Category Archives: 2013 NJ Election

Additional Analysis from the Report on 2013 Rutgers-Eagleton Final Election Polls

Last week we released a report from Langer Research Associates commissioned by the Eagleton Institute of Politics examining the reasons for our mis-estimates of the U.S. Senate race in October and the gubernatorial race in November 2013. In addition to examining question order priming effects, which were found to be the primary cause of the mis-estimates, Langer Research Associates examined some general operational aspects of the Rutgers-Eagleton Poll to assess the degree to which any of them influenced the estimates. None of these items were found to be causes of the mis-estimates, but Langer’s assessment provides useful guidance for the Poll.

This week we provide a brief summary of the operational portions of the report, along with our responses, which in some cases include changes in some of our processes going forward. We do this as part of our commitment to transparency and our educational mission.


Likely Voter Modeling

Issue: Pre-election polls have to estimate who will vote, which is done through likely voter modelling. This means identifying through a series of questions which respondents are likely to turn out to vote and which are not.

Summary of Langer Analysis: The Rutgers-Eagleton Poll employed a series of questions including self-reported likelihood of voting, awareness of the election date (for the Special Senate election), reported last time voting, following news of the election, and attention to debates. The calculation of likely voter was applied independently to the Senate and Governor elections. The Poll employed a scoring methodology, which assigned points to each response. The likely voter modeling was conceptually sound. However the cutoff point used greatly overestimated actual turnout. Likely voter models that overstate turnout include non-voters in their vote-preference estimates, which can compromise the accuracy of these estimates.

Likely voter modeling, however, was not the culprit in the discrepancies in the 2013 Rutgers-Eagleton estimates. Constructing three tighter likely voter models with turnout estimates as low as 32 percent in the Senate race and 38 percent in the gubernatorial contest made no substantive difference in vote-preference estimates.

Rutgers-Eagleton Poll Response: The LV screens were relatively loose in these reports, due to the size of the original sample, which simply did not allow tightening screens since that would result in too small samples. The Rutgers-Eagleton Poll staff did test varying screens and found few differences, as also determined by the Langer analysis. So the looser screens were chosen for reporting. One potential revision for future LV screens would be to use propensity scoring rather than a cutoff approach, which would allow all cases where the likelihood of voting was greater than zero to remain in the sample, weighted to reflect their relative propensity to vote. One attempt to improve the estimates that was employed was an adjustment to reflect the greater likelihood of Republicans turning out, beyond what was appearing in the LV screens. Such an approach is not industry standard and should not be employed in the future. As it turned out, this adjustment made no significant difference in the estimates. But even if it improves an estimate, we agree with the Langer report that this approach departs from best practices and should not be employed.

The Rutgers-Eagleton Poll is currently involved in a broad research project to reassess the weighting process that we use, and we anticipate the results of that project will begin to be used in polls beginning in the 2014-2015 academic year.


Weighting

Issue: Non-response generally results in variation between the sample that is completed and target population norms which are based on U.S. Census data. One potential problem could be incorrect weighting of the sample prior to reporting the results. The Rutgers-Eagleton Poll is a random digit dial (RDD) survey, requiring that respondents be asked if they are registered voters in order to determine if they should be included in the sample for the purposes of asking election-related questions. For both the October and November 2013 polls, those who responded that they were not registered to vote were immediately terminated, meaning no additional questions were asked. Thus the samples are of registered voters only and must be weighted to norms for registered voters.

Summary of Langer Analysis: The Langer Report suggests that it is standard practice to weight to demographic variables for the full population, not to the registered voter population. To do so would require not terminating non-registered voters and at least asking them a series of demographic questions. Since the Rutgers-Eagleton Polls analyzed here were of registered voters only, this option was not available. The registered voter sample was weighted to the Census Bureau Current Population Survey (CPS) from March 2012 using age, gender, race and ethnicity as target demographics. Had the Poll sampled all adults, the all adult sample would have been weighted to the Census’s American Community Survey (ACS), the standard source for weighting population surveys. It is important to note that the CPS registered voter norms were from the 2012 election, making them 11 months old in 2013, which means they do not capture voter registration changes that might occur during the election season.

Rutgers-Eagleton Poll Response: The Rutgers-Eagleton Poll will continue to generally weight samples to age, gender, race, and ethnicity targets for the population from which the sample is drawn. We will make sure that we are using the most recent available norms at all times. We may do more “all adult” samples which will include subsamples of registered voters, and will allow the full sample to be weighted to current ACS norms. However, in doing so we will necessarily have smaller samples of registered voters since to increase the overall sample size would require additional financial resources not currently available.


Sampling, including cell phones

Issue: Surveys of NJ residents must be based on a probability sample of landline and cell phone respondents in New Jersey. In this survey, respondents were asked if they were registered voters and were terminated if they were not. A second area of investigation is the relative share of cell phone calls placed as part of the sample. NJ has one of the lowest cell phone-only penetration levels, but nonetheless a significant number of residents cannot be reached without dialing cell phones.

Summary of Langer Analysis: The sampling process appears appropriate for both cell phones and landlines. However, the termination of non-registered voters means the sample must be weighted to norms for registered voters, which are generally less current than adult population norms. As noted above, the Langer report suggests that non-registered respondents should be retained for the collection of demographic data before termination.

Given the increasing use of cell phones and in particular the increasing proportion of cell-phone-only households in the United States, the inclusion of a robust sample of cell phones is a necessary practice. The Rutgers-Eagleton Poll uses an overlapping dual frame sample that includes a sample of cell phone respondents regardless of whether or not they have landlines. Estimates from the federal National Health Interview Survey (NHIS) indicate that, as of December 2011, 16.5 percent of New Jersey adults used cell phones only, and an additional 24.7 percent relied “mostly” on cell phones. These proportions surely have increased since then. The October and November Rutgers-Eagleton polls, using an overlapping dual frame design, included 17 percent and 23 percent cell phone interviews (weighted to 22 and 26 percent, respectively), with 4 and 7 percent cell phone-only respondents (weighted to 6 and 8 percent, respectively), well below available NHIS estimates.

