The divide between young voters and older voters was as stark this year as it was in 2008. While Obama lost ground among voters younger than 30, he still won this age group by 24 points over Mitt Romney (60% to 36%). He also maintained a slimmer advantage among voters 30 to 44 (52% Obama, 45% Romney), while losing ground among those 45 to 64 and those 65 and older.

Among all voters 30 and older, Obama ran behind Mitt Romney (48% for Obama, 50% for Romney). Four years ago, Obama edged John McCain, 50% to 49%, among all 30+ voters.

In Florida, Ohio, Virginia and Pennsylvania, Obama also failed to win a majority of voters 30 and older. Yet he swept all four battleground states, in part because he won majorities of 60% or more among young voters.

Just as critically, young people made up as large a share of the overall electorate as they did in 2008, according to the national exit poll (19% in 2012, 18% in 2008). As recently as September, young voters were significantly less engaged in the campaign than they had been four years earlier. But their interest and engagement levels increased in the campaign’s final weeks. In the Pew Research Center’s final pre-election survey, as many registered voters under 30 said they were giving a lot of thought to the election as did so in the last weekend of the 2008 race.

Obama’s support among young voters declined among many of the same subgroups in the overall electorate in which he lost ground, particularly whites, men and independents. Obama won a majority of white non-Hispanics under 30 in 2008, but lost this group to Romney this year. In contrast, Obama won young African Americans and Hispanics by margins that were about as large as in 2008.

His losses among young voters since 2008 might have been even greater, but for the fact that the under 30s are by far the most racially and ethnically diverse age group. Just 58% are white non-Hispanic, compared with 76% of voters older than 30. A recent report by Pew Social and Demographic Trends found that minorities are on track to become a majority of the overall population by 2050.

Young voters continue to identify with the Democratic Party at relatively high levels and express more liberal attitudes on a range of issues – from gay marriage to the role of the federal government – than do older voters. In fact, voters under 30 were as likely to identify as Democrats in the 2012 exit poll as they had been in 2008 (44% now, 45% then). And they are the only age group in which a majority said that the government should do more to solve problems.

The last two presidential elections have had the widest gaps in voting between young and old of any election since 1972. This year, 60% of those under 30 backed Obama, compared with just 48% of those 30 and older; in 2008, the gap was 16 points (66% of under 30 supported Obama vs. 50% of those 30 and older).

This year’s 12-point difference between old and young this year was identical to the gap in 1972, when 46% of voters 18-29 supported George McGovern compared with just 34% of those 30 and older.

As Pew Research observed a year ago in The Generation Gap and the 2012 Election there were only modest generational differences in presidential vote preferences between 1976 and 2004.

Changes in the Youth Vote: 2008-2012

While Obama carried the youth vote overall, his support declined from 2008 among all young voters and among key subgroups. In particular, Obama lost ground among young whites, men and independents.

Only 44% of white voters under 30 backed Obama, while 51% voted for Romney. This is a substantial change compared with 2008, when Obama carried the young white vote by 10 points (54% to 44%). Far more young blacks and Hispanics backed Obama than Romney, and there was little fall off in his support among these groups from 2008.

Obama also lost support among young men. Overall, 53% of men under 30 supported Obama, down from 62% in 2008. Fully 66% of young women voted for Obama, similar to the 69% who voted for him in 2008.

However, Obama lost support among both white men and women. Overall, 41% of white men supported Obama while 54% supported Romney. In 2008, Obama won the vote among white men, 52% to 46%. While white women voted for Obama over McCain by a 56% to 42% margin four years ago, they were divided this year (48% voted for Obama, 49% for Romney).

Surprisingly, Obama’s vote also declined among young black men, by 14 points, while holding steady among young black women.

Youth Vote in Battleground States

In four of the key battleground states, Obama may not have won without the youth vote. In Florida, 47% of voters 30 and older backed Obama, while 52% supported Romney. But by winning younger voters by two-to-one (66% to 32%), Obama eked out a narrow one-point victory in the state (50% to 49%).

The pattern was similar in Ohio – 48% of voters 30 and older voted for Obama. When younger voters were included, his share of the vote increased to 50% (vs. 48% for Romney). And in both Pennsylvania and Virginia, the youth vote helped push Obama over 50%.

In 2008, Obama won half or more of the vote among those 30 and older in all four states. Thus, he probably would have been able to carry the vote, even with less support from young voters.

The Demography of Young Voters

The racial and ethnic composition of young voters has shifted dramatically over the last four presidential elections. Just 58% of voters age 18-29 identified as white non-Hispanics, while 18% were Hispanic, 17% were African American and 7% identified as mixed-race or some other race. The share of young voters who are white has declined 16 points since 2000, when 74% of voters under 30 identified as white and 26% identified as nonwhite (including 12% who were African American and 10% Hispanic).

This stands in sharp contrast to older voters. Fully 76% of voters 30 and older were white, down only six points from 2000. Only 24% of voters 30 and older were nonwhite, including 12% who identified as black and 8% as Hispanic.

Younger voters also are less likely to be religious than older voters. Only 36% of young voters attend religious services at least weekly, compared with 51% of older voters.

And when it comes to religious affiliation, 47% of young voters say they are Protestant or Christian, but just 16% are white evangelicals. By contrast, 62% of voters 65 and older are Protestant, including 28% who are white evangelicals.
Young voters are three times as likely as voters 65 and older to say they are not affiliated with a religious tradition; 19% of young voters say they have no religious affiliation, compared with just 6% of voters 65 and older.

Party Identification and Issues

Young voters were not only more likely to back Obama, but they are far more likely than older voters to identify as Democrats than their older counterparts. A plurality of voters younger than 30 (44%) consider themselves Democrats while far fewer identify as Republicans (26%) or independents (30%). Young voters are as Democratic as they were in 2008 when 45% identified as Democrats (26% as Republican and 29% as independent).

