April 2, 2019

Wason Center National Survey of 2020 Voters

"This survey is of 1,001 likely 2020 presidential election voters drawn from a nationally representative random probability sample using voter file sampling. A full description of our rigorous, AAPOR-standard methodology can be found after the analysis of and the demographic breakdown of survey respondents appears before the methodology."

Summary of Key Findings

1. Provides an overview of partisanship of the American electorate in the 2020 cycle and offers partisanship with Independent "leaners" sorted into their respective parties.
2. Provides an overview of the ideological distribution of the 2020 electorate with a new division of moderates into "lean conservative" and "lean liberal" categories.
3. Provides a breakdown of partisanship of the 2020 electorate, by nuanced ideology.
4. Results of 2020 general election ballot question between President Trump and the Democratic Party's nominee, as well as the vote breakdown by party and by ideology.
5. Results of 2020 general election ballot question with an Independent candidate included on the ballot, as well as vote breakdown by party and by ideology.
6. Presents data visualizations (word clouds) of voters' one-word, open-end responses to the question, "What is the first word that comes to your mind when you hear _____?" Voters were asked to provide answers for the following prompts: Congress, the Supreme Court, Republican, and Democrat. Data for Congress and the Supreme Court is presented for all respondents only. Data for Republican and Democrat are presented for all voters, Democrats on Democrats, Democrats on Republicans, Republicans on Republicans, Republicans on Democrats, and pure Independents on both Republicans and Democrats.
7. Presents findings from the question, "Our nation's major news organizations publish fake news stories for political purposes."
8. Presents findings from the question, "Views of Democrats/Republicans threaten the well-being of the country."
9. Presents findings from the question, "The blame for America's dysfunctional politics falls as much on my own political party, or party I most often vote for or agree with, as the other political party."
10. Presents findings from the question, "When I vote, I am very concerned with keeping the other party's candidate from winning."
11. Presents findings from a supplemental experimental national survey* that measures Americans' responsiveness to partisan cues in terms of their policy preferences. *Data for this analysis comes from the Wason Center Political Attitudes Experimental Survey. Survey demographics and methodology can be found after the National Survey demographics and methodology at the bottom of the page.

Analysis

A durable finding from political science research is that people who self-identify as “Independent” but who admit to leaning towards one party or the other in a follow-up question behave very similarly to their partisan counterparts. Increasingly, research is treating these Independent “leaners” as soft partisans and distinguishing “pure” Independents from them. This is important because Independent “leaners” account for the majority of people who fall into the Independent category and including leaners in the Independent category may create a distorted picture of how fluid the American electorate really is.

The vote preferences of Independent “leaners” demonstrate how closely these voters mirror the preferences of their partisan counterparts, with the relationship especially strong for Democrats. 86% of Republicans and 96% of Democrats indicate they plan to vote for their respective party’s nominees. Although Republican leaners are less supportive of Trump (60%) than Democratic leaners are of the Democrat (78%), in general, leaners are indistinguishable from their partisan counterparts.

Despite the high levels of polarization and partisan acrimony in American politics, American voters still largely consider themselves ideologically moderate, with 51% of voters describing their ideology this way. Digging deeper, moderates are evenly split, with about half leaning liberal and half leaning conservative. Similar to previous studies, “conservative” is a far more robust response than its counterpart “liberal.” While 21% of voters self-identify as conservative, just 9% self-identify as liberal. Just 9% of voters consider themselves to be strong conservatives while just 7% of voters consider themselves to be strong liberals.

