Polling Results in the Competitive States

President leads in three states that would cement an Electoral College victory

The Electoral Vote application at the New York Times has allocated all but nine of the states to either Barack Obama or Mitt Romney.  In the chart below I have tabulated the number of likely-voter polls conducted in each of those nine states between June 1st, when Romney lined up enough delegates to win the nomination, and October 19th.   Rather than examine the size of the difference between the candidates, I took a simpler approach and just counted up the number of polls in which the President held a lead over Mr. Romney.  I then used a statistical test called “chi-squared” to see whether the President led in “too many” polls for it to just be the result of chance.

Battleground-Polling-Tests

So, to take the case of Nevada, there have been 20 likely-voter polls conducted between June 1st and October 19th, and President Obama has led in 18 of them.  If the race were really tied, each candidate should have led in ten of those twenty polls on average.  Statistics enables us to ask whether the President’s tally of eighteen polls is large enough to reject the notion that the President is tied or behind in Nevada.  The calculated statistic in the table above is 12.80.  The column headed with “p” reports the probability that the race is tied.  In this case there are only three chances in 10,000 the President actually trails in Nevada if he is leading in eighteen out of twenty polls.

Two other states besides Nevada also appear firmly in President Obama’s coalition, Wisconsin and Ohio.  If the President does indeed win these three states, Ohio, Wisconsin and Nevada, and retain his leads in “leaning” states like Michigan and Pennsylvania, he will win precisely 271 Electoral Votes, or one more than he needs to gain re-election.  Replacing Nevada with New Hampshire would leave the President one EV short at 269.

The bottom half of the chart replicates the same analysis using only polls taken since September 1st.  The President is still a significant favorite in all three states based on just the more recent polling as well as all the polls since June 1st.

 

Polling Update

A bunch of new polls today, October 23rd, show more positive results for the President than we have seen recently. Our trend model still detects no evidence that the race has stagnated like it did in 2004, but the President has not rebounded since his drop in the polls after October 4th.  As of now, he is predicted to hold an 0.8% lead on Election Day.

If the Gallup Tracking outlier (the -7 value) is excluded, the predicted lead for the President jumps to 1.1%.

What can we learn from the Kerry/Bush 2004 debate effect?

2004-debate-effect

This graph presents the pre-debate and post-debate trends in polls conducted during the 2004 election.  RCP does not report whether the data represent registered voters or only likely voters for that year, so I have had to include both types of polls in the chart above.  The change in the campaign’s trajectory after the September 30th, 2004, first debate was dramatic.  Mr. Kerry erased nearly nine points from President Bush’s predicted margin in the polls on Election Day.  Kerry also halted the President’s rapid gains in the polls; the daily rate of increase for George W. Bush between July 4th and September 30th was 0.12 percentage points per day, twice the rate scored by President Obama in 2008.

Unfortunately for Senator Kerry he never advanced any further in the  polls after that.  Kerry stemmed Bush’s advance but could not manage to pick up that last remaining two percent that would have brought him even with the President.  Polls after the first debate average Bush+2 with no visible or statistically measurable trend.  According to the 2004 polls, the campaign stagnated after the first debate.

Whether we might expect a similar pattern in 2012 cannot yet be determined.  All we can tell at the moment is that Mr. Romney picked up almost as much ground (+6%) as Kerry did eight years ago (+8%).  There is not yet enough data to tell whether the campaign will stagnate at the current small Romney lead, or whether the President might recover from the blow in the weeks ahead.

Winning the Electoral College while losing the popular vote

Discussions have surfaced again over the possibility of a President being elected without a popular vote majority.  The relationship between the winning candidate’s Electoral Vote margin and his popular vote margin in elections since 1960 looks like this:

We can use the estimated equation, 3.78*Popular_Vote-1.33 to solve for the popular vote proportion that predicts an Electoral College tie of 0.5.  That value is .484, meaning that the winner’s share of the two-party vote could be as low as 48.4% and still result in an Electoral College victory.  (Excluding the 1980 outlier from the analysis only changes this value to 48.6%.)

