In the late 1960s, President Nixon popularized the term silent majority. Nixon defined the silent majority as mainstream Americans who did not join protests against the Vietnam War nor participate in public discourse. Preoccupied with a vocal minority, the news media overlooked the silent majority.
President Trump believes there is a silent majority as well. And, he thinks that majority will carry him to victory.
This is wishful thinking. Trump’s presidential approval ratings fall short of a majority. And his supporters are strident and politically engaged – they are not silent.
However, Trump may be right that some people are wary about expressing their support publicly. After all, he is a polarizing and controversial figure. Are people really hiding their support for Trump?
Let’s take a look.
We know people conceal their beliefs in order to please others and win social approval. Researchers call this the social desirability effect. In opinion surveys, people select answers they believe are more acceptable, in-line with popular viewpoints, or politically correct instead of choosing answers that reflect their true feelings.
For example, when asked if they voted on Election Day, non-voters often claim they did vote. They are uncomfortable revealing their abstention. Similarly, for candidate preference, voters may declare support for an African American candidate when in fact they intend to vote for a White candidate – fearful that admitting their true preference could draw criticism of racial motivation.
In a nutshell, to avoid appearing unacceptable or unusual, people report behaviors and opinions that draw social approval. We are social beings and prefer acceptance to exclusion.
If people are reluctant to disclose unpopular behaviors and attitudes, they may well conceal support for a divisive figure like President Trump. After Trump’s surprise 2016 victory, analysts considered this very possibility.
Since that time, Trump’s navigated an array of disruptive investigations, advanced contentious immigration policies, survived contentious impeachment hearings, and absorbed widespread criticism for handling the pandemic and his response to racial unrest.
Given these circumstances, a strong stigma to admitting support for Trump should exist.
Automated voice versus live call surveys
One way to examine the existence of stigma is to compare public opinion polls that use different data collection methods. The firm Rasmussen uses pre-recorded automated phone methodology. Respondents answer questions after hearing a digitally-recorded voice. Absent a human interviewer, social pressures disappear.
Most other organizations, including CNN, use trained phone interviewers. With an actual person asking questions, social desirability may influence answers.
Therefore, during an automated survey, Trump fans should feel relatively free to disclose their true preferences – leading to higher approval ratings. Responding to a live call, however, Trump supporters are less likely to express approval and may even refuse to answer. The graph below illustrates the point.
Rasmussen’s automated polls consistently show higher Trump approval than CNN’s live calls. [i] In fact, Rasmussen’s early June survey demonstrates Trump’s approval 8% higher than CNN.[ii] In addition, the percentage of Rasmussen respondents that refused to answer the approval question is much smaller than CNN refusals – 4 to 5% of CNNs respondents refused compared to 1% for Rasmussen. In other words, instead of disclosing a stigmatized preference during live calls, Trump supporters simply refused to answer.
While there is not a silent majority, there appears to be a small group that conceals its support. In surveys that remove social pressures, Trump approval increases. This effect, though small, may be important in swing states with razor-thin margins.
A pollster distinguished for accurately predicting Trump 2016 victories in Michigan and Pennsylvania, recently claimed that social desirability effects are more severe today than four years ago. If the pollster is correct, Democrats should examine polls critically, acknowledging Trump’s level of support may be higher than reported.
And, Republicans should keep their collective fingers crossed that social desirability does in fact mask Trump’s actual level of support.
[ii] Besides mode of data collection, “house” effects which include sample selection, question wording, and weighting can in part explain the differences shown in the Figure. See Charles Franklin, Survey mode and presidential approval for details.