While the economy remains the dominant issue in this fall’s midterm elections, the issue of abortion has increased markedly in importance among Democrats following the Supreme Court’s decision ending the federal guarantee of a right to legal abortion in the United States.

A majority of registered voters (56%) say the issue of abortion will be very important in their midterm vote, up from 43% in March. Virtually all of the increase has come among Democrats: 71% of Democratic and Democratic-leaning registered voters rate abortion as very important; fewer than half (46%) said this in March. By contrast, views among Republicans and GOP leaners have shown almost no change since then (41% now, 40% then).

The two parties are essentially tied on midterm voting intentions: 44% say that if the election were held today, they would vote for the Democratic candidate in their district or lean toward the Democrat, while 42% would vote for the Republican or lean Republican. One-in-ten registered voters say they are not sure, while 4% favor a candidate other than a Republican or a Democrat.

A larger share of Republican than Democratic voters say they have given “a lot” of thought to the upcoming elections. However, Democrats are now almost as likely as Republicans to say it “really matters” which party gains control of Congress in this fall’s midterms, which marks a change since March, when a significantly smaller share of Democrats than Republicans said this.

The new national survey by Pew Research Center was conducted among 7,647 adults, including 5,681 registered voters, from Aug. 1-14. It was largely completed before the FBI’s search of Donald Trump’s residence at Mar-a-Lago, Florida, as part of an investigation into whether Trump took classified records from the White House, and the enactment of a sweeping Democratic-backed bill aimed at addressing climate change, health care costs, corporate taxes and other issues.

Voters’ views about the importance of several issues – not only abortion – have changed since the spring. Compared with March, larger shares say gun policy and violent crime are very important in their voting decisions. As with abortion, these increases have come largely among Democrats. Over this period, there have been declines in the shares of voters in both parties who rate foreign policy, energy policy and the coronavirus outbreak as major issues.

Republicans continue to view the economy as by far the top issue in the upcoming elections. Nine-in-ten Republican voters view the economy as very important, roughly 20 percentage points higher than any other issue.

Among Democrats, 77% view health care as a very important voting issue, while about two-thirds or more say the same about abortion and gun policy (71% each), Supreme Court appointments (69%), the economy (67%) and climate change (66%).

Four years after a midterm election in which there was the highest voter turnout in decades, 68% of registered voters say it really matters which party wins control of Congress this fall; that is identical to the share of voters who said this in August 2018.

Republican and Democratic voters are now about equally likely to say partisan control of Congress really matters (72% of Republican voters vs. 69% of Democrats). The share of Democrats saying the outcome really matters has increased 9 percentage points since March (from 60% to 69%), while Republicans’ views have shown little change (70% in March).

Still, a larger share of Republicans (41%) than Democrats (34%) say they have given “a lot” of thought to the upcoming midterms.

The survey finds that among all adults, Joe Biden’s job approval rating remains in negative territory: 37% approve of his job performance as president, while 60% disapprove. That is unchanged since early July (37% approve) and comparable to Donald Trump’s job rating at about this point in the 2018 congressional campaign (40%).

More voters continue to view their midterm votes as an expression of opposition to Biden than support for him. However, the share of voters who say Biden is not much of a factor in their vote has increased since March among members of both parties.

Currently, about half of registered voters (49%) say Biden is not much of a factor in their midterm vote, while 31% think of their vote as a vote against Biden and 19% see it as a vote for him. The share saying Biden is not much of a factor in their voting decision has increased 11 percentage points since March.

Today, six-in-ten Democrats say Biden is not much of a factor in their midterm vote, up from 47% in March. Fewer Democrats view their vote as an expression of support for Biden than did so five months ago (36% today, 46% then).

Republicans also view Biden as less of factor in their congressional vote. Currently, 62% of Republican voters see their vote as being against Biden, while 37% say the president is not much of a factor in their voting decision. In March, 71% of GOP voters said their vote was against Biden, compared with 26% who said he was not much of a factor.