Another stable finding from survey research are the differences in the ideological compositions of the Republican and Democratic Party coalitions. Republican voters are more ideologically robust, with the modal category for Republicans being “conservative.” Democrats have always been less liberal, or at least less willing to think of themselves as liberal. Here, voters who choose the moderate category are also forced to choose whether they lean towards liberalism or conservatism. As such, for Democrats the modal category in these data contains the word “liberal.” However, the way ideology is conventionally coded, moderate is not distinguished in this way and if you compare these data with other data you will see that the plurality, if not majority, of Democrats describe their ideology as “moderate.” One of the most interesting findings here is that despite being offered a less robust category, Republicans still prefer to describe their ideology as “conservative” although more Republicans populate the “moderate, lean conservative” category than you might find in other surveys where the moderate category is not structured this way. There are three other important findings from this analysis. After lagging for the last two decades the percent of voters who consider themselves “strong liberals” is beginning to catch up to the percent who identify as “strong conservatives.” Second, although they continue to decline, 8% of Democrats describe their ideology as “conservative.” Finally, most pure Independents identify as moderates with “lean conservative” outnumbering “lean liberal” by about 7 points.

Heading into the Democratic Party’s primary season, President Trump is at an 11-point disadvantage against his generic Democratic Party opponent, a margin well outside the survey’s +/- 3.2 point margin of error. More importantly, Trump would need all voters who currently report being undecided (10%) to break in his favor in order to close the gap. Should he run for reelection, President Trump will do so facing the most hostile electoral environment an incumbent president has faced in the modern era. His predecessors, Barack Obama, George W. Bush, Bill Clinton, George H.W. Bush, and Ronald Reagan began their reelection campaigns with better approval ratings, and without the complications of scandals. All were reelected except the senior Bush, who lost to Bill Clinton despite approval ratings that averaged in the mid-50s throughout the election year. Perhaps the closest analogy to Trump in terms of approval ratings is Jimmy Carter, whose average approval rating for his term was 45%. Carter, of course, lost his reelection bid. Trump’s average approval rating for his first 27 months in office has ranged between the high 30s and the mid-40s. President Trump will be the least popular president to run for reelection in the history of polling.
Driving the 11-point advantage for the Democrats on the two-candidate ballot is the nearly universal support for the party’s nominee from Democrats as well as a sizable advantage among pure Independents. While Trump is drawing in just 79% of Republicans (with Independent leaners included), the generic Democrat is receiving 91% of the votes from Democrats (with Independent leaners included) as well as 39% of the vote from pure Independents. Interestingly, there is robust support for a 3rd party option among pure Independents, suggesting that a significant portion of this voting bloc does not support Trump, but isn’t keen on supporting the Democratic Party’s nominee, either.

The future Democratic Party nominee enjoys uniform support among the party’s most ardent ideological supporters, but President Trump has a small faction of strong conservatives planning mutiny. 12% of voters who describe their ideology as strongly conservative intend to vote for the Democratic Party’s nominee in what one can only assume is a protest vote. There are also signs of unrest within the Republican’s ideological coalition in the moderate, lean conservative camp. Although a plurality of these voters (42%) intend to vote for Trump, 19% say they will vote for the Democrat and another 26% report being undecided. 12% of liberals also report being undecided.

Although the Democratic Party’s nominee holds a large advantage over President Trump in a two-person contest, the introduction of an Independent candidate completely reshapes the race, making it far more competitive for President Trump. This is a product of two things. First, for every one voter that moves their vote choice away from Trump to the Independent candidate, five move their vote away from the Democratic Party’s nominee to the Independent. Overall, the Independent candidate earns support from 16% of voters, and increases the number of undecided voters considerably, from 10% on the two-candidate ballot to 16% with an Independent in the race.

Defection to the Independent candidate comes primarily from Democrats and pure Independents. On the two-person ballot question, 39% of pure Independents broke for the Democratic Party’s candidate compared to just 5% for President Trump. 12% of voters were undecided and 44% of voters chose the “third party” option. The introduction of an Independent candidate reveals that almost all of the robust support for the third party preference on the first ballot question is actually for an Independent third party. With an Independent candidate as a distinct option, 42% of pure Independents choose the Independent candidate and the third party option drops to 0%. The introduction of an Independent candidate option also increases the number of pure Independents who are undecided (27%).

