Election Polling 101

Yesterday’s Quinnipiac University Poll of Florida voters was criticized by Democratic pollster David Beattie for the partisan composition of its sample.  He writes “In short the results raise some concerns because more Republicans than Democrats are interviewed, which is not going to happen on Election Day in Florida. They model that 32% of the turnout will be Republican, 29% Democrat and 32% Independent in the 2012 election, this is NOTHING like what the reality will be.  There will be a net 2 percentage point or more Democratic registration advantage on election day.”  

One problem with Mr. Beattie’s analysis is that he assumes that we know prior to the election what the exact partisan composition of the electorate will be on Election Day.  One only has to look at past Florida exit polls to see that the party distribution of the electorate can change from election to election.  In 2008, there were more Democratic voters than Republicans by a 37-34 percent margin.  However, in 2004, Republican voters outnumbered Democrats by 4 points 41-37 percent. 

We are currently using the same methodology that we used when we accurately predicted the Florida presidential results in 2004 and 2008.  Respected New York Times polling analyst Nate Silver found that we were the most accurate poll in predicting the 2010 elections.

It should also be noted that Mr. Beattie is comparing apples to oranges.  He cites our party identification numbers and then compares them to party registration figures.  Major independent pollsters and political scientists have long recognized that party identification is a better measure of political attitudes than party registration.  Some people register with one party and even though their political attitudes have changed, they don’t bother to change their party registration.  Party identification measures whether people generally consider themselves a Democrat, a Republican, or an Independent, regardless of party registration.    It is important to understand that all of the major polls – Gallup, Pew, Quinnipiac, CBS, ABC, and NBC- do not weight their data by party.  Those polls, as does Quinnipiac, weight their data for things such as gender and age to match the Census data because these things don’t change.

There is always going to be some small fluctuation in party identification in polls.  It might be due to sampling error or short term events.    These small movements are generally not worth mentioning.   If however, there is a big change that holds up over time that would be a significant finding.  Such a phenomena is very rare in public opinion.  It is called a “party realignment” and happens when there is a major event, like the Depression in the 1930’s, which causes a substantial number of people to permanently change their party identification.

In our recent poll, we found that in party identification Republicans outnumbered Democrats 32-29 percent, which is not a statistically significant difference.    In our three prior polls the Republican percentage was between 29-31 percent and the Democratic percentage was between 32 and 33 percent. This sample is a little more Republican than those earlier polls.  But the difference is so small that it is not meaningful.  Even with the slightly more Republican sample, the results of the matchup between Obama and Romney were the same for the last 3 polls. In each survey, Romney was ahead of Obama by a statistically insignificant 3 point margin.

All of these Quinnipiac polls have shown that the party distribution in Florida is about evenly split.  Given our poll numbers it is certainly possible that there will be a small Democratic advantage on Election Day or there could be a small Republican advantage, or it could be a tie.

One early indicator of a potential problem for President Obama is the enthusiasm gap in Florida.  As we’ve seen in other states, Democrats are less enthusiastic about voting than Republicans.  So even though the party breakdown is roughly evenly split, Republicans could outnumber Democrats on Election Day if they are more excited about voting in November.

It should also be pointed out that we are about 10 months from the election.   This was a poll of registered voters.  It is too early to be talking about likely voters.  We don’t even know who the Republican candidate for President will be, although it is obviously looking good for Romney.   No pollster can tell you what is going to happen in the election in November in January, not even Mr. Beattie.  A poll is a snapshot in time.  We will have a better idea of what will happen in the election come September and October, with the most accurate polling right before the election.

Finally, I would suggest that whenever a partisan pollster criticizes the methodology of an independent pollster, it be taken with a big grain of salt.  The goals of a campaign pollster are very different from an independent pollster.  The job of a campaign pollster is to help their candidate get elected.   As an independent poll, Quinnipiac’s goal is to provide accurate polling data to the public.  Unlike campaign pollsters, Quinnipiac is completely transparent about its methods and poll results.    I would suggest that partisan pollsters also be completely transparent and disclose to the news media their own methodology and poll results.  If campaign pollsters are going to go on the record criticizing other polls, shouldn’t they be expected to go on the record explaining how they do their own polls? We welcome any members of the news media to visit our Polling Institute and see how we do our surveys.  We wonder whether Mr. Beattie or other critics, partisan or independent, are willing to make the same offer.

