Harris Poll/BYU Study Finds More People would be Angry or Upset if Mitt Romney was President than if Barack Obama or Hillary Clinton was President Using Experimental Survey Design to Gauge Underlying Voter Sentiments

As the race towards the White House heads into the February 5^th, super duper Tuesday primaries, there are a lot of unanswered questions which linger. Will Americans vote for a female president? A black president? A Mormon president? Utilizing an experimental survey design called a "list experiment", the answers to these questions may be a little clearer. The most striking is that just under half (47%) of Americans would be upset or angry if Mitt Romney was serving as President, one-third (33%) of adults would be angry or upset if Barack Obama was serving as President and just under this (30%) if Hillary Clinton was serving as President.

Granted, a great deal of this may have to do with the individuals themselves, so how do the American people respond to the various characteristics? Well, just over one-quarter (27%) would be angry or upset if a Mormon was serving as President and almost the same number (25%) if a woman was serving as President. A very small number, just under one in ten Americans (9%), would be angry or upset if a Black person was serving as President.

These are the results of a nationwide Harris Poll of 2,302 U.S. adults surveyed online between January 15 and 22, 2008 by Harris Interactive^®. This survey was conceived and developed by Harris Interactive and was not commissioned by any organization. Harris Interactive worked closely with the Center for the Study of Elections and Democracy at Brigham Young University in developing these questions and experiment and then in the analysis of the results.

When the pairings between the candidate and their personal characteristic are compared to each other, it seems that having a Mormon as President does not make people as angry or upset nearly as much as having Mitt Romney serving (27% versus 47%). The same situation is true regarding a Black person or Barack Obama serving as President as more say Obama serving would make them angry or upset compared to a Black person serving (9% versus 33%). Whether Hillary Clinton or a woman is serving does not seem to make much of a difference as the just slightly more people say having her specifically serve would make them angry or upset compared to the generic woman (25% versus 30%).

One of the main factors in the Democratic nomination fight these past few weeks has boiled down to a question of race. Who will White Democrats and Black Democrats vote for? When race is looked at, the numbers do not change substantially. Only one in ten (10%) of non-white adults are angry or upset at the notion of a Black person serving as President with 13 percent of White adults feeling the same way — an imperceptible change. However, when the idea of Barack Obama serving as president is examined, dramatic differences between Whites and non-Whites emerge. No non-white adult is angry or upset at the idea of Barack Obama serving as President compared to two in five (41%) of White voters.

While not as much in the forefront as race, this election has also seen the issue of gender emerge. Among men, gender is an issue. Just over two in five men (43%) say the idea of a woman serving as President makes them angry or upset while 48 percent of men are upset or angry about the prospect of Hillary Clinton serving as President. As might be expected, gender doesn’t appear to be as serious an issue among women. Only 12 percent of women say a woman serving as President would make them angry or upset while 21 percent of women say Hillary Clinton serving as President would make them angry or upset.

When it comes to these issues, Republicans and Democrats, not surprisingly, have some major differences in how angry or upset these all make them. When it comes to a black person or woman serving as president, Democrats are fine with either, while Republicans are much more likely to say a woman would make them angry. Just one in ten (11%) of Republicans say a Black person serving as President would make them angry or upset while 16 percent of Democrats say the same. There is a much larger difference when it is a woman serving as President as one in five Democrats (20%) say that would make them upset or angry while two in five (41%) Republicans say that.

When the actual candidates themselves are examined, there is almost no difference for Democrats between Senators Clinton and Obama. One-quarter (25%) of Democrats say Hillary Clinton serving as President would make them angry or upset while almost the same number of Democrats (23%) say Barack Obama serving as President would make them angry or upset. Republicans are also of similar mind as just under half (48%) say Barack Obama serving as President would make them angry or upset while over half (56%) of Republicans say Hillary Clinton serving as President would make them angry.

On religion, Republicans are not angry or upset about a Baptist serving as President while one-quarter (25%) of Democrats are, although the difference is not statistically significant. When it comes to a Mormon serving as President, Republicans are slightly more likely than Democrats to be angry or upset (34% versus 27%). A larger and more significant difference is seen if Mitt Romney was serving as President – 41 percent of Democrats would be angry or upset while 51 percent of Republicans would be angry or upset. When it is Mike Huckabee serving as President, only one in ten Republicans (9%) would be angry or upset while three in ten (30%) Democrats would feel the same way.