Rutgers-Eagleton Poll Response: The Rutgers-Eagleton Poll has increased the cell phone target to 30 percent of the sample and will monitor whether this provides a reasonable share of cell phone-only households. We are also collecting additional information from respondents including the number of adults in the household (for landlines) and the number of adults sharing a cell phone (for cell respondents). These data will help with improving weighting calculations.


Question Wording & Field Dates

Issue: Field dates and question wording are other potential causes of differences in survey estimates.

Summary of Langer Analysis: The review finds no indication that either field dates or wording influenced Senate or gubernatorial vote preference estimates in these surveys. Question wording, while different in each survey, in all cases was balanced and neutral.

Rutgers-Eagleton Poll Response: Question wording was slightly different between the Senate and gubernatorial head-to-head questions. The biggest difference was that voters who responded don’t know in the gubernatorial question in October we not asked about which way they leaned. They were asked this in November, and the Senate vote asked about leaners in October.

 

Leave a comment

Filed under 2013 NJ Election, NJ Senate 2013 Special Election

Analysis of Rutgers-Eagleton 2013 Pre-election Polls Released

Following inaccurate results for final pre-election polls in October 2013 (NJ Special Senate) and November 2013 (NJ Governor), the Eagleton Institute of Politics commissioned an outside study by Gary Langer of Langer Research Associates of New York to identify reasons for the outcomes of these polls. Today, The Eagleton Institute of Politics and Eagleton Center for Public Interest Polling are releasing this analysis to the public as part of a commitment to transparency and education.

The Rutgers-Eagleton Poll reported a final pre-election poll for the special Senate election between then-Newark Mayor Cory Booker, a Democrat, and Republican former Bogota Mayor Steve Lonegan in which Booker held a 22-point lead. Booker ultimately won by 11 points. In the final November gubernatorial pre-election poll, Rutgers-Eagleton had Republican Governor Chris Christie ahead of his Democratic challenger state Senator Barbara Buono by 36 points: Christie won by 22.

The Langer report identifies the primary reason for the inaccurate results as the failure to put the “head-to-head” questions, which asked respondents for their vote intention, at or near the beginning of the questionnaire. Because these questions were asked after a series of other questions, it appears that respondents were “primed” to think positively about Governor Chris Christie in the November survey, which then may have led Democrats and independents in particular to over-report their likelihood of voting for the Governor. A similar process occurred with the October Senate poll, where voters were first reminded of how little they knew about Lonegan and how much they liked Booker before being asked the vote question.

Ruth B. Mandel, director of the Eagleton Institute of Politics stated that, “In response to these results, Eagleton chose to contract with an independent, highly respected, outside survey research firm to review its recent work and offer suggestions for improvement.” She added, “The Institute is committed to contributing to political knowledge in New Jersey and nationally with credible, impartial data. When we saw we had a problem, we knew we had to learn why and what to do about it.”

“Gary Langer and his colleagues spent many hours examining multiple aspects of our polling to understand what went wrong,” said David Redlawsk, director of Eagleton’s Center on Public Interest Polling (ECPIP) and professor of political science at Rutgers. “We are grateful for the efforts they put in and the advice they have provided, both in terms of this specific issue and general operations of the Rutgers-Eagleton Poll. The results of this report will make what we do even better.”

The Rutgers-Eagleton Poll has been a valued source of information about the views of New Jersey residents for over 40 years. As an academic-based survey research organization, ECPIP strives to be transparent and accessible. “We have a special obligation to take our educational mission seriously, which includes informing the public as well as learning from our own errors.” Redlawsk notes that survey research results released by the Rutgers-Eagleton Poll, for example, aim to meet the transparency standards set by the American Association of Public Opinion Research (AAPOR). Further, in recent years, the Rutgers-Eagleton Poll has been providing open informal insights and perspectives about survey research from Redlawsk and members of his staff through its blog at http://eagletonpollblog.wordpress.com. And for many years full data from the Poll has been freely available, generally after a one-year period, at http://eagleton.libraries.rutgers.edu/.

Langer’s major finding is that the order in which the head-to-head ballot test questions were asked most likely added inadvertent bias to the results in both the October and November Polls, although the results came out in opposite partisan directions in the two polls. Decisions made by ECPIP to maintain the standard set of questions about political figures including Cory Booker and ratings of Chris Christie at the beginning of the questionnaire worked to particularly prime Democrats in the November poll and Republicans in the October poll to support the candidate from the other party – Christie or Booker.

Redlawsk noted that the cause was a decision to maintain an ongoing four-year series of questions about Governor Christie that have been asked at the very beginning of a Rutgers-Eagleton NJ Poll since the governor’s inauguration. “We made this decision purposefully to maintain the integrity of our time series,” said Redlawsk. “This long-term research has greatly informed our understanding of public opinion about Governor Christie, and we had concerns that moving these questions after a head-to-head vote question would bias those results for the same reason we ended up biasing the vote questions.”

Most pre-election head-to-head polls focus only on the election and do not include long batteries of additional questions. The Rutgers-Eagleton Poll was unable to field separate pre-election surveys and thus combined the head-to-head polls with the regular surveys of New Jersey public opinion. “In retrospect, this was the wrong choice when one goal was to be as accurate as possible with pre-election numbers,” noted Redlawsk. “We should have either fielded a separate poll or just focused on our long-term work, rather than trying to do both at the same time.”