By comparison, Democrats had just a three-percentage point advantage in party identification among voters 30 and older. Among those voters, 37% identify as Democrats, 34% as Republicans and 25% as independents.

A third of young voters described their political views as liberal, while 41% said they were moderate and just 26% said they were conservative. Among voters 30 and older, 23% said their political views were liberal, 41% said their views were moderate and 37% described themselves as conservative.

Consistent with their strong vote for Obama and their Democratic Party affiliation, young voters also have more liberal views on a variety of issues. A majority (59%) said that the government should do more to solve problems, while 37% said the government is doing too many things better left to businesses and individuals.

That is a somewhat narrower margin than in 2008 (69% vs. 27%), but they were still far more likely than older voters to say that the government should do more to solve problems.

Young voters also were far more approving of the 2010 health care law: 53% said it should be expanded (34%) or left as it is (19%). That compares with 42% of 30 and older who favored retaining the health care law or expanding it.

On social issues, 64% of voters younger than 30 said abortion should be legal in all or most cases, compared with 58% of voters 30 and older. And they are far more likely than older voters to support allowing gay marriage. Fully 66% of young voters favored their states legally recognizing gay marriage, compared with 45% of voters 30 and older (and just 37% of those 65 and older).

The 2012 presidential campaign was a frustrating experience for many voters, who say the campaign was more negative than usual and had less discussion of issues than in most previous campaigns. Both Obama and Romney get mixed grades for the job they did reaching out to voters, as do campaign consultants, the press and pollsters. On most measures, voters’ views of campaign 2012 fall short of the election four years ago.

Similarly, voters do not have a particularly rosy outlook on national politics going forward. Fully 66% say that relations between Republicans and Democrats will either stay about the same (52%) or get worse (14%) over the next year. And while 56% of voters think Obama will be successful in his coming term, that is down from the 67% who thought his first term would be successful at this point four years ago.

While broad majorities of all voters want Barack Obama (72%) and the Republican leadership (67%) to work with the other side to get things done over the coming year, each party’s political base sends mixed signals. Only about half (46%) of Republicans want GOP leaders to work with Obama to get things done, while about as many (50%) say they should stand up to Obama, even if less gets done. The message to Obama from Democrats is only somewhat more conciliatory: 54% want the president to try to work with Republicans, but 42% do not.

Republicans and Republican leaners remain of the view that the GOP leaders should move in a more conservative direction, not a more moderate one, by a 57% to 35% margin. Democrats and Democratic leaners, meanwhile, continue to support more moderation from their political leaders: Nearly six-in-ten (57%) want Democratic leaders to move in a moderate direction, while 33% want them to move in a more liberal direction.

These are the principal findings of the Pew Research Center’s quadrennial post-election survey, conducted Nov 8-11, 2012 among 1,206 voters who were originally interviewed before the election. The poll finds that, despite expressing strong criticisms of the campaign, most voters say they were satisfied with the choice of presidential candidates and believe they learned enough about them over the course of the campaign to make an informed choice. The presidential debates, in particular, stand out as positive – about two-thirds (66%) say they were helpful in learning about the candidates.

Republican voters are about as likely as Democratic voters to say they learned enough about the candidates to make an informed choice and to have found the debates helpful.

Republicans also are less critical of their candidate, and their party, today than they were after John McCain’s loss in 2008. Fully 75% of Republican voters give Mitt Romney a grade of A or B for the way he conducted his campaign. In the post-election survey four years ago, 63% of Republican voters gave McCain an A or B. In that regard, more GOP voters say they were satisfied with their choice of candidates this year than after the 2008 election (57% now, 38% then).

Republicans also are more positive about the GOP’s performance in the campaign. Six-in-ten GOP voters (60%) give their party grades of A or B for the way it conducted the campaign; just 44% rated the party that positively after the election four years ago.

But Republicans give the voters much lower grades than in 2008 – just 29% give a grade of A or B, down from 47%. In fact, Republicans’ grades for the voters equal the lowest grades given by members of either party dating back to 1988. In 1996, after Bob Dole lost to Bill Clinton, just 30% of Republicans gave the voters positive grades.

The survey finds that internet has again grown as a source of campaign news. Nearly half (47%) of voters say the internet was a main source of campaign news over the course of the election, up from 36% four years ago. The internet now far surpasses newspapers (27%) as a main source of campaign news, though it still trails television (67%).

In this vein, virtually all voters (92%) who followed the returns on election night tracked them on television, and 34% followed the returns on the internet. Slightly more than a quarter of voters (27%) were “dual screeners,” using both television and the internet to get information. Among voters younger than 40, 39% of those who followed returns on election night kept track both by watching TV and following online.

Election night is also a social experience for some voters: 16% of those who followed election returns did so with friends, while 8% used social networking sites like Facebook and Twitter to track the results. Obama supporters were more likely to watch returns with friends, and to use social networks to follow results, than were voters who supported Romney.
Aside from long lines, few voters report having had problems casting their ballots. Some 37% of the voters interviewed say they cast their ballot before Election Day, 19% in person and 17% by mail. Many early voters cite convenience as the main factor they cast their ballots early, but for those who went to early voting sites the lines were often just as long as for those who waited until Election Day.

While most say the voting process in their area was managed well and that they are very confident that their votes were accurately counted, confidence about the vote nationwide is down from 2008. About three-in-ten (31%) voters say they are very confident that the votes across the country were accurately counted this year, down from 43% four years ago. Just 21% of Romney voters say they are very confident in the accuracy of this year’s vote, down from 29% among McCain supporters four years ago. Skepticism is also up among Obama supporters, 42% are very confident that the nation’s votes were accurately counted, down from 56% after the 2008 election.

November 14, 2012

Postelection talk of “lessons learned” is ­often exaggerated and misleading, and so it is in 2012.

A week after President Obama won re-election, two themes are dominant. First, that Mr. Obama kept his job because key elements of his base—notably young people, African-Americans, Latinos and Asian-Americans—turned out for him. Second, that the growing size of these voting blocs represents a decisive challenge for the Republican party.

