Gallup released a poll last Wednesday that claims 52 percent of United States residents over the age of 18 support the impeachment and removal from office of President Donald Trump.
“52% of Americans now support Trump’s impeachment and removal,” a story highlight of the Gallup News article published on October 16 claimed.
But Gallup has confirmed to Breitbart News that it does not know how many illegal aliens or non-citizen legal residents were among the 1,526 respondents questioned in the poll conducted between October 1 and October 13. Gallup also confirmed that it does not know how many of the poll respondents are registered voters.
As a result, it appears that the Gallup Poll released last Wednesday, which has been widely reported in the mainstream media as an indicator that public support for impeachment is on the rise, dramatically overstates the percentage of American citizens–and particularly American citizens who are registered voters–who support impeachment.
Breitbart News asked Gallup to provide demographic data of poll respondents in these seven key areas:
1. Number of respondents who are illegal aliens
2. Number of respondents who are not US citizens but are legal residents
3. Number of respondents who are US citizens but not registered voters
4. Number of respondents who are US citizens and registered voters
5. Number of respondents who responded in Spanish
6. Number of Hispanic respondents
7. Number of African-American respondents
“For clarification, we did not ascertain respondents’ status for the items referenced in your questions 1 [number of respondents who are illegal aliens] or 2 [number of respondents who are not US citizens but are legal residents] . We also did not ask about registered voter status, but we do include that from time to time, particularly during election years,” Lydia Saad, Director of U.S. Social Research for Gallup, told Breitbart News in an email received on Tuesday.
Gallup’s Saad added that questions about the citizenship of respondents, as well as their voter registration status “are not standard demographics.”
“We can answer your questions #5-7 based on the standard demographic information collected in our Oct. 1-13 poll. The other questions are not standard demographics, and were not included in this poll,” Saad told Breitbart News in an email received on Tuesday:
Please note that I’m providing the figures based on weighted data so they accord with the sample composition on which the reported findings are based. (Weighting targets reflect 2018 CPS/Census targets.)
5. Number of respondents who responded in Spanish: 86 respondents, or 5.6% of the sample
6. Number of Hispanic respondents: 249 respondents, or 16.3% of the sample
7. Number of African-American respondents: 193 respondents (non-Hispanic blacks), or 12.6% of the sample
The 1,526 respondents to the poll conducted between October 1 and October 13 apparently included an unknown number of illegal aliens currently residing in the United States as well as non-U.S. citizens who are legal residents, in addition to U.S. citizens who are not registered to vote, as well as U.S. citizens who are registered to vote.
Gallup described the poll’s methodology in a statement released last week:
Results are based on telephone interviews conducted October 1-13, 2019 with a random sample of –1,526— adults, ages 18+, living in all 50 U.S. states and the District of Columbia. For results based on this sample of national adults, the margin of sampling error is ±3 percentage points at the 95% confidence level.
Interviews are conducted with respondents on landline telephones and cellular phones, with interviews conducted in Spanish for respondents who are primarily Spanish-speaking. Each sample of national adults includes a minimum quota of 70% cell phone respondents and 30% landline respondents, with additional minimum quotas by time zone within region. Landline and cell phone telephone numbers are selected using random digit dial methods. Gallup obtained sample for this study from Dynata. Landline respondents are chosen at random within each household on the basis of which member has the next birthday.
Samples are weighted to correct for unequal selection probability, non-response, and double coverage of landline and cell users in the two sampling frames. They are also weighted to match the national demographics of gender, age, race, Hispanic ethnicity, education, region, population density, and phone status (cell phone only/landline only/both and cell phone mostly). Demographic weighting targets are based on the March 2018.
Current Population Survey figures for the aged 18 and older U.S. population. Phone status targets are based on the January-June 2018 National Health Interview Survey. Population density targets are based on the 2010 census. All reported margins of sampling error include the computed design effects for weighting.
Gallup also apparently heavily skewed the weighting of sample respondents–a sample that most likely included illegal aliens and non-citizen legal residents–by dramatically decreasing the percentage of Republicans included in the weighted final results that showed 52 percent support impeachment, while dramatically increasing the percentage of Democrats included in those weighted results.
The raw data used by Gallup from respondent answers in compiling its weighted results can be seen here.
The actual number of self-identified Republicans who were surveyed in the unweighted sample of 1,526 respondents was 462, or 30.2 percent of the actual number of respondents. Only six percent of Republicans favored impeachment and removal of the president. Gallup weighted those results downward so that each Republican who responded to the survey was only counted as 85 percent of a respondent, so only 393 Republicans were included in the “weighted” sample of 1,526, which made up 25.7 percent of the weighted sample.
In contrast, the actual number of self-identified Democrats who were surveyed in the unweighted sample of 1,526 respondents was 419, or 27.5 percent of the actual number of respondents. A whopping 89 percent of Democrats favored impeachment and removal of the president. Gallup weighted those results upward so that each Democrat who responded to the survey was counted as 107 percent of a respondent, so 450 Democrats were included in the “weighted” sample of 1,526, which made up 29.5 percent of the weighted sample.
The actual number of self-identified Independents who were surveyed in the unweighted sample of 1,526 respondents was 612, or 40.1 percent of the actual number of respondents. A majority– 55 percent of Independents favored impeachment and removal of the president. Gallup weighted those results upward so that each Democrat who responded to the survey was counted as 107 percent of a respondent, so 655 Independents were included in the “weighted” sample of 1,526, which made up 42.9 percent of the weighted sample.
Although Gallup did not provide the actual number of respondents in the unweighted sample who answered “yes” to the question “Should the president be impeached and removed from office?” a simple calculation suggests that the number was 737 out of 1,526, or 48 percent, which is four percent less than the 52 percent reported in the weighted sample.
Similarly, the Gallup Poll weighted respondents by race significantly.
In the unweighted sample, 23.4 percent of respondents, or 359, were Non-White, while 73.9 percent, or 1,132, were White.
In the “weighted sample, however, 32.7 percent of respondents, or 502, were Non-White, while 65.2 percent, or 1,000 were White. (30 respondents, or two percent, were not classified by race.)
The response provided by Gallup to the Breitbart News question of the number of Hispanics and African Americans included among the sample of poll respondents appears to be consistent with this Non-White sampling. Hispanic respondents in the weighted sample totaled 249 (16.3 percent), while African-American respondents in the weighted sample totaled 193 (12.6 percent). By simple mathematical calculations, there were 60 “other” Non-Whites in the weighted sample (3.9 percent), bringing the total to 502 out of 1,526 or 32.7 percent of the weighted sample.
According to the poll (using the “weighted” sample), 68 percent of Non-Whites answered “yes” to the impeach and remove question, while 44 percent of Whites answered yes, resulting in an overall 52 percent answering “yes.”
The poll results using samples unweighted by race–but still including an unknown number of illegal alien and non-citizen legal resident respondents– yielded a different result: 49 percent answering “yes,” as opposed to 52 percent.
Notably, according to Pew Research, “Black Hispanic, Asian, and other racial ethnic groups accounted for 26.7 percent of all votes cast in 2016.” By that standard, the Gallup Poll overweights Non-Whites by six percent.