The Langer report on the cause of the pre-election poll mis-estimates is available to the public now on the Rutgers-Eagleton Poll website at http://eagletonpoll.rutgers.edu (PDF).

 

Leave a comment

Filed under 2013 NJ Election, Buono, Chris Christie, Cory Booker, NJ Senate 2013 Special Election, Steve Lonegan

Update: Reviewing our pre-election polls

As was obvious, the Rutgers-Eagleton Poll was pretty far off on both pre-election polls last fall. In October we overstated the vote for now-Senator Cory Booker, while in November we also overstated the vote for Gov. Chris Christie. Over the last month we have been working with an outside consultant to examine our processes and results for those two pre-election polls. We anticipate having a report very soon. One thing that is very clear is that while our head-to-head vote was wrong, the other routine data we collected in those polls continued to track with both our prior results and the results of other polls at about the same time. So the focus is on our head-to-head questions, their design, and location in the questionnaire in particular, as well as our weighting strategies. In the end, we hope to be able to learn what we need to allow us to do a better job with pre-election polling.

These problems have given us pause as we move forward on our regular Rutgers-Eagleton polling. We spent a lot of time reviewing all of our numbers and benchmarks. If we had seen problems in the questions that did not ask directly about the vote (like job performance and favorability questions) we would be much much worried about our regular polls. But we do not see significant problems there. And as we run benchmarks against other polls taken at the same time, we see convergence in general, other than the vote itself.

Once we have the report from our outside consultant we will release it here. In the meantime, we are looking carefully at our newest polling results to ensure that we are confident in the numbers before we release them.

Leave a comment

Filed under 2013 NJ Election

#1 Gov Chris Christie – Rutgers-Eagleton Poll’s 2013 Top 5 Countdown

1.) Chris Christie’s banner year of unprecedented popularity ends with a reelection landslide.

Since Hurricane Sandy struck New Jersey in October 2012, Gov. Christie’s ratings have seen unprecedented highs throughout 2013, ending the year with 65 percent of voters having a favorable impression toward him, 68 percent approving of the job he is doing, and 59 percent awarding him an A or B as governor. Christie – following his own administration’s motto – has certainly been stronger since the storm, with his widely acclaimed Sandy leadership and resulting personal popularity driving his reelection victory over Democratic State Sen. Barbara Buono in this past November’s gubernatorial election. The governor easily won a second term by double-digits on Election Day against his virtually unknown, unsupported, and under-funded opponent – after a campaign year that continually showed little contest and increasingly good news from Christie, who received unparalleled support from across the political aisle, independents, women, and minority voters. Christie has become a media darling and front-page news both state and nationwide, and a 2016 presidential bid seems all but inevitable as early polling shoes him to be a top contender. The governor’s Sandy-driven success has all come in spite of New Jerseyans’ lower ratings of Christie on key issues that they deem most important in the state – 44 percent of NJ voters still disapproved of Christie’s handling of the economy and 50 percent disapproved of his job on taxes when last polled in November. Only time will tell how long the rally around Christie will last as he enters his second term and as the national spotlight placed on him continues to grow brighter.

Christie

And that’s it for our Top 5 for 2013. There were plenty of other stories over the year, and lots of additional polling. But these are the ones we think stand out.

All the best for a  Happy New Year from the Rutgers-Eagleton Poll!

1 Comment

Filed under 2013 NJ Election, Chris Christie

A Closer Look by the ECPIP Staff… Gender Effects in the 2013 NJ Gubernatorial Election?

The Candidate Gender Effect That Wasn’t?
The Subtle Role Gender Played in the 2013 New Jersey Gubernatorial Election

Ashley Koning

Ashley Koning is Manager of the Rutgers-Eagleton Poll and a Ph.D. candidate in Political Science at Rutgers University.

The past election cycle in New Jersey was notable for many reasons, not least of which was that State Sen. Barbara Buono was the first woman to run for the executive office as a major party candidate since Gov. Christie Todd Whitman more than a decade ago.  It initially seemed like the governor’s race could have been the perfect storm for a gendered campaign, given Buono’s stark contrast to Gov. Christie – the very embodiment of “Jersey guy” toughness – and her partisanship, ideology, and agenda that revolved around the types of compassion and social issues frequently stereotyped as feminine, “communal,” and emotionally charged.  After all, we know through a whole host of research that the entire process is gendered for women candidates from selection through Election Day as they continue to be stereotyped – even if not intentionally – by both voters and by the media (see the Center for American Women and Politics or the Political Parity project for more on this).  Gov. Whitman even speculated decades ago that her gender affected pre-election polling during her run for governor back in 1993, consistently underestimating her eventual win (though this has been debated in the literature, both for and against).

Yet it was Gov. Christie who cried foul about his physical appearance being put in the campaign spotlight after a comment he claimed Buono made that alluded to his weight issues.  And the only gender gap the media spoke of was the one in Christie’s favor among women voters.  In a race where gender could have potentially had a significant impact, any type of gender effect from Buono’s candidacy seemed to be non-existent.

But sometimes there is more than meets the eye in these cases.  Therefore, in our final pre-election poll from October 28 – November 2, we conducted a small “list” experiment to test whether Buono’s gender actually had any effect on why respondents thought she was behind in the polls throughout the entire campaign.  For a campaign season that rarely – if ever – explicitly mentioned Buono’s gender as a pro or con to her candidacy, the results of our survey experiment paint somewhat of a different story.

First, what is a “list experiment”?