Other important findings from the survey

Republicans remain skeptical midterm elections will be conducted fairly. Majorities of registered voters say they are confident that this fall’s elections will be conducted fairly and accurately (65% are very or somewhat confident) and that all citizens who want to vote will be able to (75%). These views are little changed since March. Republicans are far more likely than Democrats to have little or no confidence the elections will be conducted fairly (55% vs. 17%). And about a third of Democrats (34%) are not confident that all citizens will be able to vote, compared with just 15% of Republicans.

Views of Biden’s personal traits have turned more negative since early last year. As Biden’s job approval rating has declined since the early months of his presidency, so too have the public’s evaluations of his personal traits. Currently, a 54% majority of adults say Biden stands up for his beliefs. That is his highest rating among six traits included in the survey; in March 2021, 66% said he stood up for his beliefs. Biden gets his lowest ratings for being inspiring; 31% say this describes Biden.

Most Republicans continue to want Trump to remain a major figure. A majority of Republicans and Republican-leaning independents (63%) say they want Trump to continue to as a major national political figure for many years to come. Among those who express this view, more want Trump to run for president in 2024 (39% of all Republicans favor this) than to support another candidate who shares his view (23%). These views have changed modestly since last September, when 67% of Republicans favored Trump continuing as a major figure and 44% wanted him to run again for president.

With less than three months until the midterm elections, registered voters are about equally likely to back Democratic and Republican candidates for the U.S. House. Among all registered voters, 44% say they would vote for or lean toward the Democratic candidate in their House district, and 42% would vote for or lean toward the Republican candidate. One-in-ten registered voters say they are not sure, while 4% favor a candidate other than a Republican or a Democrat.

Demographic differences in voters’ preference are largely consistent with recent presidential and midterm elections.

Among registered voters, just under half of women (47%) say they would vote for the Democratic candidate if the elections were held today, while 38% would vote Republican. Men are more likely to favor Republican candidates, by a narrow margin: 47% of men would vote for a Republican candidate while 41% would choose a Democrat.

About half of White voters (51%) say they would vote for the Republican candidate in their district, while 38% would vote for the Democratic candidate. By contrast, seven-in-ten Black voters would vote Democratic; just 6% of Black voters favor a Republican candidate. As was the case in March, Asian American voters favor Democratic candidates by roughly two-to-one (57% vs. 26%), while Hispanic voters favor Democrats by 25 percentage points.

Younger voters also tend to favor Democratic candidates for the House over Republican candidates. Roughly half of voters ages 18 to 29 (54%) would support a Democratic candidate if the election were held today, while 22% would support a Republican. Nearly a quarter of voters under age 30 (24%) say they would vote for a candidate from another party or are not sure of their voting preference.

Among voters ages 30 to 49, 44% prefer Democratic candidates while 40% favor Republicans. Voters ages 50 and older tilt modestly toward the GOP: 47% say they would vote for or lean toward a Republican candidate, while 42% would vote for or lean toward a Democrat.

As in recent elections, voters with postgraduate degrees continue to be one of the demographic groups most supportive of Democrats. A majority (57%) say they would vote for a Democratic candidate for the House, while 39% favor a Republican. Voters with bachelor’s degrees are more divided: 47% favor Democrats, 42% Republicans. Voters without bachelor’s degrees are about as likely to prefer a Republican (43%) as they are a Democrat (40%).

Congressional vote preferences and views of Biden

Voters who disapprove of Joe Biden’s performance in office but not very strongly are the most divided when it comes to the upcoming congressional elections – though they are more likely to favor Democratic candidates than Republican candidates.

Roughly four-in-ten in this group (43%) say they would vote for or lean toward a Democratic House candidate in their district if the election were held today, while 29% would vote for or lean toward a Republican; another 28% are not sure or favor another candidate.

Those who disapprove of how Biden is handling is presidency – but not so strongly – make up 17% of registered voters. This group is a mix of Democrats and Republicans and includes substantial shares who identify as independents but lean toward one of the two major parties: 31% are Democrats, 27% are Democratic leaners, 22% are Republicans and 17% are GOP leaners.

Voters who strongly disapprove of Biden (43% of voters) and those who approve of his job performance (37% of voters) are more one-sided in their midterm vote preferences. Among strong disapprovers of Biden – the vast majority of whom are Republicans or Republican leaners – 82% favor the GOP candidate in their district.