Both points are true, but most observers are overstating the gravity of the GOP’s problem. In particular, they are paying too little attention to how weak a candidate Mitt Romney was, and how much that hurt Republican prospects.

Here is what the exit poll found. Mr. Romney’s personal image took a hard hit during the primary campaign and remained weak on election day. Just 47% of exit-poll respondents viewed him favorably, compared with 53% for Mr. Obama. Throughout the campaign, Mr. Romney’s favorable ratings were among the lowest recorded for a presidential candidate in the modern era. A persistent problem was doubt about his empathy with the average voter. By 53% to 43%, exit-poll respondents said that Mr. Obama was more in touch than Mr. Romney with people like themselves.

Mr. Romney was never fully embraced by Republicans themselves, which may have inhibited the expected strong Republican turnout. Pew’s election-weekend survey found Mr. Romney with fewer strong supporters (33%) than Mr. Obama (39%). Similarly, a much greater percentage of Obama supporters (80%) than Romney supporters (60%) told Pew that they were voting for their candidate rather than against his opponent.

Continue reading at WSJ.com

Voters had a mixed reaction to the outcome of the 2012 presidential election in the hours immediately following Barack Obama’s victory. Reflecting the narrow vote margin between the candidates, 44% of voters overall said they were happy with the outcome, while 43% said they were unhappy, according to a Pew Research Center report based on survey data collected by Google among a sample of internet users.

When voters were asked for a single word that described their reaction to Obama’s victory, Obama voters said they were “relieved” and “happy.” A substantial number of Obama voters also mentioned words like “great,” “elated,” “yes” and “good.”

Romney voters generally said they were “disappointed” or “sad” about the election outcome. Romney voters also used the words “disgusted,” “sick,” “horrified” and “scared,” to describe their reaction to Obama’s victory.

Overall, 41% of those surveyed – including both voters and nonvoters – said they were happy that Obama was reelected president; 37% were unhappy and 22% neither. A majority of nonvoters (55%) said they were neither happy nor unhappy with the election outcome, while 29% said they were happy and 16% said they were unhappy.

The reaction to the election was divided along partisan lines. Fully 90% of Obama voters were happy that their candidate won while 88% of Romney voters expressed unhappiness.

This report examines Google Consumer Surveys, a new tool developed by Google that interviews a stratified sample of internet users from a diverse group of about 80 publisher sites who allow Google to ask one or two questions of selected visitors as they seek to view content on the site. The sample is stratified on age, gender and location; these demographic characteristics are inferred based on the types of websites the users visit, as recorded in their DoubleClick advertising cookie and their computer’s internet address, and then is weighted by these same characteristics to parameters for all internet users from the Current Population Survey. It is neither an “opt in” survey nor a recruited panel but does not constitute a probability sample of all internet users.

The Pew Research Center remains committed to rigorous, probability-based sampling and to dual frame telephone surveys for measuring public opinion, tracking long-term trends and conducting in-depth analyses of the interrelationship of demographic characteristics and social and political values and attitudes. We continue to evaluate the performance of dual frame telephone surveys, as in our study of the impact of survey nonresponse earlier this year. It showed that “despite declining response rates, telephone surveys that include landlines and cell phones and are weighted to match the demographic composition of the population continue to provide accurate data on most political, social and economic measures.”

It is important to critically evaluate new methodologies, as our traditional methods face growing challenges, especially increasing nonresponse and rising costs. To evaluate the results obtained using Google Consumer Surveys, the Pew Research Center, in consultation with Google, embarked on a series of tests covering a wide range of topics and question types to compare results from Pew Research telephone surveys to those obtained using the Google Consumer Surveys method. This testing is ongoing. This report describes the findings of the evaluation thus far and provides a description of the Google Consumer Surveys methodology. The analysis and conclusions are solely those of the Pew Research Center.

Pew Research and Google Comparisons

From May to October, 2012, the Pew Research Center compared results for more than 40 questions asked in dual frame telephone surveys to those obtained using Google Consumer Surveys. Questions across a variety of subject areas were tested, including: demographic characteristics, technology use, political attitudes and behavior, domestic and foreign policy and civic engagement. Across these various types of questions, the median difference between 43 results obtained from Pew Research surveys and using Google Consumer Surveys was 3 percentage points. The mean difference was 6 points, which was a result of several sizeable differences that ranged from 10-21 points and served to increase the mean difference.

Differences between the Pew Research surveys and Google results occur for a number of reasons. Given that Google Consumer Surveys does not use a true probability sampling method, and its sampling frame is not of the general public, differences in the composition of the sample are potentially of greatest concern. A comparison of several demographic questions asked by Pew Research indicates that the Google Consumer Surveys sample appears to conform closely to the demographic composition of the overall internet population. Communication device ownership and internet use also aligns well for most, though not all, questions. In addition, there is little evidence so far that the Google Consumer Surveys sample is biased toward heavy internet users.

Some of the differences between results obtained from the two methodologies can be attributed to variations in how the questions were structured and administered. During the evaluation period, we typically tried to match the question wording and format. However, some exceptions had to be made since many of the questions were part of longstanding Pew Research trends and had to be modified to fit within the Google Consumer Surveys limits and the different mode of administration (online self-administered vs. interview-administered by telephone).

The context in which questions are asked could also explain some of the differences; questions in Pew Research surveys are asked as part of a larger survey in which earlier questions may influence those asked later in the survey. By contrast, only one or two questions are administered at a time to the same respondents in the Google Consumer Surveys method.

The Google Consumer Surveys method is a work in progress and the Pew Research Center’s evaluation began shortly after its inception and continued for six months. The testing is ongoing, and we will continue to evaluate their methodology.