List experiments have increasingly become a popular research tool in political science as a way to unobtrusively tease out attitudes towards sensitive subjects that respondents may not want to express explicitly – whether because of a wish to sound socially desirable and not go against commonly held societal beliefs, to maintain congruency within oneself regarding other views and aspects of his or her own life, or in order to not disparage the interviewer based on perceived characteristics such as gender and race.  Respondents only become more comfortable expressing socially undesirable or unacceptable answers when a degree of anonymity can be obtained.  Therefore, survey list experiments enable this desired anonymity by allowing respondents to not have to directly state their views to the interviewer but rather provide a much more coded and private – and thus hopefully more accurate – response.

In a list experiment, the survey sample is randomly split into different groups; one group acts as the baseline or “control” condition group, while the one or more remaining groups act as the experimental condition(s) to test the sensitive issue in question.  All groups are then given a list and asked only to specify how many items on the list they would choose; those in the experimental group(s) have one additional item on their list related to the target issue than those in the control group.  Respondents then report only a number based on how many items they support or oppose, never specifying exactly which item(s) was (were) left out.

For example, we asked the following question to more subtly explore the impact of gender on Buono’s campaign:

“There has been a lot of speculation about why Barbara Buono has been so far behind Chris Christie in the polls during this entire election season.  I am going to read you three possible reasons for it. I would like to know how MANY of these reasons you agree with.  Just tell me how many, I do not need to know which ones.”

 We then split the sample in half, randomly assigning half of our respondents to each group.  The first half were given three statements about Buono in a random order and asked whether they agreed with none, one, two or all three of them:

Her views are too liberal for New Jersey.
She is not well known throughout the state.
She did not receive enough support from the Democratic Party.

The other half of the sample received the same three statements, plus a fourth statement that addressed Buono’s gender.  All four were presented in a random order, and these respondents were asked whether they agreed with none, one, two, three, or all four of them. The fourth statement was:

            She is a woman.

Since the ONLY thing that differs between the two lists is the addition of the target item in the second list, we can use some very simple statistics to see if it ha any effect, that is, to see if  some voters think Buono was disadvantaged by being a woman. Because the two groups are random, if both got the same list we should get about the same mean number of responses for both groups. But since the second group has an extra item – the target item – any difference in the mean between voters in the second group and those in the first has to be because of the extra item.

So we compare the means (averages) of both groups, and look at the difference between the first group (the “control” list) and the second group (the “experimental” list) to determine the percentage of respondents who supported or opposed the target item.  While we will not know whether individual respondents agreed or disagreed with this item, we will get a more accurate read on  how many voters felt Buono’s gender was a disadvantage than we would if we asked the question outright..

So, was there a candidate gender effect?

The short answer is yes, but a modest one.  The first “control” group that received only three items selected 1.77 items on average.  The second group – our “experimental” group – that received the four items selected 1.92 items on average.  We take the difference of these two averages and multiply it by 100 to get the percentage of respondents who actually agreed that Buono’s gender influenced her continual double-digit lag behind Christie in the polls throughout the campaign. The result is 15 percent of NJ voters thought Buono being a woman candidate contributed to her support deficit during the race.

We know that there are no significant differences in characteristics between the two random groups – such as respondent gender, age, race, partisanship, ideology, and vote choice.  Furthermore, the only difference between the experimental versions assigned to these two groups is the item specifying Buono’s gender; all other items are the same, and all items are presented in a random order.  Therefore, we can with a great deal of confidence say that this extra gender item is making the difference – and according to tests of statistical significance, this difference is not just by chance.

The Candidate Gender Effect That Not-so-evidently Was

While 15 percent may not initially seem like a large amount, that is about one in seven voters who believed Buono’s gender to be an issue in her performance during the campaign.  If we generalize this a bit to the more than two million New Jersey voters who voted for governor this past election, that means more than 300,000 voters may have believed in some way that Buono being a woman candidate contributed to her struggles.  While Buono no doubt eventually lost for a large variety of reasons – Christie’s untouchable and skyrocketing post-Sandy popularity, lack of support from Democrats at the elite and mass level, and difficulties fundraising compared to Christie’s insurmountable war chest just to name a few – her gender may be one of them despite the apparently non-existent role it seemed to play publicly.

Buono in fact did a few points better in the actual election results than in the last pre-election average forecast by Real Clear Politics, and though Christie did slightly better than the average predicted as well, Buono’s increase from the average to the actual was twice Christie’s (Christie went up 1.6 points from an average of 58.8 percent to an actual of 60.4 percent; Buono went up 3.4 points from an average of 34.7 percent to an actual of 38.1 percent).  Perhaps there is some more support here for the polls underestimating Buono in part due to gender, but the clearest takeaway here is that gender still plays a notable role in campaigns – even when it seemingly does not.

Leave a comment

Filed under 2013 NJ Election, Buono, Chris Christie

A Closer Look by the ECPIP Staff… Internet use in the Gubernatorial Election

The Internet, the Gubernatorial Election, and Age

Ian McGeown, Liz Kantor and Max Mescall

Ian McGeown is an Aresty Undergraduate Research Assistant with the Eagleton Center for Public Interest Polling, and is a sophomore at Rutgers University. Liz Kantor is an Aresty Undergraduate Research Assistant with the Eagleton Center for Public Interest Polling and a School of Arts and Sciences Honors Program at Rutgers University in the class of 2016. Max Mescall is a research intern at the Eagleton Center for Public Interest Polling and a junior at Rutgers University.

In recent years, the Internet has become a wealth of information on nearly every topic. For example, many politicians and institutions have a presence on a host of social media sites – such as Facebook, Twitter and YouTube – as do newspapers, who more and more frequently put their own content online. Notable generational gaps in how much news is obtained through Internet sources exist, however (for a great summary of generational differences in various online activities in general, see here: http://www.pewinternet.org/Infographics/Generational-differences-in-online-activities.aspx). The recent gubernatorial election in New Jersey highlights this trend.

