USIPC/NPNA Pre-Election Naturalized Citizen Voter Survey
Naturalization does not guarantee a feeling of security.
More than half of these registered naturalized citizens say they feel less secure in their citizenship than they did a few years ago. The concern is not evenly shared across communities.
How secure do naturalized citizens feel today?
Responses about citizenship security compared with a few years ago
Citizenship insecurity and views of enforcement are strongly associated
Share in each citizenship-security category who say enforcement has gone too far
Hispanic respondents report markedly greater citizenship insecurity
Percent feeling less secure and percent saying immigration enforcement has gone too far, by race/ethnicity.
Large majorities oppose broad immigration enforcement and citizenship restrictions.
Most respondents oppose deporting all undocumented immigrants, raids at schools, hospitals, and churches, and expanding efforts to revoke citizenship from naturalized citizens.
Has the current administration’s enforcement gone too far?
What should happen to undocumented immigrants living in the U.S.?
Three distinct policy approaches
Large majorities oppose mass deportations and raids at sensitive locations
Share supporting versus opposing each policy, combining strongly and somewhat responses
On citizenship policies, respondents favor protection over restrictions
Support and oppose sum to 100% because the fielded scale has four responses
Economic pressure is pervasive.
Food, housing and health costs rank at the top of voters’ stated issue priorities. The survey also asks about difficulty paying bills, who is responsible for prices and which party voters trust to address those problems.
What is extremely or very important when deciding how to vote?
Ranked by combined “extremely” or “very important” responses; respondents rated each issue separately.
Household payment strain
report trouble paying at least one of the listed expenses in the past year.
How family finances compare with a year ago
Actual household payment difficulties in the past year
Multiple responses allowed; percentages are shares of all respondents.
Who do voters blame for prices?
Two related measures: “a great deal” and “a great deal or some.”
Which party do naturalized voters trust on major issues?
Four-way responses, not a two-party-only measure
Each issue was rated separately. Both equally and neither remain in the denominator.
The financial squeeze differs across communities
Percent saying family finances have worsened
A large pool of undecided voters remains even as leaners are counted.
The unleaned generic congressional ballot and its undecided follow-up tell different, complementary stories. Voting plans are survey responses—not an election turnout forecast.
Initial generic congressional preference
How initially undecided voters lean
Only respondents initially undecided on B1; unweighted n=250
One in four initially undecided; leaners narrow the pool
Unleaned and leaned generic congressional ballot
Two columns use the same 1,000 respondents. “Including leaners” is derived from B1 and the B2 follow-up; it is not an election projection.
Self-reported intention to vote
Self-reported intentions include respondents saying they have already voted; not a turnout forecast
Voting motivation compared with past elections
Who has contacted these registered voters about voting?
Respondents could report multiple contact sources; the “no one” response is separate.
The voter-to-voter comparisons—including 2024 presidential recall, 2026 House preferences, and undecided leaners—appear in Section 6: 2024 to 2026.
Dissatisfaction extends beyond any single question.
Ratings of the president, the direction of the country and the political parties provide context for how registered naturalized citizens are approaching November.
Approval of President Trump
Direction of the country
Presidential job approval by issue
Percent approve and disapprove of handling of five policy areas.
Favorability of parties and congressional parties
The residual group selected “never heard of / can’t rate.”
Party identification
Self-described political ideology
The Democratic advantage dipped in 2024 — and rebounds sharply in 2026.
Democratic margin (D+) among naturalized citizens, 2022–2026
The Democratic advantage fell sharply between the 2022 midterm and the 2024 presidential election. The USIPC/NPNA 2026 pre-election survey finds that it rebounds in current House preferences.
A varied naturalized electorate—and an important weighting distinction.
The charted shares below use the final survey weights. These show the composition of the survey estimates, not new independent Census benchmarks.
Weighted race/ethnicity composition
Derived, mutually exclusive race5 variable; Hispanic of any race, others non-Hispanic
Census region composition
From the derived region variable
Age of naturalized registered-voter respondents
Verified directly from raw age, not the mislabeled ranges in the preliminary methods draft
Languages most often spoken at home
Most frequently reported countries or birthplace categories
Share of weighted sample; the dataset contains aggregated “Other” birthplace categories
Where naturalized voters get political news
Multiple responses allowed; percentages sum to more than 100%.
About the survey
The USIPC / NPNA Pre-Election Naturalized Citizen Voter Survey contains 1,000 foreign-born naturalized U.S. citizens age 18 or older who reported being registered to vote. Interviews were conducted from October 1–11, 2026 using a mixed-mode sample. Respondents were recruited via online panels, text-to-survey, and traditional phone interviews. The sample includes a 200-person Hispanic oversample and 100-person Asian oversample. Weights were calculated using iterative proportional fitting ("raking"). Target proportions were determined using the American Community Survey (ACS). Separate weights for registered voters were calculated using the Current Population Survey (CPS) Voting and Registration Supplement and additional calibration of recalled 2024 presidential vote (51% Harris and 47% Trump).
Weight range 0.205–2.515; weighting design effect 1.20; effective N 832.
Question wording is available by hovering over findings on desktop. Some results are multi-select; their sums can exceed 100%.