Methodology of the Google Consumer Surveys

The Google Consumer Survey method samples internet users by selecting visitors to publisher websites that have agreed to allow Google to administer one or two questions to their users. There are currently about 80 sites in their network (and 33 more currently in testing). These include a mix of large and small publishers (such as New York Daily News, Christian Science Monitor, Reader’s Digest, Lima, Ohio News and the Texas Tribune), as well as sites such as YouTube, Pandora and others. Google is attempting to assemble a diverse publisher network covering a range of content (e.g., news, reference, arts and entertainment), size and geography. The results page for each question shows the proportion of respondents from these publisher content groups. Google excludes publishers whose sites include or link to various types of potentially offensive content. (See McDonald et al. for further information about the methodology, as well as a report on Google’s own comparison of results with external benchmarks.)

Google Consumer Surveys selects potential respondents by using inferred characteristics of visitors to the network of publisher sites to attempt to create a sample of internet users that matches national parameters for age, gender and location for the internet using population, based on estimates derived from the Census Bureau’s 2010 Current Population Survey’s Internet Use Supplement. In a stratified-sampling process, the selection of respondents, done in real-time by computer algorithms, attempts to fill each survey with the proper proportion of individuals by age, gender and location (region, state and/or zipcode) needed for all active surveys. For example, if a male in the 18-24 age group living in the Western U.S. visits a publisher in the network and is available to receive a survey, the system will randomly select among the available questions to present to that user. Users are selected by the system and cannot opt in to any survey.

Although respondents cannot volunteer to take part in the study, the resulting sample is a non-probability sample of internet users. It is unknown whether visitors to the network of publisher sites are fully representative of all internet users or what proportion of internet users are covered by the publisher network. All members of the internet using population do not have a known chance of being included in the sample. As a result, no meaningful margin of error can be calculated for projecting the results to the internet population. In addition, the non-probability sampling may result in more variation from sample to sample.

The demographic targeting used in selecting respondents is based on inferred information. Geography is inferred through a respondent’s IP address, while gender and age are inferred based on the types of websites the users visit as recorded in their DoubleClick advertising cookie. The system also deposits a short-term cookie to prevent users from being asked to participate in the same survey more than once. Errors associated with inferred demographic characteristics can influence the sampling and weighting process, even if these inferred demographics are not used in the analysis. For approximately 30-40% of the users, demographic information is not available – either because their cookies are turned off but more often because the algorithm cannot determine a trend from the websites visited as recorded in their DoubleClick advertising cookie that would suggest what gender or age they are. For results reported on the weighted sample, respondents without inferred demographic information on the variables used in weighting are excluded.

Weighting is done with multiple-cell crosstabs, where the sample size permits, that combine age, gender and location (state or region depending on the most specific geography for which a reliable estimate is available). If some variables are not available, the weighting will adjust to use any of the three characteristics that are available.

The point at which users receive the question prompt varies by publisher site. For example, questions may appear after a user attempts to access any content, views a certain number of articles or attempts to access particular types of content (such as a photo gallery). Users may complete the initial question shown to them, request an alternative question, complete some other action (such as logging into an account, signing up to receive emails, or sharing the content on social media), or decide not to view the content on that site.

Only one or two questions can be administered to the same respondent and currently there is no ability to administer questions to the same respondents over time. This may increase response rates by reducing respondent burden, but is also one of the key limitations of the Google Consumer Surveys method. Much of the political and social research conducted using survey data seeks to explore the relationship among attitudes and behaviors; such analyses require multiple questions to be asked of the same respondent. Similarly, the ability to administer only one or two questions to the same respondent means that few measures of demographic characteristics are available for analysis.

It is also difficult to ask complex questions using the Google Consumer Surveys platform. There is a limit of 125 characters on question stems and 44 characters on response options. In addition, a maximum of five response categories can be offered. These limitations mean that longer questions cannot be asked or have to be substantially modified, potentially affecting how people comprehend and answer the question.

The brevity of a Google survey does confer one important advantage, which is that surveys can be fielded very quickly: 1,000 or more responses can be obtained in a matter of a few hours, though most surveys typically run for one or two days. Consequently, Google Consumer Surveys can be used for gathering immediate reactions to events that would be difficult and expensive to measure using telephone surveys and for tracking reactions to measure how they evolve in the short and long term. These include qualitative responses to events, such as verbatim or “one word” reactions.

Demographic Characteristics

Based on tests of several demographic variables, the profile of internet users who respond to Google Consumer Surveys is similar to that of internet users in Pew Research Center surveys. The profile of Google Consumer Surveys respondents shown here is not based on Google’s inferred demographic information, but on demographic questions that were asked of respondents to Google Consumer Surveys.

As discussed in more detail below, there can be substantial errors in how individual people are classified using Google’s inferred demographics (See “Assessing Google’s Inferred Demographics” below.) But in this test, Google Consumer Surveys achieved a representative sample of internet users on gender, age, race/ethnicity, marital status and home ownership when compared with internet users in Pew Research Center surveys.

The gender balance and age profile of internet users in Pew Research surveys and Google Consumer Surveys were fairly similar. In addition, both Google Consumer Surveys and Pew Research reached a similar share of white and non-white internet users.

Each source found that about half of internet users are married while about half are not, and the specific status of the unmarried (widowed, divorced, never married or living with a partner) also were very similar. And in both the Pew Research survey and the Google Consumer Surveys, 63% of internet users said they owned their home.

Weekly church attendance among internet users was comparable in the Pew Research survey and the Google Consumer Surveys. Volunteerism rates were similar in both surveys, although slightly more internet users say they volunteered in the past 12 months in the Pew Research survey than using Google Consumer Surveys (51% vs. 45%).

On two other measures of social and political engagement – talking with neighbors and contacting a public official – there were substantial differences between the results from the Pew Research and Google survey. Nearly six-in-10 (58%) in the Pew Research survey say they talk with their neighbors weekly or more, compared with 43% using Google Consumer Surveys. Nearly twice as many in the Pew Research survey as in the Google surveys said they contacted a public official in the past 12 months (34% vs. 18%). On both of these measures, however, Google results were closer to the estimates from the Current Population Survey’s Civic Engagement Supplement.

