immigration and American attitudes. Public Opinion 10, 47-50

Simon, R. J. and Alexander S.H. (1993) The ambivalent welcome: Print media, public opinion and immigration. Westport, CT: Praeger.

Smith, J. P. & Edmonston, B. (1997) The new Americans: Economic, demographic, and fiscal effects of immigration. Washington D.C.: National Academy Press.

Sorensen, M.& Krahn, H. (1996) Attitudes toward immigrants: A test of two theories. The Alberta Journal of Educational Research. 421, 3-18.

Starr, P. D. & Roberts, A. E. (1982) "Attitudes toward new Americans: Perceptions of Indo-Chinese in nine cities," Research in Race and Ethnic Relations. 3, 165-186.

U.S. Department of Homeland Security. (2003) Yearbook of immigration statistics, 2002. Washington D.C.: U.S. General Printing Office.

U.S. Department of Justice, Immigration and Naturalization Service. (2001) Statistical yearbook of the immigration and naturalization service. Washington D.C.: U.S. General Printing Office.

U.S. Department of State. (2002) "Foreign-born population surges in United States." http://usinfo.state.gov/usa/diversity/ pr061002.htm. Retrieved August 8, 2004.

CARLOS GARCIA 

San Jose State University

Department of Sociology 

LORETTA E. BASS

University of Oklahoma

Department of Sociology

Table 1. Reported Attitudes toward English-only Use in
Schools and Using Tax Money to Teach English to Immigrants

                                         English should
                                           not be the
                                          only language
                                          used in public
                                            schools **

Characteristics       Total      % *      Total       %

Total 18 years
and older
  Agree                                     252      33.9
Sex Male                157      39.8       152      30.3
  Female                238      60.3       229      36.2
Age 18-27                60      15.2        60      60.0
  28-37                  71      18.0        68      35.3
  38-47                  79      20.0        78      30.8
  48-57                  83      21.0        81      29.6
  58-67                  36       9.1        34      35.3
  68-77                  36       9.1        32      21.9
  > or = 78              24       6.1        23       4.4
  Don't know              6       1.5
  No answer
Marital Status
  Married               220      55.7       213      30.5
  Not Married           173      43.8       166      38.0
  Don't know              2       0.1
  No answer
Education
  High school           113      28.6       107      34.6
  grad or less
  Some college          130      32.9       127      29.1
  College graduate      152      38.5       147      37.4
  or greater
Race/Ethnicity
  White                 343      85.6       330      30.3
  Non-White              49      13.7        48      58.3
  Don't know              3       0.8
  No answer

                      Tax money should    Tax money should
                      be used to teach    be used to teach
                         English to          English to
                         immigrant           immigrant
                        children **          adults **

Characteristics        Total       %       Total       %

Total 18 years
and older
  Agree                 303      78.9       287      74.0
Sex Male                151      76.2       156      74.4
  Female                233      80.7       232      73.7
Age 18-27                58      84.5        57      75.4
  28-37                  68      79.4        70      74.3
  38-47                  77      79.2        79      68.4
  48-57                  79      78.5        80      70.0
  58-67                  36      75.0        36      80.6
  68-77                  36      72.2        36      77.8
  > or = 78              24      83.3        24      87.5
  Don't know
  No answer
Marital Status
  Married               215      76.5       217      72.2
  Not Married           167      81.4       170      76.5
  Don't know
  No answer
Education
  High school           107      61.7       108      54.6
  grad or less
  Some college          127      80.3       128      75.0
  College graduate      150      90.0       152      86.8
  or greater
Race/Ethnicity
  White                 334      79.0       336      74.4
  Non-White              47      76.6        49      71.4
  Don't know
  No answer

Source: Societal Attitudes Survey, a national telephone survey
administered by the University of Oklahoma Public Opinion Learning
Laboratory. September 2001.

* Percentages may not sum to 100% due to rounding. ** Percentages
refer to those who thought that English should not be the only
language used in public schools, and that tax money should be used
to teach English to immigrant children and immigrant adults.

Table 2. Odds of Not Agreeing that English Should
Be the Only Language Used in Schools

                        Model 1          Model 2
                          Odds             Odds

Value                   SE Ratio         SE Ratio
Age                  (.01) .97 ***    (.01) .97 ***
Sex
  Male (R)                (R)              (R)
  Female              (0.24) 1.46      (.24) 1.52 *
Marital Status
  Married (R)             (R)              (R)
  Not Married          (.24) 1.34       (.24) 1.41
Education
  High School             (R)              (R)
  or Less (R)
  Some                 (.30) .79        (.32) .67
  College
  College              (.29) 1.45       (.30) 1.15
Race
  White (R)               (R)              (R)
  Non-White          (.34) 2.80 ***   (.35) 2.72 ***
Warmth toward                         (.06) 1.20 ***
immigrants
Warmth toward
undocumented
immigrants
Intercept
-2 Log Likelihood         39.6             48.4
Ratio
Degrees of Freedom         6                7
                         n=376            n=369