One of the more challenging issues pollsters have had to grapple with this year is the sensitive nature of the various characteristics of the candidates running for President. Is someone’s race, gender, or religion really an issue? When asked directly, the answer tends to be no – of course the fact that someone is Black or a woman is not a problem. But is this the case when the voters are actually voting? Using this list experiment analysis, the answer may be a little clearer. First, race doesn’t appear to be an issue whether asked indirectly in a list experiment or more directly. When asked directly only 14 percent of voters say they would not vote for an otherwise qualified Black person compared to 9 percent who are angry or upset at a black person serving as president. Religion appears to be something that voters reveal preferences on when asked directly. A Baptist candidate encounters relatively little resistance when asked directly (15%) or in the list experiment (15%) and the percentage unwilling to vote for a Mormon (31%) is comparable to the percent angry or upset about a Mormon serving as president (27%). Gender fits between race and religion. Americans are not likely to say they won’t vote for a woman when asked directly (16%), but the list experiment shows that 25% of voters, have trouble with a woman serving as president and that this percentage goes up dramatically among men (43%) and Republicans (41%)

Many surveys have asked voters directly about supporting an otherwise qualified candidate who is a woman, black, Mormon, Baptist or any other number of traits. However, asking about a candidate directly allows a voter in a survey to hide any reservations about that candidate especially if the trait is sensitive. One way to unmask these potentially hidden sentiments is through an experimental survey design called a "list experiment." Using a list experiment, attitudes that are potentially socially unacceptable are measured without tipping off the survey respondent that their hidden sentiments are being explored.

A list experiment relies on randomly assigning participants in a survey to either a control group or one of several treatment groups (eight in this case). All of the groups were asked the following: "Below are four/five things that sometimes make people angry or upset. After you read all four/five statements, please indicate how many of them upset you. We do not need to know which ones upset, just how many." The four item list, the control group was shown the following list:

• The way gasoline prices keep going up;
• Professional athletes getting million dollar plus salaries;
• Requiring that seat belts be used when driving;
• Large corporations polluting the environment.

The eight treatment groups were all shown the same four items as above and each had one additional statement added to the list. The eight statements tested were:

• A Mormon serving as President;
• Mitt Romney serving as President;
• A woman serving as President;
• Hillary Clinton serving as President;
• A black person serving as President;
• Barack Obama serving as President;
• A Baptist serving as President;
• Mike Huckabee serving as President.

The idea is to see if any of the treatment statements causes people to be angry or upset. In order to analyze the results, a difference is calculated by taking the average number of items for the control group and comparing it to each treatment group. Because the survey respondents are randomly assigned to the treatment or control group, any perceptible difference between the two can only be attributed to the additional item added to the list for the treatment group. The difference between the treatment and control group represents the percentage of people who were angry or upset by the additional item.

"Below are four/five things that sometimes make people angry or upset. After you read all four/five statements, please indicate how many of them upset you. We do not need to know which ones upset you, just how many."

Base: All adults

Note: For this list experiment, our sample was broken into nine different groups. One group of respondents was shown the first four statements as a control group. The other eight groups were each shown the same four statements and then ONE additional statement. This table shows the four control statements and then the percent of adults made angry or upset by the additional of the fifth statement for each of the additional statements.

* Not statistically significant at the 95% confidence level.

"Below are four/five things that sometimes make people angry or upset. After you read all four/five statements, please indicate how many of them upset you. We do not need to know which ones upset you, just how many."

Base: All adults

* Not statistically significant at the 95% confidence level.

"Below are four/five things that sometimes make people angry or upset. After you read all four/five statements, please indicate how many of them upset you. We do not need to know which ones upset you, just how many."

Base: All adults

* Not statistically significant at the 95% confidence level.

"Below are four/five things that sometimes make people angry or upset. After you read all four/five statements, please indicate how many of them upset you. We do not need to know which ones upset you, just how many."

Base: All adults

* Not statistically significant at the 95% confidence level.

"If you thought he or she was otherwise well qualified, would you vote for any of the following who was running for President of the United States?"

Base: All adults

Note:    Percentages may not add up to 100% due to rounding     

This Harris Poll^® was conducted online within the United States January 15 and 22, among 2,302 adults (aged 18 and over). Figures for age, sex, race/ethnicity, education, region and household income were weighted where necessary to bring them into line with their actual proportions in the population. Propensity score weighting was also used to adjust for respondents’ propensity to be online.

All sample surveys and polls, whether or not they use probability sampling, are subject to multiple sources of error which are most often not possible to quantify or estimate, including sampling error, coverage error, error associated with nonresponse, error associated with question wording and response options, and post-survey weighting and adjustments. Therefore, Harris Interactive avoids the words "margin of error" as they are misleading. All that can be calculated are different possible sampling errors with different probabilities for pure, unweighted, random samples with 100% response rates. These are only theoretical because no published polls come close to this ideal.

Respondents for this survey were selected from among those who have agreed to participate in Harris Interactive surveys. The data have been weighted to reflect the composition of the adult population. Because the sample is based on those who agreed to participate in the Harris Interactive panel, no estimates of theoretical sampling error can be calculated.

J33083

Q600

The center for the Study of Elections and Democracy (CSED) at Brigham Young University is a nonpartisan academic research center seeking to increase knowledge about the practice of American democracy. CSED is committed to the production and dissemination of research that meets high academic standards, is useful to policy makers, and informs citizens. Founded in 1998, CSED has enjoyed significant success securing grants from foundations and other sources totaling several million dollars. CSED-sponsored research has been published in leading academic journals and presses in the areas of campaign finances, voting technology and election reform, presidential and congressional elections, religion and politics and democratic deliberation. CSED scholars frequently provide expert commentary on national and local politics in areas related to their research. The views expressed in this release do not necessarily represent the views of Brigham Young University and its sponsoring institution. More information about CSED can be found at http://csed.byu.edu.