Internet and Technology Use

Given the Google surveys’ reliance on internet users visiting particular websites, it is especially important to determine the extent to which internet and technology use among Google’s respondents conforms to the broader population of internet users. Google’s own analysis of visitors to the Google Consumer Surveys publisher network shows that heavier internet users are more likely to appear, but the magnitude of this bias is relatively small. Comparisons of measures of device ownership and internet use in Pew Research surveys and Google Consumer Surveys confirm this.

In general, the percentage who said they owned particular devices and engaged in various online activities were fairly similar in the Pew Research surveys and the Google Consumer Surveys. The percentages of internet users saying they owned a tablet and e-readers were about the same in both the Pew Research survey and Google Consumer Surveys.

In the Pew Research survey, 15% of internet users said they use Twitter, compared with 18% using Google Consumer Surveys. The number saying they donated to charity online also was comparable; 25% in the Pew Research survey and 28% using Google Consumer Surveys. Social networking use was somewhat lower in the Google Consumer Surveys (57%) than in the Pew Research survey (63%), as was getting news online (70% vs. 77%, respectively).

However, there was a difference in smartphone ownership and searching for health information online. Google’s samples reported lower levels of smartphone ownership, when asked in the same way as in the Pew Research survey, and fewer said they searched for health information online.

The Pew Research question on smartphone ownership asks “Do you currently own a smartphone, such as a Blackberry, iPhone, Android or Windows phone?” In response, 55% of internet users in a telephone survey said that they did, compared with 39% in a Google survey. However, in a separate test using different question wording, respondents were asked “What type of mobile phone do you currently own?” and were offered Android, iPhone, Blackberry, Windows phone and “other type of mobile phone” as separate choices. In this version, 53% of Google respondents reported having one of the types of smartphones.

There also was a large difference in the percentage who said they looked for health or medical information online; in a Pew Research survey 71% of internet users said they did this, compared with 52% in a Google survey.

Political Attitudes and Policy Views

Across several political measures, the results from the Pew Research Center and using Google Consumer Surveys were broadly similar, though some larger differences were observed.

On party identification, the Google sample included slightly more Republicans (27% vs. 24%) and more conservatives (40% vs. 36%) than the Pew Research survey’s sample. Similarly, ratings of Obama’s job approval were more negative using Google Consumer Surveys (at the time, 45% vs. 50% approved of Obama job performance). In a September comparison, more voters reached using Google Consumer Surveys supported Obama’s re-election than in the Pew Research survey (57% vs. 51%).

Views about the size and role of government were similar in a Pew Research survey and the Google survey. In both, more respondents said they prefer a smaller government providing fewer surveys than a bigger government providing more services.

Reported frequency of voting also was little different in the Google Consumer Surveys and the Pew Research survey. A majority of respondents to the Pew Research survey (69%) reported voting always or nearly always, compared with 65% in a Google survey.

There were larger differences between the Pew Research results and those obtained using Google Consumer Surveys on several domestic policy issues tested. But taken collectively, the direction of the differences were not consistently in a liberal or a conservative direction. On the issue of same-sex marriage, opinion was more divided in the Pew Research survey than in the Google survey. In the Pew Research survey, 48% favored and 44% opposed allowing gays and lesbians to marry legally. In the Google survey, more favored allowing same-sex marriage, by a 59% to 41% margin.

The Pew Research survey found more support for Obama’s policy to allow illegal immigrants brought to the U.S. as children to remain in the country and apply for work permits (63% approve vs. 33% disapprove) than using Google Consumer Surveys (52% approve, 48% disapprove).

Opinion about the health care legislation passed by Obama and Congress in 2010 was divided in the Pew Research and Google surveys, both before and after the Supreme Court ruling upholding most of the legislation. The results of the two surveys were similar, especially after accounting for possible mode differences.

On the issue of global warming, more in the Pew Research survey said there is solid evidence that the average temperature on earth has been warming over the past few decades (67% vs. 57% using Google Consumer Surveys). But the percentage of people saying that warming is occurring mostly because of human activity was similar in the two surveys.

Across a variety of foreign policy issues, results from the Pew Research surveys and those obtained using the Google Consumer Surveys method were quite comparable. When it comes to economic and trade policy toward China, slightly more respondents in both said that it is more important to get tougher with China than to build a stronger relationship with China,

On the issue of withdrawing U.S. troops from Afghanistan, similar percentages in both said Obama is handling this about right. But more said that Obama was not removing troops quickly enough in the Google survey (36% vs. 28% in the Pew Research survey). A majority of the public approved of the use of drones to target terrorists in other countries in both approaches, but support was somewhat higher using Google Consumer Surveys than in the Pew Research survey (63% vs. 55%).

By about two-to-one, in both surveys, more said that good diplomacy rather than military strength is the best way to ensure peace. This was tested in two versions of a long-term trend question about political values. One version, which the Pew Research Center began tracking in 1987, asks if the respondent agrees or disagrees that “the best way to achieve peace is through military strength.” The other asks respondents to choose between two alternatives: one is the same as the original question, while the other is that “good diplomacy is the best way to achieve peace.” In Pew Research telephone surveys, fewer respondents chose military strength in the forced choice format, compared with the agree/disagree format. For both versions of the question, Google Consumer Surveys produced nearly identical results to the telephone surveys.

Reactions to the Presidential Debates

In a series of tests after each presidential debate, the Pew Research surveys and Google Consumer Surveys produced similar reactions. Both approaches found that Romney was widely viewed by registered voters who watched the debate as doing the better job. Romney had a 72% to 20% margin over Obama in the Pew Research survey on who did the better job in the first debate.