Unsurprisingly, younger generations tended to get information on the 2013 governor’s race in New Jersey through social media sites more than older generations. For 18-39 year olds, for example, 16 percent got information about the race on Twitter, while only 6 percent of 40-64 year olds did the same; no one over the age of 65 received information from the same site. Facebook also generated a similar trend, with 32 percent of those under 40 obtaining some information about the race from the site.  This is compared to only 21 percent of those between the ages of 40 and 64, and 11 percent of people 65 and older, who got information from here. Meanwhile, 26 percent of people between 18 and 39, 11 percent of those between 40 and 64, and 10 percent of people over 65 years old reported getting information off of the video sharing website YouTube.

This trend ceases, however, when it comes to “traditional” online media sources, such as online newspapers and blogs. Here, there is only a small 5-point difference between the youngest and oldest age brackets, with 15 percent of the under 40 crowd reporting that they got information on the gubernatorial election from these sources, while 10 percent of those between the ages of  40 and 64, and 13 percent of those 65 and over, state the same. These numbers suggest that it’s not necessarily using the Internet itself for election information that promotes generational gaps but rather the types of sources being accessed on the Internet.

In sum, among those who used the Internet to get information on the governor’s race, younger voters were most likely to use social media sites – particularly Facebook – more than traditional news outlet sites or blogs, as well as more likely to use social media than older voters.  Yet it also seems that everyone is using Facebook the most overall, as it was the main source of information for the governor’s race among 40-64 year olds and a close second for Internet users who are 65 and over; these Internet users in the oldest age group tended to go to blogs and news sites slightly more for their election information.  Twitter was the least popular among Internet users within each age group.

Therefore, the Internet is making it increasingly easier to get news and information about politics, but not everyone chose this method during the gubernatorial campaign season this past year – and even among those who did, differences remained in what types of sites they choose to access for information once online.  These differences are especially evident when analyzing voters by age.

internet

Leave a comment

Filed under 2013 NJ Election

A Closer Look by the ECPIP Staff…More on the Internet and the NJ Gubernatorial Campaign

The Internet, the Gubernatorial Election, and Vote Choice

 Caitie Sullivan and Mihir Dixit

 Caitlin Sullivan is a data visualization and graphic representation intern at the Eagleton Center for Public Interest Polling and a senior at Rutgers-New Brunswick with plans to graduate in January 2014.  Mihir Dixit is also a data visualization intern. He is a first year undergraduate student in the School of Arts and Sciences at Rutgers.

Both gubernatorial candidates this past election cycle in New Jersey were quite tech savvy throughout the campaign – maintaining Twitter accounts, Facebook pages, YouTube channels, and more.  It is therefore interesting to take a look at how big a role the Internet played among the electorate and how voters differently used the Internet throughout the campaign to learn about and interact with the candidates.  While Christie’s frequent usage of the web – particularly social media – is widely known, his supporters this past election season were surprisingly not as “connected” to the Internet as their governor or Buono supporters.  Whether Twitter, Facebook, YouTube, or blogs and news sites, those Internet users who voted for Buono were more likely than Christie voters to say they used these online tools.

The differences between Christie and Buono voters in what online tools they use mainly stem from social media usage.  More Internet-using Buono voters than Christie voters used Twitter to keep track of the election – 12 percent versus 7 percent.  There was also a clear difference in the percentage of Internet-using voters for each candidate who reported using Facebook to monitor the election. Twenty-one percent of Christie voters used Facebook to acquire information about the race, versus 26 percent of Buono voters.  The widest margin between Christie voters and Buono voters was for YouTube. While the video sharing website remained unpopular with 12 percent of Christie voters, almost double the number of Buono voters – 22 percent – reported having used it as an informational resource.  Christie and Buono voters were more similar in their use of blogs or news websites: 12 percent of Christie voters and 13 percent of Buono voters said they used this source for election-related news.

These differences in online sources are most likely attributable to the different characteristics underlying Christie voters and Buono voters and not necessarily directly to the vote choice itself.  Despite Christie’s frequent activity online, his supporters – many who are Republicans like Christie – tend to be older and therefore not as likely to use the Internet and especially social media.  Buono supporters, on the other hand, are made up of mainly Democrats and younger voters, who are especially prone to using the Internet.  Therefore, the differing usage of these various Internet tools between Christie voters and Buono voters may be more about the demographics typically most associated with each candidate’s party than simply who they would vote for in the election.

internet

Leave a comment

Filed under 2013 NJ Election, NJ Voters

A Closer Look by the ECPIP Staff… Internet Use and the NJ Gubernatorial Election

Clicking on and Connecting to the 2013 New Jersey Governor’s Race

Ashley Koning

Ashley Koning is Manager of the Rutgers-Eagleton Poll and a Ph.D. candidate in Political Science at Rutgers University.

The Internet has become an increasingly integral tool for both voters and politicians alike during campaign seasons, and this past year’s gubernatorial race in New Jersey was no exception.  Gov. Christie has always been technologically savvy – with Twitter accounts, a YouTube channel, thousands of hits on his YouTube videos, and a continually updated Facebook page.  Christie’s Democratic opponent, State Sen. Barbara Buono, was also quite active on social media and the Internet in general during the race, often relying on YouTube and Twitter to disseminate messages at low cost due to limited campaign funds.  Voters, likewise, had many opportunities to interact with the candidates online throughout the campaign –watching live streaming video of the debates, “liking” or “following” the candidates, reading live blog updates on Election Day, and more.