Large majorities of voters who approve of the way Biden is handling his job as president – 93% of whom are Democrats or Democratic leaners – say they would vote for or lean toward a Democratic candidate for U.S. House if the midterm elections were held today. About nine-in-ten voters who strongly approve of Biden’s job performance (93%) would vote for a Democrat, as would a similar share of those who approve of his performance but not so strongly (86%).

As in 2018, most voters say it ‘really matters’ which party controls Congress

The 2018 midterm elections featured the highest turnout rate among eligible voters in a U.S. midterm election in 100 years. Currently, registered voters are assigning a similar level of importance to this year’s elections as they did at the same point in the 2018 calendar.

Roughly two-thirds of voters (68%) say that it really matters which party wins control of Congress, identical to the share who said this in August 2018.

Today, about seven-in-ten Republicans (72%) and Democrats (69%) say that it really matters which party wins control of Congress. By comparison, at the same point in 2018, Democrats were slightly more likely than Republicans to say this (72% of Democrats vs. 66% of Republicans in August 2018).

Democrats are now 9 points more likely to say it really matters which party wins control of Congress than they were in March of this year. Among Republicans, the share who say this is essentially unchanged since March.

Age, racial and ethnic differences in election engagement

As in past elections, there are demographic differences in the extent to which voters see partisan control of Congress as really mattering, and in the extent to which they have given the upcoming elections a lot of thought.

Older voters are much more likely than younger voters to say they have given a lot of thought to the elections at this stage of the campaign. Half of voters ages 65 and older say this, compared with 41% of voters 50 to 64, 25% of those 30 to 49, and just 20% of voters under 30.

Voters ages 65 and older are also much more likely than those under 30 to say it really matters which party wins control of Congress in November (82% vs. 50%). Six-in-ten voters ages 30 to 49 and 71% of voters 50 to 64 say this.

White voters (72%) are more likely than Hispanic (59%), Black (58%) or Asian American (56%) voters to say partisan control of Congress really matters. And four-in-ten White voters say they have given a lot of thought to the election at this stage in the campaign, compared with 30% of Hispanic, 27% of Black and 17% of Asian voters.

Men (71%) are somewhat more likely than women (66%) to say that it really matters which party wins control of Congress. Men are also more likely than women to say they’ve given a lot of thought to the upcoming elections (40% vs. 33%).

Educational differences are similar to those seen in other recent elections, with the most highly educated voters more likely to say partisan control of Congress really matters; 73% of college graduates and 66% of voters without degrees say this. However, voters with college degrees are no more likely than those without to say they have so far given a lot of thought to the midterms.

Voter engagement and midterm preferences

While preferences for this year’s House elections are closely divided between Republican and Democratic candidates among all registered voters, the GOP holds an edge among those who say they have given a lot of thought to this year’s congressional elections at this stage in the contest: 52% say they would vote for or lean toward a Republican candidate if the election were held today, while 44% would vote for or lean toward a Democratic candidate.

Those who say it “really matters” which party wins control of Congress are evenly divided: 48% favor Democratic candidates while 48% favor Republicans.

Few voters who have given a lot of thought to the elections at this stage or who say it really matters which party wins control of Congress – just 4% and 5%, respectively – say they are not sure who they will vote for or that they plan to vote for a candidate other than a Republican or a Democrat.

Voters’ feelings about possible election outcomes

Similar shares of voters say they would feel positive emotions if Democrats were to maintain control of the U.S. House of Representatives in the midterm elections as if Republicans were to take control.

Overall, the share of voters who say they would react positively to Democrats maintaining control of the U.S. House of Representatives is similar to the share who say the same about Republicans gaining House control in the midterms.

Nearly half of registered voters (48%) say they would feel either excited (10%) or relieved (38%) if Democrats maintained control, while 46% would feel positively if Republicans regained the House (13% excited, 33% relieved).

Republicans are 9 percentage points more likely than Democrats to say they would feel excited if their party held control of the House in the next Congress (26% vs. 17%). Meanwhile, Democrats are 6 points more likely to say they would feel angry if the other party controlled the House (32% vs. 26%).

Both Republican and Democratic voters are somewhat more likely to say they would be angry if the other party controlled the House following the elections than they were prior to the 2018 midterms.