Similarly, Romney had a 57% to 16% lead over Obama according to the Google Consumer Surveys reaction, with 27% saying both candidates did about the same. In the Google reactions, Romney’s lead widened from the night of the debate to the next day.

By contrast, Obama was seen as winning the second debate and third debates, but by more modest margins. By a 48% to 37% margin, more debate watchers said in the Pew Research survey that Obama did the better job in the second debate. The Google Consumer Surveys reaction showed similar results: 50% said Obama did the better job while 32% said Romney did the better job. Views about who did the better job in the second debate changed little from the night of the debate through the following weekend.

Registered voters who watched the second debate also were asked using Google Consumer Surveys for a one-word impression of Obama and Romney in the debate. The top reactions to Obama’s performance included “liar,” “great,” “president” and “strong.” For Romney, the top reactions included “presidential,” “liar,” “awesome” and “great.”

Both the Pew Research survey and Google Consumer Survey showed Obama winning the third presidential debate, but the margin was much wider in the Pew Research survey. In the Pew Research survey, voters by a 52% to 36% margin said Obama did the better job. The Google survey found 43% of voters saying Obama did a better job vs. 37% for Romney.

The public’s reaction to the vice-presidential debate was divided in both the Pew Research survey and Google Consumer Survey. Among voters who watched the vice-presidential debate, 47% said Joe Biden did the better job while 46% said Paul Ryan did the better job, according to the Pew Research survey conducted Oct. 12-14. The Google Consumer Surveys reaction, conducted over a similar period, also found a divided reaction to the vice-presidential debate; 38% said Biden did the better job while 42% chose Ryan; 20% said they did the same.

Assessing Google’s Inferred Demographics

The demographic characteristics Google uses in sampling and weighting and what it provides for use in analysis are inferred based on information about the types of websites respondents have visited as recorded in their DoubleClick advertising cookie. But there is no publically available analysis of how well these inferred demographics match up to actual demographic information as reported by respondents. To assess this, Google Consumer Survey respondents were asked their gender and age so that the survey responses could be compared to the inferred data.

For 75% of respondents, the inferred gender matched their survey response. About eight-in-ten whom Google inferred were men (79%) said they were male when asked. Similarly, 72% of women based on Google’s inferred information said they were female when asked. Among those for whom Google did not infer gender, 58% said they were male and 42% female.

For age, the pattern is more mixed. Because Google limits the number of response categories for an individual question to five but provides inferred age in six categories, age was asked twice, of separate samples of respondents, collapsing different age categories for each.

In the first comparison, from 23% to 65% report an age that was in the same category as their inferred age, that averages to about 44% among all respondents. But when adjacent age categories also are included, about 76% report an age that is the same or close to their inferred age by Google.

Although there are errors at the individual respondent level in Google’s inferred demographic information, especially for those in the middle age-ranges, correlations between substantive questions and gender and age are consistent with those found in Pew Research surveys.

For example, on the question of whether people prefer a smaller government or a bigger government, more men than women said they prefer a smaller government in both the Pew Research survey and the Google survey. The age pattern also was similar, with younger people being less likely in both surveys to prefer a smaller government.

In both surveys, men and women were about equally likely to say they always vote. And in both the Pew Research survey and the Google survey younger people were far less likely than older people to say they always vote.

The age pattern on presidential approval was quite similar in the Pew Research survey and Google Consumer Surveys; young people were more likely to approve of the job Obama is doing as president in both samples. However, fewer older people using Google Consumer Surveys approved of Obama’s job performance than in the Pew Research survey.

Overall, Obama benefited from relatively strong turnout – both nationally and in key battleground states – among young people and minorities. Obama won voters younger than 30 by a somewhat smaller margin than he did four years ago, but these voters made up about as large a share of the electorate as they did in 2008, according to national exit polls conducted by the National Election Pool. Moreover, African Americans, Latinos and Asian Americans backed Obama by huge margins.

Nationally, nonwhite voters made up 28% of all voters, up from 26% in 2008. Obama won 80% of these voters, the same as four years ago.

Obama’s support from nonwhites was a critical factor in battleground states, especially Ohio and Florida. In Ohio, blacks were 15% of the electorate, up from 11% in 2008. In Florida, Hispanics were 17% of the electorate, a slight increase from 14% in 2008. While minority compositional gains were not huge, they offset a strong tilt against Obama among white voters. Nationally, Romney won the white vote, 59% to 39%.

Owing to comparatively strong turnout from Obama’s base, the composition of the electorate in 2012 was strikingly similar to that of four years ago: Nationally, Democrats made up 38% of all voters, Republicans 32% and independents 24%. This enabled Obama to win re-election even though he lost the independent vote. Romney won independents by a 50% to 46% margin. Four years ago, Obama carried independents by 52% to 44%.

Evidence of the demographic transformation of the electorate is seen in the profiles of each candidate’s supporters: Fully 89% of Romney’s were white non-Hispanics, compared with just 56% of Obama’s supporters. Romney managed to better McCain’s showing among whites by four percentage points – and still lost the election.

Continued support from women was also an important factor that helped Obama stay above water in the swing states of Florida, Virginia and Ohio. He held or improved his 2008 margins among women in all three states, while losing support among men. Nationally, Obama won women voters by nearly as much he did in 2008.

Obama’s advantages among women, minorities and young people helped offset his continued weakness among white voters. Obama’s falloff in white support was broad-based, crossing many demographics – working class whites, college graduates and even women.

The economy was clearly the dominant issue of the election, but the extent to which voters held Obama accountable for their economic frustrations was limited. Nationally, 59% said the economy is the most important issue facing the country, but voters overall were divided over which candidate could do a better job on the issue: 48% said Obama, while 49% said Romney. However, by a 53% to 38% margin, voters blamed Bush, not Obama, for the economic problems the nation is facing.

