The Political Assimilation of Immigrants and Their Descendants

Many skeptics of immigration reform claim that immigrants and their descendants will not politically assimilate and will consistently vote for bigger government for generations.¹ Political survey data suggest that this fear is unwarranted, as the political differences between immigrants and native-born Americans are small and, in most cases, so small that they are statistically insignificant. In the cases where the differences are significant, the descendants of immigrants rapidly assimilate into America’s political culture by adopting mainstream ideologies, political party identifications, and policy positions held by longer-settled Americans. The policy and political views of immigrants and their descendants are mostly indistinguishable from Americans whose families have been here for at least four generations. As a result of these small differences in opinion and the subsequent rapid assimilation of immigrants, they and their descendants are unlikely to alter America’s aggregate political attitudes.

This analysis is based on responses from the nationwide General Social Survey (GSS), a large biennial in-person survey of the demographic characteristics and attitudes of Americans that has been conducted since 1972.² The GSS records the respondent’s political party identification, ideological positions, and opinions on the spending levels of various government programs as well as the place of birth of the respondents, their parents, and their grandparents. The political opinions of the immigrants and their descendants can be measured directly in the GSS, which produces more accurate results than other polling methods.³ The first-, second-, third-, and fourth-generation or greater categories are defined in Table 1.

This paper uses the fourth-generation or greater group of respondents (henceforth fourth generation) as our assimilation benchmark because their families have been Americans for the longest period of time. Assimilation is the convergence of political opinions of each generation with those of fourth-generation Americans.

Measuring the Political Assimilation of Immigrants and Their Descendants We compare the political party identifications, ideological identifications, and policy preferences of first- , second- , third-, and fourth-generation Americans. The GSS asks only two questions about political party and ideological identification, so we included those responses in our review. For policy preferences, we selected the most controversial areas affecting government spending— such as levels of taxation and welfare spending—that divide Americans ideologically. These issues reveal more about potential differences among the groups in our study than would be the case of narrower questions such as how much should be spent on the space program, mass transit, or parks and recreation. We also check whether each generation’s opinions can be explained by random variation or if the differences vary in a statistically significant⁴ way from the fourth generation. If the variations of any generation can be explained by random chance compared to the fourth generation, then the opinions of the generation in question do not vary in a statistically significant way compared to the fourth generation. In other words, the political differences between the generations are very small or nonexistent if the comparison is statistically insignificant.

Political Party Identification
Figure 1 shows that first-generation immigrants are much more likely to be “independent,” about as likely to identify as “Democratic,” and less likely to identify as “Republican.” The first-generation’s differences in party identifications are statistically significant compared to the fourth generation because immigrants are much more likely to identify as “independent,” not because they are more likely to be Democratic. The political identification differences between the second- and third-generations compared to the fourth generation (see Figures 2 and 3) are so small that they are statistically insignificant.

Political Ideology
For political ideology, the small differences among each generation of immigrants and the fourth generation are statistically insignificant, as there is very little variation in ideological outlook. The ideological opinions of the first-generation are slightly more “liberal” than those of the fourth generation (see Figure 4). The second generation is more likely to be “moderate” and “liberal” (see Figure 5). The ideological differences between the third generation and the fourth generation are still statistically insignificant (see Figure 6). The third generation is virtually indistinguishable from the fourth generation.

Public Policies
This section analyzes responses from each generation on specific policy questions. The first question is whether the government should do more or less (see Figure 7). This very general question produces what appears to be a startling difference between the first generation and subsequent generations. To a statistically significant extent, the first generation is more likely to say that the government should do more. The responses for the second and third generations are statistically indistinguishable from the fourth generation. The first generation’s support for a more active government is noteworthy but general survey questions such as this are poor at gauging actual policy preferences. Questions about specific policies produce a more accurate picture of immigrant policy preferences.

Responses to the following specific policy questions reveal that immigrants and subsequent generations have policy opinions that are very similar to fourth-generation Americans on government spending on welfare programs (Figure 8), federal income taxes (Figure 9) government assistance to the poor (Figure 10), and Social Security benefit levels (Figure 11). For Figures 8-11, the responses by the first, second, and third generations all differ to a statistically insignificant extent compared to the fourth generation.

Conclusion
Immigrants have very similar political and policy views when compared to fourth-generation Americans. For those policy and political questions where first-generation immigrants have statistically significant differences of opinions, the second and third generations do not vary compared to the fourth generation, meaning that political assimilation is complete by the second generation. Immigrants are rapidly assimilating into American political life without upsetting the current ideological and political balance. The rapid pace of political and ideological assimilation of immigrants and their children should assuage the concerns of those who oppose immigration reform for this reason.

Notes
1. See Phyllis Schlafly, “Amnesty Is Republican Party Suicide,” Townhall, October 29, 2013, http://townhall.com/columnists/ phyllisschlafly/2013/10/29/amnesty-is-republican-party-suiciden1732723/page/full; Eagle Forum, “How Mass (Legal) Immigration Dooms a Conservative Republican Party,” Eagle Forum Report, January 2014, http://www.eagleforum.org/immi gration.html; John Fonte and Althea Negai, “America’s Patriotic Assimilation System Is Broken,” Hudson Institute Culture and Society Briefing Paper, April 2013, http://www.hudson. org/files/publications/Final04-05.pdf; and James G. Gimpel, “Immigration’s Impact on Republican Political Prospects, 1980 to 2012,” Center for Immigration Studies Backgrounder, April 2014.
2. The General Social Survey is only conducted in English, which might bias the results for the first-generation because English speakers could have political opinions that vary significantly from non-English speakers. This potential concern does not affect the second, third, and fourth generations because they are either universally English-fluent or near to it. See Pew Research Center, “Second-Generation Americans: A Portrait of the Adult Children of Immigrants,” Pew Research Center, February 7, 2013, p. 49.
3. See Brian Duncan and Stephen J. Trejo, “Tracking Intergenerational Progress for Immigrant Groups: The Problem of Ethnic Attrition,” American Economic Review 101, no. 3 (2011).
4. We measure statistical significance through a Pearson ChiSquare test at 95 percent confidence.