Although moderates are the most likely voter group to support an Independent candidate, the Independent candidate draws support from across the spectrum, which may reflect the undefined nature of the question’s phrasing. The modal category for support for the Independent is moderates who lean conservative. 28% of these voters chose the Independent and an equal percentage report being undecided in this scenario. Interestingly, an Independent candidate receives significant support across all three liberal categories, including strong liberals. This illustrates why the electoral picture for Democrats changes significantly when an Independent candidate is added to the ballot. 15% of strong liberals report they would support the Independent, compared to 22% of liberals and 19% of moderate, lean liberal voters. These numbers suggest the Democrats may be vulnerable at both ends of their ideological spectrum. Choose a nominee who is too far to the left, and moderates and liberals may be susceptible to an Independent candidate, but choose a candidate that is too centrist, and progressive Democrats may find themselves looking for an alternative. Conversely, ideological conservatives are far less receptive to an Independent candidate. Conservative is the modal ideological category for Republican voters.

For the following questions, voters were asked for the first word that popped into their heads for Democrats, Republicans, and Congress. The Congress frequency distribution clearly reflects the frustration Americans have been expressing with Congress in surveys over the polarized era. Voters use a variety of different descriptors - dysfunctional, ineffective, useless, corrupt, lazy, incompetent - to describe the same thing, a broken Congress.

The most common word used is dysfunctional, appearing 33 times. All told, 61% of the words used by voters about Congress are negative.

**Note: Column 1 response is "Dysfunctional", column 7 response is "Government"

Voters clearly associate Democrats with liberalism. The word liberal is by far the most dominant word in their frequency distribution. Voters also associate the word Democrat with Obama and Clinton in nearly equal measure. Socialism, which is typically an attack word in American politics, is the most frequently mentioned negative term.

**Note: Column 10 response is "Republican"

The frequency distribution created with words Democrats use to describe themselves is clearly distinguishable from the frequency distribution made with words from Republican voters, although "liberal" is prominent in both. Unlike the frequency distribution from Republicans describing themselves, Democrats use a variety of words to describe themselves including two different terms to describe ideology - liberal and progressive. Clinton, socialism, liar, and liberal are dominant in the frequency distribution made from Republicans describing Democrats.

**Note: Column 10 response in Democrats on Democrats is "progressive"

Pure Independents use a vocabulary all their own to describe Democrats. Yes, liberal makes an appearance, but it is not the dominant word. Instead, pure Independents rely on the opposite term, Republican, when prompted with the word Democrat. Pure Independents are also more likely to associate Democrats with their Speaker of the House, Nancy Pelosi, than with Clinton or Obama.

**Note: Column 2 response is "Republican", column 8 response is "Government"

The Republican frequency distribution shows that voters identify the word Republican both with the party’s ideology, conservatism, and with Donald Trump. Combined, these two responses make up more than a third of all responses.

**Note: Column 1 response is "Conservative"

Although Trump is a popular response among Republicans, Republicans clearly associate conservative ideology most closely with their party.

**Note: Column 1 response is "Conservative"

Although Trump and conservative are the most common words associated with the word Republican by Democrats, Democrats rely on a variety of negative words such as racist, bad, and greedy. Others are merely descriptive such as “red,” the party’s electoral map color, or politics.

**Note: Column 2 response is "Conservative"

Like their Republican and Democratic counterparts, pure Independents strongly associate Republicans with conservatism and with Donald Trump. Again, pure Independents rely on the opposite when prompted with the word Republican, returning with Democrat. The word racist is the most frequently used negative word.

**Note: Column 2 response is "Conservative", column 10 response is "Government"

The most frequently used word to describe the Supreme Court describes the Court’s position as the judicial branch in the separation of powers system. 236 voters responded with this word for the Supreme Court prompt, about a fourth of the voters. Law and judge were also common.