In our final pre-election poll from October 28 – November 2, we asked some follow-up questions to the 52 percent of voters who said they used the Internet in some way to get information about the governor’s race during the past election season.  These questions explored more in depth what Internet tools this subset accessed to get election news or interact with the candidates.  Internet-using voters were specifically asked about their usage of Twitter, Facebook, YouTube, blogs or news sites, and any candidate websites.

Unfortunately, in our first two days of polling, a coding error occurred in this internet source question that prevents us from looking at the details for respondents we interviewed those two days.  We can nonetheless explore responses for the remaining four days when the coding error was corrected.  As it turns out, even if we only look at the last fours days of our poll, we still find 52 percent saying they used the internet for the Gubernatorial campaign. Given the nature of our call center operation, this makes sense, as each day of calling collects a new random sample of NJ voters.

Among the 52 percent who used the internet to get candidate information, about two-thirds said they used one of the specific internet tools we asked about. The other third presumably used other means that we did not include.

Among the voters who used one of the five specified sources, most stuck to using just one or two instead of a combination of them: 58 percent used only one of the tools, 27 percent used two of them, and just 14 percent used three or more.

Candidate websites dominated source usage.  Almost half of all internet-using voters who received the follow-up questions – 45 percent – said they used the candidates’ websites to interact with the candidates or get information about the election.  No other source came close: only 9 percent specifically got information or interacted with the candidates via Twitter, 24 percent via Facebook, 16 percent via YouTube, and 12 percent via blogs or news sites.  When we look at the entire sample including those who do not use the Internet, the number of voters who used each Internet source becomes even smaller: only about 5 percent of all voters used Twitter, 12 percent Facebook, 9 percent YouTube, 6 percent blogs or news sites, and 23 percent candidate websites.

In general, Twitter users were typically the most or the second most likely to use one of the other four specified Internet sources: over two-thirds of Twitter users also used Facebook, over a third also used YouTube, more than 4 in 10 also accessed blogs, and over half also visited candidate websites.  But users of other sources were very unlikely to access Twitter, blog and news visitors being the most likely at 32 percent also using Twitter and candidate website visitors being the least likely at 11 percent also using Twitter.  Users of all sources were highly likely to access Facebook, on the other hand – especially other social media users (Twitter and Facebook).  But candidate websites were the most popular among all types of users, with anywhere from four in ten to six in ten users of other Internet source types visiting these particular sites.

Of course we know that these numbers will fluctuate if we take a closer look within different groups – especially among those groups that we know differ on Internet usage in general.  Therefore, throughout the next few days, some of our undergraduate staff members will take a closer look at how these trends in using the Internet during the governor’s race differ by important voter demographics – in particular, age and gubernatorial vote choice.

internet

 

Leave a comment

Filed under 2013 NJ Election, NJ Voters

Following up on our Booker – Lonegan Numbers from October

Did the Rutgers-Eagleton Poll have a “Bradley Effect” in Our Final U.S. Senate Results?

Bear with us, this is a LONG post…

In our final pre-Senate special election poll, we had Newark Mayor (and now U.S. Senator) Cory Booker up 22 points over his opponent, former Bogota Mayor Steve Lonegan. The real-world results were different – Booker’s margin was “only” 12 points or so. At the time we speculated on many reasons that our numbers could have been off on the head-to-head question, especially given that we did not see significant levels of variance with other polls on questions such as Booker and Lonegan favorability ratings. We speculated some more the day after the election looking at turnout, but also noting that we wondered if the fact that Booker is an African-American may have played a role. We have since done some fairly complex statistical analysis to examine this question. The upshot is that we see a very clear “race/ethnicity of interviewer” effect in our data; that is, our Black and Hispanic interviewers got more “Booker” votes from among the white respondents they talked to than did our white and Asian callers. And, our white callers got fewer “Booker” votes among Black and Hispanic respondents than did our non-white callers.

This is a complex phenomenon that has previously been documented by researchers, in particular in the aftermath of the 1993 Virginia governor’s race when polls badly overstated support for Doug Wilder, the African American candidate who won, but by a much smaller margin than expected. This is commonly been called the “Bradley Effect“. The argument is that respondents “guess” the race of callers and some will then adjust their responses to conform to what the believe is the caller’s expectation. Whether or not that is exactly what happens, the fact is that the data in our case seem to show exactly that happening.

Now, is the effect enough to account for being off by 10 points? That’s harder to calculate. However, our call center is very diverse – among the 113 student callers working on that poll, 25% were white, 19% Black, 47% Asian, and 11% Hispanic. Across the board our callers averaged about 7 completes per caller, with some variation by race/ethnicity. Overall, 22% of the 695 respondents for whom we have caller data were collected by white callers, 22% by Black callers, 46% by Asian callers, and 10% by Hispanic callers.

So here’s what we have – this is using all our respondents, NOT adjusting for Likely Voters. (Making that adjustment does not make any difference in our basic results.) First the unweighted responses to the question:  “Let’s talk about the Senate election in October. If the special election for the Senate seat were being held today and the candidates were [ROTATE ORDER: Democrat Cory Booker and Republican Steve Lonegan], for whom would you vote?”  (Note, that we did a followup to the don’t knows, asking how they “lean”. We will ignore this right now and focus only on the initial question.)

SNAG-005

Note we have a 22 point margin between Booker and Lonegan in the raw unweighted data, about the same as we had in the final weighted sample. The “Refused” represents people who would not answer the question at all, and the “System” are people who were not asked because they said in an initial screening question they would not be voting.

So what happens if we look at these responses by race of interviewer?

SNAG-003

Now we are only dealing with the 721 people who gave us a response to the question. Note that White interviewers got 50.3% support for Booker. But Black interviewers got 59.5%. Hispanic interviewers found even more Booker support: 62%. Finally, Asian interviewers (the largest group in our call center) found 49.9% support for Booker, pretty much the same as white interviewers.