Midterm voting issues: More voters cite the economy as ‘very important’ than any other issue

With a little less than three months until the midterm elections, the economy leads voters’ list of issues in the midterm elections: About three-quarters of registered voters (77%) say the economy is very important to their vote in the 2022 congressional election, making it the top issue out of the 15 asked about in the survey.

About six-in-ten voters say that gun policy (62%), violent crime (60%), health care (60%), voting policies (59%) and education (58%) are very important to their vote.

In the wake of the landmark Supreme Court decision on abortion earlier this year, Supreme Court appointments (58%) and abortion (56%) also rank similarly high on voters’ issue lists.

Energy policy (53%) and immigration (48%) are named by about half of voters as very important issues to their vote. Smaller shares cite foreign policy (45%), the size and scope of the federal government (42%), climate change (40%) and issues around race and ethnicity (35%) as very important to their midterm vote. Roughly a quarter of voters (28%) now say the coronavirus outbreak will be very important to their vote, the lowest of 15 issues asked about in the survey.

Since March, Democrats have become much more likely to say that several issues, including abortion and gun policy, are very important to their vote. By comparison, no issue has substantially grown in importance among Republicans.

On the issue of abortion, 46% of Democratic registered voters said in a March survey that abortion was a very important issue to their vote, months before the Supreme Court’s decision that the U.S. Constitution does not guarantee a right to an abortion. Today, 71% of Democrats say this, making it one of the most important issues for Democrats. Among Republicans, there has been little change on this issue since March.

A roughly similar pattern is seen on gun policy. In March, 54% of Democratic voters said it was a very important issue to their vote, but 71% say this today. Republicans are about as likely to view gun policy as very important today as they were five months ago (55% vs. 52%).

Democrats are also more likely to view Supreme Court appointments as a very important issue to their vote. Today, nearly seven-in-ten Democratic voters (69%) say appointments to the Supreme Court are very important, compared with 59% who said this in March. The share of Republicans who cite court appointments as very important to their vote has slipped 7 points since March, from 55% to 48%.

Among voters, both Republicans and Democrats have become less likely to view both foreign policy and the coronavirus outbreak as very important issues to their vote. In March, two-thirds of Republicans said foreign policy was a very important issue; today 54% say this. Among Democrats, about half said in March that foreign policy was very important to their vote, compared with 37% today.

On the coronavirus outbreak, just 13% of Republicans say it is a very important issue to their vote in the upcoming congressional elections, down 6 points from March. Democratic voters have also seen a similar decline in issue importance for coronavirus (46% in March, 41% now).

Little change in favorable views of Congress

Americans’ views of Congress have changed little since January. Today, roughly three-in-ten adults say they have a favorable view of Congress, up slightly from the 28% who said this in January, but down from April 2021 when more than a third (36%) said the same thing.

Democrats are about twice as likely as Republicans to say they have a favorable view of Congress (40% vs. 21%). Democrats are only slightly more favorable today than they were in January (36%) but 10 points lower than in April 2021, when half viewed Congress favorably. Views of Congress among Republicans have changed little since 2021.

About four-in-ten Americans (37%) now approve of Joe Biden’s job performance as president, while 60% disapprove. Biden’s approval rating is relatively unchanged from earlier in the summer and remains far lower than it was in the early months of his presidency.

Demographic patterns in Biden’s approval rating also are little different than in recent months, with White adults, those without bachelor’s degrees and younger adults more likely than others to disapprove of Biden.

Majorities across all age groups disapprove of Biden, though his rating is somewhat more positive among adults 65 and older (42%) than among those in younger age groups. At the same time, older adults are more likely than younger adults to express strongly negative views of Biden: While 63% of adults under 30 disapprove of Biden’s performance, just 32% say they strongly disapprove. Among those 65 and older, a smaller overall share disapproves (55%), but 46% strongly disapprove.

Biden’s personal traits viewed less positively than last fall

Public perceptions of Biden’s personal traits have turned more negative over the course of his presidency, largely paralleling the decline in his approval ratings, with some exceptions.