Because of the proximity of the survey to the Kavanaugh confirmation hearing the previous fall, Justice Kavanaugh was used 24 times and Justice Ginsburg was mentioned by 29 voters. Ginsburg has been featured in both a Hollywood movie based on her life and a popular Netflix documentary this year.

**Note: Column 4 response is "Conservative", column 5 response is "Kavanaugh", column 7 response is "Constitution"

President Trump’s rhetoric regarding the media appears to be having a massive impact on voter confidence in the media. Here we ask voters to agree or disagree with the following statement: “Our nation’s major news organizations publish fake news stories for political purposes.” We were quite intentional about the wording. We wanted to be sure respondents understood we were talking about large, traditional media sources like our nation’s major newspapers and not blogs or “fringe” sources. And we wanted to test the belief that the publication of “fake news” was done for a reason, specifically for a political reason. Our belief was that other survey research on this topic may have been worded too ambiguously. Although it is common for presidents to have contentious, even combative relationships with the press, President Trump’s treatment of the press and his criticisms of the press go far beyond the types of complaints leveled by his predecessors. Our analysis reveals that President Trump’s “war on the press” appears to be working. 57% of respondents “agree” that our nation’s major news outlets are publishing fake news stories for political purposes. Although confidence in media has been declining for the past few decades, this represents a sharp decline in trust, which carries serious implications.

Our media question was structured to also measure the strength at which voters agreed or disagreed with the statement. This graph reflects the distribution of answers. All told, 28% of voters rated the statement, “Our nation’s major news organizations publish fake news stories for political purposes” with a 10 out of 10 (indicating the strongest possible agreement). Just 11% rated it 0 out of 10 (the strongest possible disagreement), just more than a third of voters at the other end of the spectrum.

The crisis of confidence in the media is largely being driven by Republicans, the voting group specifically targeted by President Trump’s anti-press messaging. An astounding 86% of Republicans fall into the “agree” category, with 45% of Republicans rating the statement with a 10 out of 10. However, pure Independents also seem to be affected by Trump’s “fake news” attacks. 63% of pure Independents believe major American news organizations publish fake news stories for political purposes. And although nearly 70% of Democrats disagree with the statement, even 31% of Democrats believe the nation’s major news organizations publish fake news stories for political purposes.

President Trump’s inflammatory rhetoric about the media is not his only rhetoric that presents a sharp departure from the behavior of his predecessors. Trump’s criticisms of Democrats, the FBI, the Justice Department, and others are unprecedented and are amplified both in terms of scope and legitimacy by his role as president of the United States. Our analysis finds this is having a major impact on partisan animosity, particularly among Republican voters. Pew Research Center’s Political Polarization in the American Public report found that in 2014, 27% of Democrats and 36% of Republicans believed their political counterparts presented a threat to the well-being of the country. Just 5 years later, in our survey, these numbers have doubled for both parties, reaching catastrophic levels among Republican voters. Heading into the 2020 election, 7 out of 10 Republican voters believe that the views of Democrats threaten the well-being of the country, up 35 points since Pew’s study. Data presented here include independent "leaners."

31% of Republicans chose the response category 10, indicating the strongest level of agreement with the statement that the views of Democrats threaten the well-being of the country. Also notable is how Republicans score on the “no threat” end of the scale. Just 2% of Republicans chose 0, to indicate that they strongly disagreed that the views of Democrats threatened the well-being of the country; cumulatively, just 29% of Republicans score a 5 or below. Another interesting finding is how self-critical Democrats are. Just 29% score themselves as 0. Independents are more generous, with 32% scoring Democrats with a 0.

34% of Republicans chose 0 to indicate the strongest disagreement with the statement that “views of Republicans threaten the well-being of the country.” Few strongly agree with the statement. Just 2% choose 10. There are significantly fewer Democrats who strongly agree that Republicans’ views are a threat (21%) than there are Republicans who strongly agree that Democrats’ views are a threat (31%). Again, the modal category for Independents is 0 but while low, 8% of Independents chose 10 for Republicans compared to 4% who chose 10 for Democrats.