Next we look at the percentage support for BOOKER by a combination of the Respondent’s race/ethnicity and the caller’s race/ethnicity. This now uses 697 respondents for whom we have their race (a significant number always refuse to answer that question.)

SNAG-002

The raw numbers (Total Column) show that 49.6% of these white respondents supported Booker, while Booker support was 91.1% of Black respondents, 80.0% of Hispanics, and 51.2% of other. Other in this case includes Asian, multiracial, and any other response to the question. These are essentially “normal” results in that we expect Black and Hispanic voters to be more supportive of Booker.

Looking at the Total ROW at the bottom, we see that for White callers, 50.7% of all their respondents supported Booker, with a similar result (50.5%) for Asian callers. But for Black callers, 60.1% of respondents supported Booker, while for Hispanic callers it was 65.2%; both are well above the total 56.3% Booker support among this set of respondents.

More importantly, note that WHITE respondents talking to WHITE callers gave Booker 49.2% support. But when talking to Black or Hispanic callers, white respondents were more likely to report a Booker vote, at 54.9% and 58.0% respectively.  This effect has been documented in the past, including in the Wilder race for VA governor in 1993.

We see another interesting effect with non-white respondents, though we have to be very careful here since we have relatively few of them, so any one group could be highly skewed. But in general, non-white respondents who talked to white callers, were less likely to report Booker votes than when they talked to non-white callers.

All of this is interesting but it doesn’t account for the possibility that callers of different races/ethnicities may have talked to different kinds of respondents. As a simple example, if white callers were more likely to talk to Republicans (regardless of respondent race), while non-white callers talked more to Democrats, we would see the same pattern but it would not be because of the race/ethnicity of the caller. To deal with this we must do a more complex multivariate analysis to control for these kinds of differences.

We won’t go into the details of the statistical analysis here, but it was designed to control for key factors that affect the vote choice – partisanship, ideology, and voter race/ethnicity, and voter gender. That means that we make sure the differences we see in the vote by caller race/ethnicity are NOT because of these factors. We added in one more control, that for what is termed in political science as “Racial Resentment” (see also here), a measure of “subtle anti-Black feeling”. We included this because Booker is African American and research has shown that this measure helps predict the likelihood of voting for a Black candidate.

By using multivariate statistics (specifically logistic regression) to predict the likelihood of a vote for Booker based on the controls above AND the race/ethnicity of our callers, we can examine the extent to which we see caller race/ethnicity conditioning poll responses. Follow is what we find:

SNAG-001

The first row of data shows all respondents by the race of the interviewer. Results are very similar to the initial table before we control for other factors. Across everyone, voters who talked to Black and Hispanic callers were more likely to say they would vote for Booker than those who talked to white and Asian callers.

As the table shows, there are differences across the race/ethnicity of respondents. Looking only at white voters, they remain more likely to tell Black and Hispanic callers they support Booker. For Black and Hispanic voters, talking to a white caller seems to lower the likelihood of reporting support for Booker, compared to talking to non-white callers. And because the model used for this prediction controls for partisanship and other factors, we are pretty confident that the results are in fact related to the race and ethnicity of callers and the race/ethnicity of voters.

To check this, we also ran similar models with the Buono-Christie responses from the same poll (where our results were in line with everyone else’s in mid-October) which show no effects for race/ethnicity of interviewer. Even more interesting, we also tested this model with the evaluation we asked voters to give to Booker (called a “feeling thermometer rating”) on a 0-100 scale, and we found no significant effects for race/ethnicity of callers. The issue seems limited to the question of the vote itself, and not other questions.

So what does this all mean?

For the Rutgers-Eagleton Poll, it means that our pre-election numbers which overstated Booker support were, at least in part, because we have a very diverse call center, probably much more diverse than any other call center that polled on this election. It also means we will have to look more carefully at how we handle election polling when there is a non-white candidate in the mix.

And it also means that in an election like this, with an African-American candidate, polling that does not use interviewers – like computerized polls where respondents listed to a computer ask the question and respond on their phone keypads, known as “interactive voice response” – may result in more accurate results, at least for those who can be reached this way. However, IVR cannot be used to call cell phones, so at a minimum it would be necessary to combined IVR with live calling of cell phones in order to get a reasonable sample of the population. This is what Monmouth did in its pre-election polls, apparently to good effect. IVR has other issues, though, and has to be looked at very carefully.

If you’ve made it this far in this very long post, congratulations! Bottom line for us: our final pre-election Booker-Lonegan poll was off by 10 points, overstating Booker’s numbers. We now think a least some significant part of that error is due to this race/ethnicity of interviewer effect as the evidence shows.

Of course, this does NOT explain our problem in the final Christie-Buono poll, where we were off by 14 points (showing Christie up 36 points while he won by 22.) Given the evidence from the October poll where our numbers for the governor’s race fit with other polling centers results, something else must have happened in our final gubernatorial poll.  Apparently we suffered from one problem in the Senate race, but something else in the race for governor. We’re currently moving forward on trying to understand what that might have been. We’ll report more on that effort in the (we hope) not-too-distant future.

2 Comments

Filed under 2013 NJ Election, Cory Booker, NJ Voters

NEW JERSEY VOTERS EXPECT CHRISTIE TO RUN FOR PRESIDENT; THINK HIS TOUGH TALK WORKS HERE, BUT MAY NOT ‘PLAY IN PEORIA’

Click here for a PDF of the text of the release along with questions and tables.