While a majority of Americans (54%) say “stands up for what he believes in” describes Biden very (20%) or fairly (34%) well, public assessments of whether Biden is honest and cares about the needs of ordinary people are more divided. About half (48%) say “honest” describes Biden at least fairly well, while a similar share (50%) say it does not. Similarly, about half (48%) characterize him as caring about the needs of ordinary people, while roughly as many (51%) say this is not a good descriptor.

The public’s evaluation is more negative on other measures: About two-thirds say “inspiring” (68%) does not describe Biden well, while nearly as many (64%) say the same for “mentally sharp.” And although 44% say “a good role model” describes Biden at least fairly well, a larger share (55%) say this does not describe the president well.

While Biden’s ratings on each of these traits have declined substantially over the course of his presidency, the sharpest decline is seen in views of Biden’s mental sharpness – and this shift is particularly pronounced among Democrats.

Overall, 35% now say “mentally sharp” describes Biden at least fairly well, down from 54% in March 2021 and 43% last September. The share of Democrats saying this describes Biden very or fairly well has dropped to 61% today from 73% last fall and 86% in March 2021. The share of Republicans saying Biden fits this description at least fairly well has also declined over this period, although to a lesser extent, in large part because just 16% said this in March 2021 (today 6% do).

The trajectory of opinion follows a similar, if less stark, pattern for the other traits asked about in the survey.

A majority of Republicans want Trump to stay in politics; about four-in-ten say he should run for president in 2024

Overall, roughly three-in-ten Americans (31%) say Donald Trump should remain a major national political figure, while two-thirds (66%) say he should not. However, there are wide partisan divides on this question. About six-in-ten Republicans and Republican leaners (63%) say the former president should continue to play a major role in national politics, while nearly all Democrats and Democratic leaners (94%) say he should not.

The share of Republicans saying Trump should continue to be a major national political figure has declined slightly since September 2021 (67% then, 63% now).

The 63% of Republicans who would like to see Trump remain a major figure include 39% who would like to see Trump run for president himself in 2024. The remainder (23%) say while they would like Trump to remain a national political figure, they would prefer he use his position to support another presidential candidate that shares his views.

Although Republicans and Republican leaners generally say Trump should remain a major national political figure, there are demographic differences within the GOP on this question.

While half of Republicans with a bachelor’s degree or more education say Trump should remain a major political figure, that rises to 69% among Republicans with less formal education. Republicans with some college education or less are far more likely than those who are college graduates to say Trump should run for president again: 45% say Trump should run for president himself in 2024, while only about a quarter (26%) of Republican college graduates stay this.

Moderate and liberal Republicans are less likely than conservative Republicans to say Trump should remain on the national political stage (52% vs. 69%, respectively).

This report is a collaborative effort based on the input and analysis of the following individuals:

Research team

Carroll Doherty, Director, Political Research
Jocelyn Kiley, Associate Director, Political Research
Baxter Oliphant, Senior Researcher
Andrew Daniller, Research Associate
Hannah Hartig, Research Associate
Gabe Borelli, Research Associate
Amina Dunn, Research Analyst
Ted Van Green, Research Analyst
Andy Cerda, Research Assistant
Rebecca Salzer, Research Intern

Communications and editorial

Nida Asheer, Communications Manager
Talia Price, Communications Associate
David Kent, Senior Copy Editor

Graphic design and web publishing

Alissa Scheller, Information Graphics Designer
Reem Nadeem, Associate Digital Producer

Methodology

Andrew Mercer, Senior Research Methodologist
Dorene Asare-Marfo, Panel Manager
Dana Popky, Associate Panel Manager
Arnold Lau, Research Methodologist

The American Trends Panel survey methodology

Overview

The American Trends Panel (ATP), created by Pew Research Center, is a nationally representative panel of randomly selected U.S. adults. Panelists participate via self-administered web surveys. Panelists who do not have internet access at home are provided with a tablet and wireless internet connection. Interviews are conducted in both English and Spanish. The panel is being managed by Ipsos.