The good news is that a plurality of voters are willing to accept shared blame for America’s political dysfunction on behalf of their own political party, and combined, a clear majority of voters are at least willing to entertain that their own party bears some responsibility. In fact, 70% of voters fall between a 5 and 10 on the scale. Considering a choice of 0-3 to indicate a belief that one’s own party is “blameless” for our political dysfunction, the partisan breakdown of the voters who hold their party “blameless” is 18% Republican, 12% Democrat, and 21% pure Independent.

Negative voting refers to the concept that in the polarized era, voters will ignore policy, ideology, or even personal shortcomings in their own party’s nominee in order to keep the seat in question out of the hands of the opposition party. For example, in the 2017 U.S. Senate election in Alabama, Republican nominee Roy Moore received 90% of the votes of Republican voters despite credible allegations that he had molested young women. We wanted to get a sense of how self-aware voters are about negative voting, with the expectation that many voters would most likely not think of their vote decisions in this way (despite what recent electoral data might tell us).

Finally, something Republicans and Democrats agree on! Both Republican and Democratic voters agree it’s important to keep the other party’s candidate from winning. 55% of Republicans and 53% of Democrats fall into the “yes” category, compared to 45% of Republicans and 47% of Democrats in the ‘no” category. These differences are statistically indistinguishable. The voting behavior of pure Independents especially stands out here. 89% of pure Independents fall into the “no” category. This suggests that for this distinct category of voters, election campaigns may remain a contest of principles and ideas. For everyone else, elections are as much about keeping the other side from winning power as they are about anything else.

A major tenet of democratic theory is that elected officials are responsive to the views of constituents. That is supposed to come from the bottom-up; the public pressures politicians to be responsive to their policy preferences in order to earn their votes. Polarization has changed this equation, making voter responses to policy a referendum on the person or party proposing the policy, not on the merits of the policy itself. Polarization is causing voters to respond not to policies being proposed, but instead to the politicians proposing the policies. This graph reflects the results of an experimental survey used to measure the power of partisan cues on public opinion preferences among partisans. The idea for the project comes from watching the 2016 Republican National Convention, in which Ivanka Trump proposed paid family leave during her convention speech, a proposal long opposed by the Republican Party. Rather than drawing boos from the deeply ideologically conservative crowd, Ms. Trump’s proposal was met with enthusiastic applause. We wanted to see if we could move partisans completely away from party positions on issues simply by ascribing the policy proposal to their own party. We wanted to use a blunt measurement. Rather than testing changes in the degree of support for a policy, we wanted to see if we could actually flip support.

The policy proposal of a CO2 emissions trading system (“Cap and Trade”) was proposed by Democrats in Congress after they took control of both chambers in the 2006 midterms. Republican backlash to the proposal partially led to the emergence of climate denialism and later to the Tea Party. When told that a CO2 emissions trading system is proposed by “Some People,” which serves as the baseline or the control category, 50% of Republican respondents support the policy. However, when the proposal is ascribed to Democrats in Congress, support among Republicans declines by 9 points. When Republican voters are told that Republicans in Congress propose a CO2 emissions trading system, support jumps by 16 points to 66%. And when told President Trump proposes the policy, support among Republican voters skyrockets to 88% -- a 38-point increase from the baseline.

Although less responsive to partisan cues than their Republican counterparts, Democrats also shift their support when partisan cues change. All told, we tested 7 policies for evidence that partisan cues significantly shift preferences on policies among partisans. The topics included in our test are carbon dioxide emissions trading, the elimination of the alternative minimum tax, paid family leave, equal pay for equal work legislation, education vouchers, a vague statement about support of expansionary fiscal policy, and one non-partisan issue, making Grandparents Day a federal holiday.

Support for a CO2 emissions trading system among Democrats increases dramatically when Democrats are told that their Democratic Party allies in Congress propose the policy, increasing 20 points from the baseline percent of 69% to 89%.