NEW JERSEY VOTERS EXPECT CHRISTIE TO RUN FOR PRESIDENT; THINK HIS TOUGH TALK WORKS HERE, BUT MAY NOT ‘PLAY IN PEORIA’

 NEW BRUNSWICK, N.J. – A clear majority of New Jersey’s registered voters expect Gov. Chris Christie to run for president in 2016, according to a new Rutgers-Eagleton Poll. As voters geared up for last week’s gubernatorial election, 59 percent said Christie would run, while 19 percent thought he would not. Twenty-two percent were unsure.

Forty-five percent of voters thought Christie will resign to run, and 33 percent expected him to finish his term, including those who said he would run without resigning. Twenty-three percent were uncertain whether the governor will complete his second term.

Yet in last week’s election, very few voters were influenced by these expectations: more than three-quarters said Christie’s future plans would have no impact on their vote. Only 8 percent said an anticipated run made them more likely to vote for Christie, while slightly more –13 percent – said his presidential ambitions made them less likely to vote for him.

“Here is the most direct evidence yet that Sen. Barbara Buono’s attack on Christie’s apparent presidential ambitions was misguided,” said David Redlawsk, director of the Rutgers-Eagleton Poll and professor of political science at Rutgers University. “Voters think he is going to run, but most didn’t see that as a reason to oppose him, even while many expect him to resign before the end of his term.”

Most voters see no conflict between Christie considering a presidential run and New Jersey’s needs. A plurality (44 percent) views Christie’s stance on issues and extensive nationwide travel as “doing what’s best for New Jersey.” But a third thinks Christie is more focused on running for president. Another 10 percent say Christie has been doing a bit of both, while 13 percent are undecided.

“Many voters agree with Christie’s assertion that, when it comes to governing New Jersey while considering a 2016 presidential run, he can ‘walk and chew gum at the same time,’” said Redlawsk.

As for Christie’s famously confrontational style, voters agree with the governor, by a 2 to 1 margin, that it shows leadership and helps him get things done at home. Only 30 percent say it is disrespectful and hurts his ability to lead in the state.

At the same time, voters are much less sure that the tough-guy attitude will go over well on the national stage. Thirty-six percent think it will, but 46 percent say voters across the country will not like his attitude.  Just 4 percent say it will not make a difference, and another 14 percent are unsure.

Results are from a statewide poll of 804 registered voters with a margin of error of +/- 3.4 percentage points, contacted by live callers on both landlines and cell phones from Oct. 28 – Nov 2.

Tough talk and the electorate

While Christie’s tough talk approach to politics was often questioned by his opponent and the media, most voters think it reflects leadership and helps him get things done. However, Democrats are not so sure. They are slightly more likely to think Christie’s style is disrespectful and hurts his ability to lead, by a 47 percent to 42 percent margin. On the other hand, 76 percent of Republicans stand by the governor’s tough style; only 15 percent believe it hurts his ability to lead.

Independents also like Christie’s approach, and 64 percent say his blunt style works in the Garden State, while 23 percent dissent. Christie supporters overwhelmingly think his attitude helps him govern (78 percent), while 61 percent of Buono voters say the opposite.

But a tough-guy attitude may take the governor only so far, at least in the eyes of many New Jerseyans, 46 percent of who are dubious that the rest of the country will like Christie’s style.

Republicans are the most positive on this score. Just over half think Christie’s approach will go over well, but a third disagrees. Independents are evenly split on the matter, with 40 percent on either side. Democrats do not think Christie’s brashness is well suited for the national stage; 62 percent think the rest of the nation will not appreciate his approach, while 22 percent think others will accept it. Even 36 percent of the governor’s supporters think his attitude will not play out well on the national stage.

Also, beliefs about Christie’s tough-guy approach bring out a gender gap. Women are five points less likely to say Christie’s confrontational style helps him govern and are 10 points less likely to think his style will serve him well on the national stage.

Partisanship prevails on consequences

Majorities of registered voters of every political persuasion expect Christie to run for president, including just over 60 percent of both Christie and Buono voters, and about the same share of those with either favorable or unfavorable views of the governor.

Eighty percent of Republicans and 78 percent of independents reported that a potential Christie run would have no effect on their vote on Election Day. Twelve percent of Republicans and 10 percent of independents saw his presidential ambitions as a reason to vote for Christie. Democrats, however, were more negative: although 73 percent said a possible run did not matter, 20 percent said the talk was making them less likely to choose the governor. Of those who planned to vote for Buono, 28 percent said the potential presidential run made them less likely to support Christie.

“Buono’s message did not resonate with many voters, but among her supporters, some were drawn by the expectation Christie will leave the state,” noted Redlawsk. “But, for the most part, this attack made little difference.”

A majority of Democrats expect Christie to resign to run for president; just 28 percent think he will finish his new term. Forty-three percent of Republicans and 41 percent of independents also expect a resignation at some point, while about one-third of each group says Christie will stay in office. Christie voters give a slight edge to finishing over resigning, 38 percent to 36 percent.

By more than a 2 to 1 margin, voters who think Christie is already running for president say he will leave Trenton. Buono voters also expect him to step down.

Partisanship is clearly evident about whether Christie’s travel and actions in office have been in New Jersey’s best interests. Sixty-three percent of Republicans believe the governor is more focused on the good of the state, while 14 percent see him as more focused on running for president. Independents are more divided; 44 percent think Christie is doing what is best for New Jersey, and 32 percent think he is more focused on holding higher office.

A plurality of Democrats (45 percent) say Christie is preoccupied with presidential aspirations, while 32 percent see him focused primarily on what is best for New Jersey. Sixty-three percent of Christie voters say the governor is doing what’s best for the state, while 70 percent of Buono backers disagree.

1 Comment

Filed under 2013 NJ Election, Chris Christie, Christie NJ Rating, Uncategorized