Members of the foreign press visited the Quinnipiac University Polling Institute on Dec. 4. Doug Schwartz, center, director of the Quinnipiac Poll, speaks with his international guests.

The Quinnipiac Polling Institute hosted 14 foreign journalists who are based in the United States.

A wide range of political questions were fielded by Poll Director Doug Schwartz who explained the history and methodology of polling at Quinnipiac.

The group is touring the Northeast getting a read on the issues the races and the people who track them.

The group observed a poll survey then had dinner at the University Club.

Also taking part and sharing their thoughts were Mickey Carroll, April Radocchio and Tim Malloy.

The journalists in attendance were Roel Verrycken from De Tijad in Belgium,Yongxing Zhang from Xinhua News Agency, Jerome Marin from la Tribune in France, Pierre De Gasquet from Les Echos in France, Dirk Hautkapp from Westdeutsche Alligemeine in Germany, Toshiya Kobayashi from Akhata in Japan, Alf Ask from Aftenposen in Norway, Kathleen Gomes from Publico in Portugal, Alena Taranova from Radio Express the Sloval Republic, Edward Zitnik from RTV Slovenia, Elisabeth Frerot Soedergren from TV4,Sweden, Jacek Machula from TV4 Sweden, Beat Soltermann from Swiss National public Radio, Stephene Bussard from Le Temps in Switzerland.

You may view photos of the event here.

The partisan makeup of the electorate has always been one of the most  contentious issues in election polling.   There are those who believe that there is a fixed number of Democrats and Republicans in the nation.  Quinnipiac, however, believes  that while party identification is relatively stable,  it can change over time and  fluctuate temporarily in reaction to events in the news.   For example, if one of the parties is going through a particularly difficult time voters can temporarily feel “less Republican” or “less Democrat” than usual.

The exit polls have shown how the partisan composition of the electorate can change over time.   In the 2008 election, Democratic voters outnumbered Republican voters by a 7 point margin, 39-32 percent. This was an increase of 7 points in the partisan gap  from 2004 when Democrats and Republicans were tied 37-37 percent.

In Quinnipiac’s November 2  national poll,  Democrats outnumbered Republicans by 13 points, 35-22 percent.  This Democratic advantage was 10 points higher than the 3 point gap in our October 4 poll, when Democrats led Republicans 31-28 percent.  The New York Times/CBS News Poll also showed the Democrats with  a double digit advantage of 10 points over Republicans, 35-25 percent in their most recent poll. This was a 3 point increase in the Democratic advantage from their previous poll which had the gap at 33-26 percent.

Whether or not this increase in the Democratic advantage remains for any length of time will depend on future events. The increased Democratic party identification  advantage in Quinnipiac’s November 2 poll  may have occurred because the poll was taken during a week when the stock market had one of its best weeks ever, GDP numbers made it clear that a double dip recession was not occurring, and the President was on the winning side in Libya and announced that American troops would be out of Iraq by year’s end.

Like most other major national polls, we don’t weight our data toward party identification.  The only party identification target that one could possibly use would be the exit polls but the party identification in exit polls moves as we saw in 2008. 

Nobody knows ahead of time what the actual party distribution will be in the 2012 election.  If a pollster weights by party identification, they are in essence making an educated guess about what they think will happen.  This undermines the scientific nature of polling.  The pollster is basically substituting their own judgment for what the voters are saying in the poll.  If people are feeling less Republican than usual and  a pollster tries to weight up Republicans to compensate for that, the pollster is in essentially hiding something that is happening in the electorate because they think they know the voters better than the voters know themselves.

The party identification figures, along with other demographics,  are released on our website with every poll.