Data in this report is drawn from the panel wave conducted from Aug. 1 to Aug. 14 and included oversamples of Hispanic, Asian and Black adults, as well as 18-29 year old Republicans and Republican leaning independents in order to provide more precise estimates of the opinions and experiences of these smaller demographic subgroups. These oversampled groups are weighted back to reflect their correct proportions in the population. A total of 7,647 panelists responded out of 13,221 who were sampled, for a response rate of 65%. This included 6,025 respondents from the ATP and an oversample of 1,622 Hispanic respondents from Ipsos’ KnowledgePanel. The cumulative response rate accounting for nonresponse to the recruitment surveys and attrition is 3%. The break-off rate among panelists who logged on to the survey and completed at least one item is 2%. The margin of sampling error for the full sample of 7,647 respondents is plus or minus 1.7 percentage points.

Panel recruitment

The ATP was created in 2014, with the first cohort of panelists invited to join the panel at the end of a large, national, landline and cellphone random-digit-dial survey that was conducted in both English and Spanish. Two additional recruitments were conducted using the same method in 2015 and 2017, respectively. Across these three surveys, a total of 19,718 adults were invited to join the ATP, of whom 9,942 (50%) agreed to participate.

In August 2018, the ATP switched from telephone to address-based recruitment. Invitations were sent to a stratified, random sample of households selected from the U.S. Postal Service’s Delivery Sequence File. Sampled households receive mailings asking a randomly selected adult to complete a survey online. A question at the end of the survey asks if the respondent is willing to join the ATP. Starting in 2020 another stage was added to the recruitment. Households that do not respond to the online survey are sent a paper version of the questionnaire, $5 and a postage-paid return envelope. A subset of the adults returning the paper version of the survey are invited to join the ATP. This subset of adults receive a follow-up mailing with a $10 pre-incentive and invitation to join the ATP.

Across the four address-based recruitments, a total of 19,822 adults were invited to join the ATP, of whom 17,472 agreed to join the panel and completed an initial profile survey. In each household, the adult with the next birthday was asked to go online to complete a survey, at the end of which they were invited to join the panel. Of the 27,414 individuals who have ever joined the ATP, 11,651 remained active panelists and continued to receive survey invitations at the time this survey was conducted.

The U.S. Postal Service’s Delivery Sequence File has been estimated to cover as much as 98% of the population, although some studies suggest that the coverage could be in the low 90% range.¹ The American Trends Panel never uses breakout routers or chains that direct respondents to additional surveys.

Sample design

The overall target population for this survey was non-institutionalized persons ages 18 and older living in the U.S., including Alaska and Hawaii. It featured a stratified random sample from the ATP in which Hispanic, Asian and Black adults, as well as 18-29 year old Republicans and Republican leaning independents were selected with certainty were selected with certainty. The remaining panelists were sampled at rates designed to ensure that the share of respondents in each stratum is proportional to its share of the U.S. adult population to the greatest extent possible. Respondent weights are adjusted to account for differential probabilities of selection as described in the Weighting section below.

The ATP was supplemented with an oversample of self-identified Hispanic respondents from the KnowledgePanel who were of Mexican origin, Central American ancestry or who had no more than a high-school education.

Questionnaire development and testing

The questionnaire was developed by Pew Research Center in consultation with Ipsos. The web program was rigorously tested on both PC and mobile devices by the Ipsos project management team and Pew Research Center researchers. The Ipsos project management team also populated test data that was analyzed in SPSS to ensure the logic and randomizations were working as intended before launching the survey.

Incentives

All respondents were offered a post-paid incentive for their participation. Respondents could choose to receive the post-paid incentive in the form of a check or a gift code to Amazon.com or could choose to decline the incentive. Incentive amounts ranged from $5 to $20 depending on whether the respondent belongs to a part of the population that is harder or easier to reach. Differential incentive amounts were designed to increase panel survey participation among groups that traditionally have low survey response propensities.

Ipsos operates an ongoing modest incentive program for KnowledgePanel to encourage participation and create member loyalty. The incentive program includes special raffles and sweepstakes with both cash rewards and other prizes to be won. Typically, panel members are assigned no more than one survey per week. On average, panel members complete two to three surveys per month with durations of 10 to 15 minutes per survey. An additional incentive is usually provided for longer surveys. For this survey, during the last few days of data collection, KnowledgePanel members were offered 10,000 points (equivalent to $10) in addition to the regular incentive program in an attempt to boost the number of responses from panel members of Central American ancestry.