Education vouchers have long been a popular reform among Republicans but when Republican voters are told that Republicans in Congress propose them, support increases by 8 points over the baseline level of support. When Republicans are told President Trump proposes them, support increases by 18 points over the baseline. On every issue tested, Republicans respond more favorably to the President Trump prompt than to the Republicans in Congress prompt, suggesting that Republicans trust their party’s president more than their party’s congressional leaders.

For Democrats, vouchers have always been a lightning rod issue. The teachers’ unions oppose them, and Democratic politicians are often opposed to them, arguing that they weaken public schools. Despite this, 61% of Democrats in the control group support vouchers, and when Democrats in Congress are said to support them, support increases by 8 points to 69%. But Democrats increase their support almost as much when the voucher proposal is ascribed to Republicans in Congress (a 7 point increase, to 67%). Attribution to President Trump causes a 9 point drop in support among Democrats, moving to 52%.

One might think that “equal pay for equal work” is recognizable Democratic Party rhetoric. But it resonates well with Republicans. Regardless of the prompt -- Republicans, Democrats, or Some People -- on average 80% of Republicans support this policy. However, when the policy proposal is ascribed to President Trump, support among Republicans is nearly universal, coming in at 98%.

Democrats support an “equal pay for equal work” policy, and they’ll take it from anyone, even from President Trump. There is a slight decline in support for the policy when it is offered by Republicans in Congress (4 points lower than the baseline category), but the difference is not statistically significant.

The Republican Party’s history with federal family leave policy has been rocky. The Family and Medical Leave Act of 1993 passed during the Clinton Administration was modest, securing 12 weeks of unpaid leave and making it illegal for employers to fire women for having a baby. Most Republicans voted against the bill. The baseline category "some people" draws 47% from Republicans in the survey. There is a 2-point drop among Republicans when the policy is ascribed to Democrats, which is statistically insignificant. However, attributing the proposal to Republicans in Congress or to President Trump results in about a 14-point increase in support from Republican voters, moving to about 61% from the baseline of 47%.

Among Democrats, the baseline of support, "Some People," draws robust support at 68%. Support only declines modestly when the policy is offered to Democrats by Trump or Republicans in Congress, revealing no negative effect on opinion. However, there is a strong, positive effect. When told Democrats in Congress propose the plan support among Democrats skyrockets to 86%.

Public opinion data consistently shows that Americans are conservative in principle and more liberal in terms of policy specifics. This question is meant to touch on a broader, more philosophical policy. As we suspected, the less-concrete question produced strong effects. The baseline, Some People, shows support from 60% of Republicans. Support drops by 18 points among Republicans when the proposal comes from Democrats in Congress. It increases by 14 points to 74% when ascribed to Republicans in Congress and by 22 points to 82% when it comes from President Trump. When the same idea comes from Democrats in Congress versus President Trump, the spread is more than 40 points.

Democrats also show a strong response to partisan cues on the philosophical question about fiscal policy. The baseline support from Democratic voters is 56% and increases by 9 points to 65% when they are told that Democrats in Congress propose the policy. Support among Democrats declines considerably when the proposal is attributed to Republicans in Congress (47%) and to President Trump (41%), declines of 9 and 15 points, respectively.

Voters are more likely to rely on cues, such as partisan cues, when they are faced with a complex policy choice. Eliminating the alternative minimum tax is an obscure income tax proposal long championed by Republicans, which most Americans are not likely to know much about, even Republican voters. Despite the use of the phrase “tax reform,” Republican voters were deeply skeptical when told eliminating the alternative minimum tax was proposed by Democrats in Congress. Support declines by 32 points among Republicans when they are told that Democrats propose the policy.