Nate Silver’s 538 blog gives probabilities of 4.8% for a Romney popular vote victory but a loss in the Electoral College; for Obama, the probability is 2.5%.  At 7.3% this is still a chance event, but one whose chances of occurrence may be increasing.

President Obama’s loses his lead in national likely-voter polls after the Denver debate

Following the same procedures discussed in other postings here, I have regressed the size of the Obama lead in polls of likely voters taken after July 4th against the number of days remaining in the election and a dummy variable for polls taken after the October 4th debate.  The results above show that the President was following nearly the same trajectory in the polls prior to the debate as he did in 2008.  The estimated polling margin for Election Day calculated using pre-debate polls is an Obama victory of 6.0%.  The post-debate shift of -5.4 in Mr. Romney’s favor puts the Election Day margin of victory at just 0.6% in favor of President Obama.

OLS, 2012 Likely Voter Polls after July 4th (N=53)
Dependent variable: Obama lead over Romney

               coefficient   std. error   t-ratio   p-value 
  ----------------------------------------------------------
  const         6.01935      0.933589      6.448    4.39e-08 ***
  DaysBefore   −0.0534490    0.0139228    −3.839    0.0003   ***
  PostDebate   −5.38442      1.00644      −5.350    2.20e-06 ***

Mean dependent var   2.150943   S.D. dependent var   2.884949
Sum squared resid    267.6344   S.E. of regression   2.313588
R-squared            0.381610   Adjusted R-squared   0.356875

The coefficient on DaysBefore indicates that the President would pick up an average of 0.05% as each day of the campaign passes.  That rate of gain is a bit slower than in 2008 when it reached 0.06%.  In 2012, President Obama was on track to gain a full percentage point in the polls every twenty days.  If he continues on that path, he would be marginally ahead of Mr. Romney by 0.6% come Election Day.

 

2004 First Debate Results Suggest Race Could Stagnate

This graph presents the pre-debate and post-debate trends in polls conducted during the 2004 election.  RCP does not report whether the data represent registered voters or only likely voters for that year, so I have had to include both types of polls in the chart above.  The change in the campaign’s trajectory after the September 30th, 2004, first debate was dramatic.  Mr. Kerry erased eight points from President Bush’s lead after that debate, and the President’s rapid gains in the polls that year came to an abrupt halt.  After that debate, Mr. Kerry remained two points behind President Bush but could never close the gap between them.  The campaign seemed to stagnate at this point.

Whether we might expect a similar pattern in 2012 cannot yet be determined.  All we can tell at the moment is that Mr. Romney picked up almost as much ground (+6%) as Kerry did eight years ago (+8%).  There is not yet enough data to tell whether the campaign will stagnate at the current small Romney lead, or whether the President might recover from the blow in the weeks ahead.

First Post-Debate Results

President Obama’s loses his lead in national likely-voter polls after the Denver debate.

The five likely-voter polls taken since October 4th suggest the 2012 Election is now tied.  Extrapolating the President’s trend in likely-voter polls taken between July 4th and the October 4th debate would have put him ahead of Mr. Romney by the comfortable margin of six percentage points.  The five polls taken since the debate suggest that lead has been erased.  Even with so few polls, the estimated swing of 6% in Mr. Romney’s favor is highly significant with a value five times its standard error.  The 95% confidence interval for the shift in the President’s support ranges from -9.7% to -3.6%.

The President’s trend in the polls this campaign hardly deviated from his progress in 2008 before the Denver debate.  He picked up support a bit more quickly in 2008 and reached a predicted margin in the polls on Election Day of 7.3%, nearly identical to his actual margin of victory in the popular vote. The slightly slower pace of growth in the President’s support in 2012 put the President on a trajectory to reach 6.0% by November 6th. The estimated -6.0% post-debate swing eliminates all that advantage.