Obama also was aided by nascent economic optimism. By a 39% to 30% margin, more said the national economy is getting better rather than getting worse, and given the very strong correlation between that opinion and vote choice, that cut in Obama’s favor.
When it comes to how people feel about their own financial situation, more said they are worse off today than four years ago than better off, and most of those who expressed that opinion supported Romney. Yet importantly, a plurality of voters (41%) said they are doing the about same today as four years ago. These voters backed Obama by a wide margin (58%-40%).

Voters’ views of Obama were similar to those of George W. Bush eight years ago. After falling below 50% in job approval for much of the year, he reached 53% approval on Election Day – the same as Bush’s in 2004.

The biggest national trend in Romney’s favor was the public’s distaste for more expansive government. More voters said government is doing too many things than said government should do more, by 51% to 43%. But a populist streak regarding the lack of fairness in the system cuts the opposite way. Most voters (55%) say the U.S. economic system generally favors the wealthy, compared with 39% who say it is fair to most Americans. And a majority thought that Romney’s policies would favor the rich, while a plurality thought Obama policies would favor the middle class.

Romney improved his personal image substantially over the course of the campaign, but in the end he still was not as viewed as favorably as Obama. The national exit polls found that 47% viewed Romney favorably, up substantially from a low of 29% of registered voters in March during the bruising GOP primary. However, on Election Day, 53% viewed Obama favorably.

Just over half of the public (53%) say they followed news about Hurricane Sandy and the storm’s impact very closely last week, outpacing interest in the 2012 presidential election (47% very closely) and news about the U.S. economy (38% very closely).

The latest national survey by the Pew Research Center for the People & the Press, conducted Nov. 1-4, 2012, among 1,011 adults, finds that interest in the storm is particularly high in the Northeast: fully 73% of those living in this region say they are following news about Sandy very closely, compared with 53% of those in the South, 46% of those in the Midwest and 43% of those in the West.

Interest in Hurricane Sandy is far higher than for Hurricane Isaac earlier this year (31% very closely) and ranks as one of the most closely followed storms in Pew Research surveys dating to 1989. Interest is comparable to that for Hurricane Ike in September 2008 (50% very closely), but remains well below the 70% who were following Hurricane Katrina very closely in September 2005.

Interest in 2012 Campaign Lags Behind 2008

Overall, 47% say they have been following news about the candidates for the 2012 presidential election very closely, down modestly from 52% a week ago. Campaign news interest has lagged behind measures from comparable points in 2008 in four out of the past five weeks. And the high mark for campaign interest measured in 2012 (52% in the Oct. 25-28 survey) is lower than the high measured in 2008 (61% in the Oct. 17-20, 2008 survey).

Views of Economic News

As has been the case throughout 2012, most Americans (55%) say they are hearing a mix of good and bad news about the U.S. economy. A third (33%) say they have been hearing mostly bad news about the economy, up slightly from October (28%). Just 10% are hearing mostly good news about the economy, though this, too, has risen since the summer. In August, just 3% said they were hearing mostly good economic news.

The percentage of independents who say they are hearing mostly bad news about the economy has risen over the past month. In the current survey, 38% of independents say the economic news they have been hearing is mostly bad, up from 29% at the start of October. Views of Republicans and Democrats are little changed.

Less Bad News about Gas Prices

While overall views of economic news have shifted only slightly, news about gas prices has improved dramatically since early October. Overall, half (50%) say they are hearing mostly bad news about gas prices, down 22 points from a month ago. More now say they are hearing a mix of good and bad news (30%) or mostly good news about gas prices (16%) than did so in October (21% and 4%, respectively).

News about other economic sectors has been more stable over the past month. Overall, 46% say they have been hearing mostly bad news about the job situation, 37% say they have been hearing mixed news, and just 14% say they have been hearing mostly good news.

News about real estate and financial markets remains mixed, with little change in views from October.

Barack Obama has edged ahead of Mitt Romney in the final days of the presidential campaign. In the Pew Research Center’s election weekend survey, Obama holds a 48% to 45% lead over Romney among likely voters.

The survey finds that Obama maintains his modest lead when the probable decisions of undecided voters are taken into account. Our final estimate of the national popular vote is Obama 50% and Romney 47%, when the undecided vote is allocated between the two candidates based on several indicators and opinions.

A week ago the race was deadlocked, with each candidate drawing support from 47% of the likely electorate. Interviewing for the final pre-election survey was conducted Oct. 31- Nov. 3 among 2,709 likely voters. The previous survey was conducted Oct. 24-28, before Hurricane Sandy made landfall along the East Coast.

Obama’s handling of the storm’s aftermath may have contributed to his improved showing. Fully 69% of all likely voters approve of the way Obama is handling the storm’s impact. Even a plurality of Romney supporters (46%) approve of Obama’s handling of the situation; more important, so too do 63% of swing voters.

Voter turnout, which may be lower than in 2008 and 2004, remains one of Romney’s strengths. Romney’s supporters continue to be more engaged in the election and interested in election news than Obama supporters, and are more committed to voting.

The survey also indicates that voters in the nine battleground states are as closely divided as the national electorate: 49% of likely voters in battleground states support Obama while 47% back Romney.

There are many good signs in the poll for Obama. He has regained much of the ground he lost following his lackluster performance in the first presidential debate. In mid-September, Obama led Romney by eight points among likely voters, but in early October, shortly after the debate, he trailed by four points.

Nearly four-in-ten (39%) likely voters support Obama strongly, while 9% back him only moderately. A third of likely voters support Romney strongly, compared with 11% who back him moderately. In past elections, dating to 1960, the candidate with the higher percentage of strong support has usually gone on to win the popular vote.

Similarly, a much greater percentage of Obama supporters than Romney supporters are voting for him rather than against his opponent (80% for Obama vs. 60% for Romney), another historical indicator of likely victory. And far more registered voters expect an Obama victory than a Romney victory on Nov. 6 (52% vs. 30%).