Democratic voters appear generally wary of the proposal to repeal the alternative minimum tax, whether the proposal is from their own party, from Republicans or from “Some People.” But even lukewarm support for the proposal turns cold when it comes from President Trump. The baseline for Democratic voters is 33%, and their support stays relatively flat when the proposal is ascribed to either Democrats or Republicans in Congress. However, support collapses to just 13% when Democratic voters are told it’s President Trump’s proposal.

No matter how obscure, eliminating the alternative minimum tax is still a political issue, which carries with it embedded political baggage and partisan associations. In order to determine whether partisan cues impact policy preferences even for non-political issues, we polled respondents about something we hoped would be truly non-partisan: making Grandparents Day a federal holiday. We find that no issue, no matter how far removed from the political arena, is safe from the power of partisan cues. First, Republican voters are generally unreceptive to making Grandparents Day a federal holiday; just 36% support this proposal at the baseline. Ascribing the proposal to Democrats makes little difference. However, when Republican voters are told Republicans in Congress wish to make it a holiday, support increases to 47%, an 11-point jump. And when President Trump proposes it? Republican support soars, increasing to 59% (+23 points).

Overall, Democratic voters are more receptive to making Grandparents Day a federal holiday than their Republican counterparts. As were Republicans, they are much more supportive when the policy is proposed by their own party. Support among Democrats increases to 61% from the baseline of 49%, a 12-point increase. Support from the baseline is largely unchanged when the policy is proposed by Republicans in Congress or by President Trump.

Methodology: Wason Center 2020 Likely Voter Survey

The results of this poll are based on 1,001 interviews of likely voters randomly selected from the national voter file via a registered sampling vendor, including 331 on landline and 670 on cell phone, conducted February 3-17, 2019. Percentages may not equal 100 due to rounding. The margin of error for the whole survey is +/- 3.2 at the 95% level of confidence. This means that if 50% of respondents indicate a topline view on an issue, we can be 95% confident that the population’s view on that issue is somewhere between 46.8% and 53.2%. The margin of error is higher for subgroups. All error margins have been adjusted to account for the survey’s design effect, which is 1.1 in this survey. The design effect is a factor representing the survey’s deviation from a simple random sample, and takes into account decreases in precision due to sample design and weighting procedures. Subsamples have a higher margin of error. In addition to sampling error, the other potential sources of error include non-response, question wording, and interviewer error. The response rate (AAPOR RRI Standard Definition) for the survey was 17%. Five callbacks were employed in the fielding process. Live calling was conducted by trained interviewers at the Wason Center for Public Policy Survey Research Lab at Christopher Newport University. The data reported here are weighted using an iterative weighting process on age, race, sex, education, region, and population density to reflect as closely as possible the population of registered voters in the United States as anticipated for the 2020 election.

Methodology: Wason Center Political Attitudes Experimental Survey

The results of this poll are based on 1,280 interviews of Americans over the age of 18 randomly selected via random digit dialing (RDD) from a sample file purchased from a registered sampling vendor. 896 of the respondents were contacted on a phone number from a cell phone list and 384 on a landline. The interviews were conducted in March of 2017. Percentages may not equal 100 due to rounding. The margin of error for the whole survey is +/- 3 at the 95% level of confidence. This means that if 50% of respondents indicate a topline view on an issue, we can be 95% confident that the population’s view on that issue is somewhere between 47% and 53%. The margin of error is higher for subgroups. All error margins have been adjusted to account for the survey’s design effect, which is 1 in this survey. The design effect is a factor representing the survey’s deviation from a simple random sample, and takes into account decreases in precision due to sample design and weighting procedures. Subsamples have a higher margin of error. In addition to sampling error, the other potential sources of error include non-response, question wording, and interviewer error. The response rate (AAPOR RRI Standard Definition) for the survey was 19%. Five callbacks were employed in the fielding process. Live calling was conducted by trained interviewers at the Wason Center for Public Policy Survey Research Lab at Christopher Newport University. The data reported here are weighted using an iterative weighting process on age, race, sex, region, and population density to reflect as closely as possible the population of adults in the United States.