Data collection protocol

The data collection field period for this survey was Aug. 1 to Aug. 14, 2022. Postcard notifications were mailed to all ATP panelists with a known residential address on Aug. 1, 2022. 

Invitations were sent out in two separate launches: Soft Launch and Full Launch. Sixty ATP panelists and 909 KP panelists were included in the soft launch, which began with an initial invitation sent on Aug. 1, 2022. The ATP panelists chosen for the initial soft launch were known responders who had completed previous ATP surveys within one day of receiving their invitation. All remaining English- and Spanish-speaking panelists were included in the full launch and were sent an invitation on Aug. 3, 2022.

All panelists with an email address received an email invitation and up to four email reminders if they did not respond to the survey. All ATP panelists that consented to SMS messages received an SMS invitation and up to four SMS reminders.

Data quality checks

To ensure high-quality data, the Center’s researchers performed data quality checks to identify any respondents showing clear patterns of satisficing. This includes checking for very high rates of leaving questions blank, as well as always selecting the first or last answer presented. As a result of this checking, 12 ATP and seven KP respondents were removed from the survey dataset prior to weighting and analysis.

Weighting

The data was weighted in a multistep process that accounts for multiple stages of sampling and nonresponse that occur at different points in the survey process. First, each panelist began with a base weight that reflects their probability of selection for their initial recruitment survey. These weights were then adjusted to account for each panelist’s probability of being sampled to participate in this wave.

Next, respondents were placed into one of five sample groups: 1) Hispanic adults of Mexican origin 2) Hispanic adults of Central American origin 3) Other Hispanic adults with no more than a high school education 4) Other Hispanic adults with more than a high school education 5) Non-Hispanic adults. Separately within each group, the weights for ATP and KnowledgePanel respondents were scaled to be proportional to their effective sample size within that group. The ATP and KnowledgePanel respondents were then recombined and the weights were poststratified so that the weighted proportion of adults in each of the five groups matched its estimated share of of the U.S. adult population.

The weights were then calibrated to align with the population benchmarks identified in the accompanying table. These include a set of weighting parameters for the total U.S. adult population as well as an additional set of parameters specifically for Hispanic adults. Separately for each sample group, the weights were then trimmed at the 1st and 99th percentiles to reduce the loss in precision stemming from variance in the weights. Sampling errors and tests of statistical significance take into account the effect of weighting.

Some of the population benchmarks used for weighting come from surveys conducted prior to the coronavirus outbreak that began in February 2020. However, the weighting variables for most ATP members were measured in 2022. A small number of panelists for which 2022 measures were not available use profile variables measured in 2021. For KnowledgePanel respondents, many of the weighting variables were measured on this wave.

This does not pose a problem for most of the variables used in the weighting, which are quite stable at both the population and individual levels. However, volunteerism may have changed over the intervening period in ways that made these more recent measurements incompatible with the available (pre-pandemic) benchmarks. To address this, volunteerism is weighted to an estimated benchmark that attempts to account for possible changes in behavior.

The weighting parameter is estimated using the volunteerism profile variable that was measured on the full American Trends Panel in 2021 but weighted using the profile variable that was measured in 2020. For all other weighting dimensions, the more recent panelist measurements were used. For American Trends panelists recruited in 2021, the 2020 volunteerism measure was imputed using data from existing panelists with similar characteristics.

This ensures that any patterns of change that were observed in the existing panelists were also reflected in the new recruits when the weighting was performed.

The following table shows the unweighted sample sizes and the error attributable to sampling that would be expected at the 95% level of confidence for different groups in the survey.

Sample sizes and sampling errors for other subgroups are available upon request. In addition to sampling error, one should bear in mind that question wording and practical difficulties in conducting surveys can introduce error or bias into the findings of opinion polls.

Dispositions and response rates

CORRECTION (August 30, 2022): A previous version of this methodology statement incorrectly identified the years in which weighting variables for ATP members were measured. The text has been corrected to indicate that profile variables for these panelists were measured in either 2022 or 2021.  None of the study findings or conclusions are affected.

© Pew Research Center, 2022