Follow the Yellow Brick Road

I have been trying to determine whether President Obama’s current standing in the polls is running ahead or behind this time four years ago. To answer this question I’ve collected all the polling reports from RealClearPolitics for the 2008 election and all those for 2012 with fieldwork ending on or before October 7th.  From here on I will refer to the day a poll’s fieldwork ends as its “polling date.”  I then use this date to determine the number of days between the polling date and Election Day.  I can then plot the President’s course in the polls in both years using an identical metric, the number of days remaining before Election Day.

I have chosen to focus on just those polls whose polling dates fell after July 4th. By then both parties have chosen their candidates, the campaigns have begun gearing up for the long haul to November. but no conventions or debates have taken place. I am also using only polls of “likely voters.”  There are significant and complex differences in results between polls of likely and registered voters.  To avoid all those confounding problems I am sticking to just likely voters.

That leaves us with 47 polls in 2012 and 119 polls from 2008 whose polling date fell after the Fourth of July and before Election Day.  If we fit a straight line to all these polls using simply the time left before the election we get the graph above.  The individual polls are represented by blue plus signs for 2008 and red Xs for 2012.  (The red and blue colors have no partisan meaning; they are just designed to enhance visibility.)  The horizontal axis is reversed so that it begins 125 days in advance of the election  and ends on Election Day itself.

In general the graph depicts a fairly rosy view of the President’s chances.  The one disturbing outlier is the Pew poll released today which had Mr. Romney ahead by four points among likely voters.  The President faced a couple of equally bad polling results in 2008, but they occurred about a month earlier than Pew’s post-debate polling date of E-30.   In 2008, the President consistently held the lead in every poll taken over the last thirty days of the campaign.  At no time was he trailing in any poll this close to Election Day,  never mind trailing by four points.

That caveat aside, 2012 has generally followed the same trajectory as 2008, though with the President running perhaps one percentage point behind his 2008 average. This estimate is still statistically weak reaching only the 0.10 level of significance, but I have included it since the number of 2012 polls is still fairly limited.

(There does not appear to be an “interaction” effect where the slope differs significantly in 2012 as well as the intercept.  The best fit uses just a zero/one dummy for the 2012 election year along with the DaysBefore trend.)

The heavy line represents the best estimate for the trend taking all the elections together.  The 2012 trend might run about one percent lower than 2008 as represented by the narrow line.  If the campaign followed that trend, President Obama would win re-election by a margin of 53-47.  Using the graph of the Electoral College “swing ratio” I posted the other day, that margin in the popular vote would translate into just under two-thirds of the Electoral Cellege, or over 350 Electoral Votes.

 

Will the race tighten as the election approaches?

Reading Nate Silver’s 538 blog I see commentators wondering whether the margin in the polls naturally tightens as the election approaches.  For 2008 I can say with some certainty that the answer then was no.  For the current election, the results are more uncertain.

The nitty-gritty details are posted in the Technical Appendix. The headline results appear as those two solid lines that depict how the the competitiveness of each race has changed as they get nearer to Election Day.  For 2008 we have all the polls over the last 120 days of the campaign, while for 2012 only polls conducted between July 7th and and October 2nd are included.

I use the absolute value of the difference in support for the two candidates as the measure of competitiveness.  Races where Obama leads by two and races where Romney leads by two are treated identically.  I have also restricted the analysis to the 120 days prior to the election (beginning in early July) after the nominees have been decided but before the conventions.

Despite this evenhanded approach, there is not much difference between “competitiveness” and how well Barack Obama is doing.  Both John McCain and, so far, Mitt Romney have only held the lead on 12% of their polling days within the last 120.

You can see this at work in the closing stages of the 2008 campaign where the red line depicts Obama’s march to victory. There is no evidence of the 2008 race drawing closer as the election came near.

In 2012, we see a much different story, one of a campaign stuck in neutral for weeks now.  The model predicts that the margin between the 2012 candidates will still be around 3.5% on Election Day as it has been for the past three months.

I’m going to add the 2000 and 2004 elections to this analyais. That will add another open-seat contest and one where an incumbent faces reelection. We might guess that 2000 will look more like 2008, and 2004 more like 2012. We’ll see how well this hypothesis fares in a later installment.