The Enthusiasm Gap

The Democrats should be concerned about the “enthusiasm gap” in the upcoming Presidential election.  In the latest Quinnipiac University Poll in Florida, a key swing state, we found that Republicans are much more enthusiastic than Democrats about voting in the 2012 presidential election.  While 63 percent of Republicans say they are more enthusiastic about voting in this Presidential election than in past presidential elections, only 28 percent of Democrats say they are more enthusiastic.  This is a very large gap but with about 13 months to go there is still plenty of time for President Obama to fire up his base.

This enthusiasm poll question is one measure that pollsters use to try to gauge what voter turnout will look like. The partisan makeup of the electorate can have a huge impact on who wins the election. 

In 2004, Republican George W. Bush won Florida by 5 points over Democrat John Kerry, 52-47 percent.  Similarly, Republican voters outnumbered Democratic voters by 4 points, 41-37 percent, according to exit polls. 

In 2008,  Democrat Barack Obama  took Florida.  He defeated Republican John McCain by 3 points,  51-48 percent.  Just as there was an 8 point swing in the election (a 5 point Republican victory in ’04 to a 3 point Democratic victory in ’08), the partisan makeup of the electorate also shifted Democrat.  There was a 7 point swing in partisanship as Democrats went from a 4 point deficit in ’04 to outnumbering Republicans by 3 points, 37-34 percent in ’08.

It is too early for Quinnipiac to start looking at “likely voters”.  For now, we are reporting our poll results among all registered voters.  In the latest poll, self identified Democrats outnumber Republicans by 32-29 percent. The question is whether the Republicans big enthusiasm advantage will trump the Democrats slight advantage in party identification.

In 2008, President Obama won the key swing states of Florida, Pennsylvania and Ohio because he won the independent vote and drove up Democratic turnout.  In 2010, independents swung to the Republican side by voting for Republican governors and senators in all three states.  Also, the Democratic voter turnout surge in 2008, declined in 2010. What will happen in 2012?  Which way will Independents go in these critical battlegrounds?  Will there be a Democratic surge in 2012 as big as 2008?  The answer to these questions will tell us whether President Obama will be reelected or not.

According to the most recent Quinnipiac University polls, more Independents disapprove than approve of President Obama’s job performance in all three of these key swing states. In Pennsylvania 56 percent disapprove, while only 40 percent approve; in Florida 52 percent disapprove, compared to 42 percent who approve; in Ohio 54 percent disapprove, while 42 percent approve.    While these are not good numbers for the President, he still has 14 months to turn them around.

It’s also important to remember that elections are about choices.  While the President is not beloved among independents, neither are the leading GOP Presidential candidates. In Quinnipiac’s most recent national poll,  Rick Perry’s net favorability score among independents was a “negative 3” (19 percent favorable/22 percent unfavorable), with 55 percent saying they haven’t heard enough about him.  The fact that most independents don’t have an opinion of Perry means that he has a lot of room for growth  in either a positive or negative direction.  Romney is better known and better liked than Perry among Independents nationwide but also has a large “Don’t Know”.  Romney has a net favorability score of “plus 15” (39 percent favorable/24 percent unfavorable) with 35 percent who haven’t heard enough.

Unlike the Republican candidates,  almost everyone has an opinion about the President.  Independent voters are split on President Obama (45 percent favorable/48 percent unfavorable)

Both President Obama and the eventual Republican nominee will be trying to do the same things.  They both want to excite their bases to drive up their party’s turnout but at the same time they want to appeal to the center to win independent voters.  The advantage President Obama has is that he can start appealing to independent voters now because he doesn’t have to worry about campaigning for the nomination.  The Republican nominee must appeal to the “conservative” base now to win the nomination and then try to move back to the center.

A nasty fight for the nomination could also weaken the eventual Republican nominee for the general election.

Back in April Quinnipiac asked New Jersey voters to describe Governor Christie in a word.  We found that the word “bully” was the most frequent response given about Christie.  Today, we asked the same question of New York state voters about their governor, Andrew Cuomo.  The top response:  “good”. 

To illustrate registered voters impression of Andrew Cuomo responses were entered into Wordle, a website that generates a “word cloud” from text provided by the user. The cloud gives greater prominence to words that appear more frequently in the source text.

Respondents were asked:

“Thinking about New York Governor Andrew Cuomo… What one word best describes your impression of Andrew Cuomo?”