                        Model 3          Model 4
                          Odds             Odds

Value                   SE Ratio         SE Ratio
Age                  (.01) .97 ***    (.01) .97 ***
Sex
  Male (R)                (R)              (R)
  Female               (.25) 1.35       (.25) 1.37
Marital Status
  Married (R)             (R)              (R)
  Not Married          (.25) 1.41       (.25) 1.45
Education
  High School             (R)              (R)
  or Less (R)
  Some                 (.32) .79        (.33) .72
  College
  College              (.31) 1.38       (.32) 1.23
Race
  White (R)               (R)              (R)
  Non-White          (.36) 2.82 ***   (.36) 2.69 ***
Warmth toward                           (.07) 1.11
immigrants
Warmth toward        (.05) 1.18 ***   (.06) 1.14 **
undocumented
immigrants
Intercept
-2 Log Likelihood         50.7             51.9
Ratio
Degrees of Freedom         7                8
                         n=363            n=357

*** significant at p<.01, ** significant at p<.05, * significant
at p<.10

Source: Societal Attitudes Survey. A national telephone survey
administered by the University of Oklahoma Public Opinion Learning
Laboratory. September 2001.

Table 3. Odds of Agreeing that Tax Money Should
Be Spent to Teach Immigrant Children English

                        Model 5          Model 6
                          Odds             Odds

Value                   SE Ratio         SE Ratio
Age                    (.01) .99        (.01) 1.00
Sex
  Male (R)                (R)              (R)
  Female               (.27) 1.47      (.28) 1.72 *
Marital Status
  Married (R)             (R)              (R)
  Not Married         (.28) 1.68 *    (.30) 1.85 **
Education
  High School             (R)              (R)
  or Less (R)
  Some               (.31) 2.48 ***   (.33) 1.89 **
  College
  College            (.35) 5.87 ***   (.37) 3.83 ***
Race
  White (R)               (R)              (R)
  Non-White            (.40) 0.96       (.43) 1.13
Warmth toward                         (.07) 1.39 ***
immigrants
Warmth toward
undocumented
immigrants
Intercept
-2 Log Likelihood         33.3             51.3
Ratio
Degrees of Freedom         6                7
                         n=378            n=372

                        Model 7          Model 8
                          Odds             Odds

Value                   SE Ratio         SE Ratio
Age                    (.01) 1.00       (.01) 1.00
Sex
  Male (R)                (R)              (R)
  Female               (.28) 1.42      (.30) 1.68 *
Marital Status
  Married (R)             (R)              (R)
  Not Married         (.29) 1.66 *     (.31) 1.82 *
Education
  High School             (R)              (R)
  or Less (R)
  Some               (.32) 2.45 ***   (.34) 1.98 **
  College
  College            (.37) 5.07 ***   (.38) 3.78 ***
Race
  White (R)               (R)              (R)
  Non-White            (.44) .65        (.46) .75
Warmth toward                         (.08) 1.32 ***
immigrants
Warmth toward        (.07) 1.35 ***   (.08) 1.23 ***
undocumented
immigrants
Intercept
-2 Log Likelihood         52.8             34.2
Ratio
Degrees of Freedom         7                8
                         n=366            n=360

*** significant at p<.01, ** significant at p<.05,
* significant at p<.10

Source: Societal Attitudes Survey. A national telephone survey
administered by the University of Oklahoma Public Opinion Learning
Laboratory. September 2001.

Table 4. Odds of Agreeing That Tax Money Should
Be Spent to Teach Adult Immigrants English

                        Model 9          Model 10
                          Odds             Odds

Value                   SE Ratio         SE Ratio
Age                    (.01) 1.01       (.01) 1.02
Sex
  Male (R)                (R)              (R)
  Female               (.25) 1.01       (.26) 1.13
Marital Status
  Married (R)             (R)              (R)
  Not Married         (.26) 1.64 *    (.27) 1.81 **
Education
  High School             (R)              (R)
  or Less (R)
  Some               (.29) 2.46 ***   (.30) 2.04 **
  College
  College            (.32) 5.88 ***   (.33) 4.35 ***
Race
  White (R)               (R)              (R)
  Non-White            (.37) 1.18       (.39) 1.34
Warmth toward                         (.07) 1.35 ***
immigrants
Warmth toward
undocumented
immigrants                .--              .--
Intercept
-2 Log Likelihood         37.0             54.8
Ratio
Degrees of Freedom         6                7
                         n=382            n=376

                        Model 11         Model 12
                          Odds             Odds

Value                   SE Ratio         SE Ratio
Age                    (.01) 1.01      (.01) 1.02 *
Sex
  Male (R)                (R)              (R)
  Female               (.26) 1.01       (.27) 1.14
Marital Status
  Married (R)             (R)              (R)
  Not Married         (.27) 1.62 *    (.28) 1.79 **
Education
  High School             (R)              (R)
  or Less (R)
  Some               (.30) 2.31 ***   (.31) 1.98 **
  College
  College            (.34) 4.97 ***   (.35) 4.04 ***
Race
  White (R)               (R)              (R)
  Non-White            (.39) .82        (.41) .91
Warmth toward                         (07) 1.28 ***
immigrants
Warmth toward        (.07) 1.29 ***   (.07) 1.81 ***
undocumented
immigrants                .--              .--
Intercept
-2 Log Likelihood         51.6             37.0
Ratio
Degrees of Freedom         7                8
                         n=370            n=364

*** significant at p<.01, ** significant at p<.05,
* significant at p<.10

Source: Societal Attitudes Survey. A national telephone survey
administered by the University of Oklahoma Public Opinion Learning
Laboratory. September 2001.