Obama’s increases in likely voter support are most notable among women, older voters, and political moderates. Women now favor Obama by a 13-point margin (53% to 40%), up from six points a week ago and reflecting a shift toward Obama since early October. Right after the first presidential debate, the women’s vote was split evenly (47% each). Men, by comparison, favor Romney by a 50% to 42% margin, with little change in the past month.

Romney continues to lead among voters age 65 and older, by a nine point margin (51% to 42%) in the current survey. But that is only about half of the 19-point lead he held among seniors just a week ago. Political moderates now favor Obama by 21 points (56%-35%).

Another notable gain for Obama, perhaps reflecting Hurricane Sandy’s effect on the race, comes in a region he was already secure in: the Northeast. He has increased his lead over Romney from nine points (52%-43%) to 21 points (56%-35%) there over just the past week. While the storm’s impact on many parts of the Northeast has been substantial, an analysis of the polling data shows no substantial underrepresentation of voters in the most heavily affected counties.

Early voting makes up a large share of votes cast, but there is no sign that they are breaking decisively toward either candidate. Among the 34% of likely voters say they have already cast their ballot, 48% say they supported Obama, 46% Romney. That is roughly the same margin as among likely voters who plan to vote on Election Day (47% Obama, 45% Romney).

In the final days before Tuesday’s election, most of the focus will be on those likely to cast votes. But a sizable minority of adults choose not to vote or are unable to vote. By their absence, they also will affect the outcome. Nonvoters are numerous; in 2008, they constituted about 43% of the voting age population.

Nonvoters favor Barack Obama over Mitt Romney by a wide margin (59% to 24%). While most nonvoters (64%) have a favorable view of Obama, just half as many (32%) view Romney favorably.

By contrast, likely voters are evenly divided in Pew Research’s most recent national survey (47% Obama, 47% Romney). Nearly identical percentages of likely voters view Obama and Romney favorably (51% Obama, 52% Romney).

A plurality of nonvoters are independents (44%); 29% identify as Democrats and just 17% as Republicans. Likely voters include about the same percentages of Democrats (35%) as Republicans (34%); 27% of likely voters are independents.

About half of nonvoters (52%) either identify as Democrats or lean Democratic; only 27% identify as Republicans or lean Republican. Likely voters’ leaned party identification is evenly divided: 47% identify with the Republican Party or lean toward the GOP, while the same percentage identify as Democrats or lean Democratic.
However, just a quarter of nonvoters describe their political views as liberal; that is little different from the percentage of liberals among likely voters (20%). But nonvoters are far less likely than voters to describe their political views as conservative (28% vs. 44% of likely voters).

As might be expected, nonvoters express very little interest in politics or the election. A third of nonvoters say they are registered to vote. But they are far less likely than voters to give a lot of thought to the election and follow public affairs.

Demographic Differences Between Voters, Nonvoters

Nonvoters are younger, less educated and less affluent than are likely voters. More than a third (36%) of nonvoters are younger than 30, compared with just 13% of likely voters.

Just 13% of nonvoters are college graduates and about the same percentage (14%) have family incomes of $75,000 or more. Among likely voters, 38% are college graduates and a third (33%) have family incomes of $75,000 or more.

While most voters are married, most nonvoters are not. Nearly two-thirds (64%) of nonvoters are unmarried, compared with 40% of likely voters.

Nonvoters also are much more likely than voters to be Hispanic: 21% of nonvoters are Hispanic, which is three times the percentage of Hispanics among likely voters (7%). About six-in-ten (59%) nonvoters are white non-Hispanics. By contrast, white non-Hispanics make up about three quarters (74%) of likely voters.

Nonvoters and Voters on the Issues

Nonvoters express more liberal opinions than voters on a wide range of domestic and foreign policy issues. But the differences are far less pronounced on social issues, such as gay marriage and abortion.

Far more nonvoters than voters favor activist government. About half of nonvoters (52%) say the government should do more to solve problems, while 40% say the government is doing too many things better left to businesses and individuals. The balance of opinion is reversed among likely voters: 56% say the government is doing too much, while 39% say the government should do more to solve problems.

By 46% to 31%, more nonvoters favor keeping the 2010 health care law in place than repealing the law; 23% do not express an opinion. Voters are more evenly divided, with 49% favoring the law’s repeal and 43% saying it should remain in place; just 8% do not express an opinion.

There are smaller differences in the opinions of voters and nonvoters on tax policy: 55% of nonvoters and 51% of likely voters say it would better for the economy to maintain current tax rates on income under $250,000 while raising taxes on income about that level. A third of nonvoters, and 42% of likely voters say it would better to lower tax rates for all Americans by 20%, while limiting some tax deductions.

On foreign policy, a clear majority of likely voters (56%) favor withdrawing U.S. forces from Afghanistan as soon as possible; 40% say they should remain until the situation there is stabilized. Support for an immediate withdrawal is even more widespread among nonvoters (67% favor).

Nonvoters also are less supportive of taking hard line against Iran’s nuclear program: 45% say it is more important to take a firm stand against Iran, while 40% say it is more important to avoid a military conflict with Iran. Among likely voters, 62% favor taking a firm stand against Iran while just 31% say it is more important to avoid a military conflict.

On gay marriage and abortion, however, there are at most modest differences between nonvoters and likely voters. And there are no significant differences in opinions about the impact of immigrants on the country.

Identical percentages of nonvoters and likely voters favor allowing gays and lesbians to marry legally (49% each). Somewhat fewer nonvoters oppose gay marriage (36% vs. 42% of voters ), while more have no opinion (15% vs. 9%).

Nonvoters have the same views on abortion as likely voters: 54% of nonvoters say abortion should be legal in all or most cases, while 39% say it should be illegal in all or most cases. Among likely voters, 55% favor legal abortion in all or most cases; 38% say it should be mostly illegal.

Like voters, nonvoters express mixed opinions about the impact of the growing population of immigrants on the U.S.: 27% say it has been a change for the better, 34% a change for the worse, while 34% say it has not made much difference.