NOTE: Based on registered voters: Figures show the number of respondents who offered each response: these numbers are not percentages.  This table reports only those words that were mentioned at least six times.

Good                        76    Surprise                    15
Competent                   51    Fair                        14
Trying                      51    Hardworking                 14
Effective                   39    Intelligent                 14
Honest                      38    Energetic                   13
Okay                        35    Tenacious                   13
Politician                  32    Adequate                    12
Tough                       24    Diligent                    10
Arrogant                    23    Dissatisfied                10
Leader                      23    Integrity                    9
Determined                  22    Committed                    8
Efficient                   21    Opportunistic                8
Ambitious                   20    Smart                        8
Liberal                     20    Consistent                   7
Strong                      20    Progressive                  7
Dedicated                   18    Action                       6
Aggressive                  17    Alright                      6
Satisfactory                17    Bully                        6
Decisive                    16    Capable                      6
Excellent                   16    Courage                      6
Forceful                    16    Focused                      6
Confident                   15    Go-Getter                    6
Disappointed                15    Phony                        6

The Quinnipiac University Poll has recently found that while Republican governors have been experiencing big gender gaps in their job approval ratings, Democratic governors and President Obama are seeing relatively small gender gaps. Why?

Public opinion polls have long shown that men are more likely than women to approve of the job Republican office holders are doing and conversely women are more likely than men to like the job Democrats do in office. Typically, the gap between the percentage of men and women approving of an elected official is relatively small (in single digits), and can usually be explained away by differences in party identification. Women are much more likely than men to consider themselves Democrats. For example, in Quinnipiac’s July national poll, 41 percent of women considered themselves Democrats compared to only 28 percent of men. Consequently, women are more supportive than men of Democratic officeholders.

While President Obama has a 7 point gender gap, with 50 percent of women approving of his job performance compared to 43 percent of men, this gap almost vanishes when we factor in party identification. For example, there is almost no difference between Democratic women and Democratic men – they both universally approve of the President. Similarly, there isn’t much of a difference between Republican men and Republican women – almost all of them disapprove of the President. There was a small, statistically insignificant difference in President Obama’s approval rating among Independent men and Independent women.

Like President Obama, the two Democratic governors that Quinnipiac regularly polls on – New York’s Andrew Cuomo and Connecticut’s Dannel Malloy also have small gender gaps. Malloy’s job approval gender gap is just 2 points, with 37 percent of men approving and 39 percent of women approving. Cuomo’s gender gap is also small; it’s just 5 points and actually goes in the opposite direction than what we would expect given historical patterns. Women are slightly less supportive of Cuomo, although both men (67 percent) and women (62 percent) overwhelmingly approve of the job he is doing.

Unlike the Democratic officeholders, all of the Republican governors have double digit gender gaps that can’t be explained simply by differences in party identification. The gender gap for Pennsylvania Governor Tom Corbett is 18 points (48 percent of men approve vs. 30 percent of women); for New Jersey governor Chris Christie it is 17 points (53 percent of men approve vs. 36 percent of women); for Florida governor Rick Scott it is 11 points (35 percent of men approve vs. 24 percent of women); for Ohio governor John Kasich the gender gap is 11 points (44 percent of men approve vs. 33 percent of women); and for Virginia Governor Bob McDonnell it is 11 points (61 percent of men approve vs. 50 percent of women).

One thing that all these Republicans have in common is their approach to budget deficits. They all have relied on spending cuts, rather than tax increases. Just as there is a big gender gap for their overall job approval ratings, there is also a big gender gap for the way they’ve handled their budgets. Women are much less likely than men to approve of the way these governors are handling their state’s budgets.

These Republican governors have either tepid or low overall job approval ratings, except for Governor McDonnell who is at 55 percent. They need to do better among women to lift their overall scores.

All of these governors have plenty of time to improve their standing among women. They aren’t up for reelection until 2013 or 2014. However, one might wonder whether there will be any spillover effect in the 2012 elections. Will women, upset over the budget cuts proposed by Republican governors, show up in unusually large numbers or vote in unusually large numbers for Democrats? Democrats are certainly hoping that will be the case.