Princeton Election Consortium

A first draft of electoral history. Since 2004

What would it take for the House to flip?

August 12th, 2016, 8:00am by Sam Wang

As I wrote recently, the Speaker of the House, Paul Ryan (R), has raised the possibility that his party could lose its majority in November. This fear can account for a variety of recent political events, including the entry into the presidential race of a total unknown who just happens to be the policy director of the House Republican caucus.

To take control of the House, Democrats need to gain at least 30 seats over their current 188 seats. This is a tall order, but not impossible. A change of this size has happened in two out of the last five Congressional elections: a 31-seat gain for Democrats in 2006, and a 63-seat gain by Republicans in 2010. In the modern era of polarization, 1994-2014, a change of that size has happened in 3 out of 11 elections. Since 1946, it has happened in 10 out of 35 elections.

What would it take for the House to change hands? [Read more →]

→ 45 CommentsTags: 2016 Election · House

Politics & Polls #7: The Coattail Effect, Gerrymandering & Third-Party Candidates

August 11th, 2016, 2:00pm by Sam Wang

On Politics & Polls (iTunesSoundCloud, PodOmatic): Donald Trump has attempted to ease tensions within his party by endorsing House Speaker Paul Ryan, Senator John McCain and Senator Susan Collins. But do Republican candidates actually want the endorsement? Could Trump have inverse coattails, and drag other candidates down? Listen to Episode 7!

Correction: I referred to the Republican Senator from New Hampshire as Susan Collins. I meant to say Kelly Ayotte. Collins represents Maine!

→ 8 CommentsTags: 2016 Election

Can Third-Party Candidates Help Save The Republican Downticket?

August 10th, 2016, 8:36pm by Sam Wang

On Monday, the Princeton Election Consortium got its 10 millionth view since it became a WordPress site in 2008. Traffic in July 2016 was over 50 times larger than July 2012. Thank you, both old and new readers!


The Presidential cake is baking. Hillary Clinton’s lead over Donald Trump has increased in national polls to around 7%. Since January, the Clinton-over-Trump national margin has been 4.7 ± 2.3 % (average ± SD). The corresponding SD for January-August 2012, a re-election year, is 1.4%. Before that, Wlezien and Erikson’s data on Presidential campaigns from 1952 to 2008 show higher values for SD in all years, as high as 11.0% (1964). So 2016 is a very stable year so far. As much as 2012, it appears that voter attitudes are extremely well set. This fits with the fact that even after making increasingly incendiary statements, Trump’s support is still around 40%.

Clinton’s advantage in the Electoral College is smaller, as evidenced by a Meta-Margin of 3.7%, which is defined as how much current state poll margins would have to move in order to create an electoral tie. 3.7% may not sound like much, but Clinton’s Meta-Margin has not gone outside the 2.5%-4.5% range since May, when our calculation for 2016 began. All in all, measured in terms of public opinion, 2016 is a contender for the most stable Presidential race since World War II. Of course, we still have nearly three months for movement to occur.

A focus on downticket races is in the air. [Read more →]

→ 24 CommentsTags: 2016 Election · House · President · Senate

Why follow polls?

August 5th, 2016, 10:00am by Sam Wang

Goal Thermometer - PECBefore the 2016 campaign season, I had reservations about re-starting up this site’s polling analysis. However, there was one big reason in favor of doing it. It has to do with your readership of the site – and how you can best influence the outcome.

The biggest reason not to re-start the site was the market for statistical politics, which looked saturated. In 2004, this strange hobby was made newly possible by an abundance of polls. Now there are many sites, most notably the NYT’s The Upshot, ESPN’s FiveThirtyEight,, and HuffPollster (which provides our polling data). The Princeton Election Consortium may now seem redundant.

Yet that is not the case. [

→ 128 CommentsTags: 2016 Election · House · President · Senate

Politics & Polls #6: Data And The Political Media (also, upvote us!!)

August 4th, 2016, 1:50pm by Sam Wang

This week in Politics and Polls, Julian Zelizer and I discuss how data should inform coverage – but sometimes doesn’t. And also the role of storytelling!

Also, we try our spiffy new sound setup, which involved sticking my head into a canvas box. Thankfully, there is no video of this.

Like what we’re doing? Please upvote us in iTunes! (Open it in iTunes, then click Ratings and Reviews.) Enough votes and we become New and Noteworthy! [WWS] [iTunes] [SoundCloud]

→ 6 CommentsTags: 2016 Election

Why is the PEC polls-only forecast so stable?

August 3rd, 2016, 4:30pm by Sam Wang

Brad DeLong explains some mechanics of the ESPN/FiveThirtyEight polls-only forecast:

  • Takes recent past polls to estimate a current state of the race…
  • Estimates an uncertainty of our knowledge of the current state…
  • Projects that that current state of the race will drift away from its current state in some Brownian-motion like process…
  • Calculates the chance that that process–starting from the estimated present–would produce 270 electoral votes for one candidate or the other.

This is also a near-perfect description of the “random-drift” calculation in the second line of the banner above. My main quibble here is that he says it’s not a forecast. Actually, it is. It assumes random drift from now to November 8th, Election Day. That is a forecast!

The issue is that it is a forecast based on current polls only. In my view, such a forecast has the conceptual flaw of overweighting the freshest data. It amounts to reporting a snapshot of today’s conditions, blurred out a bit. To make an analogy, this is like forecasting the temperature a week from now using a single temperature reading. You should not predict weather using a single reading taken on a hot day or in the middle of the night. Jon Stewart had a routine like this once to make fun of deniers who say that because it is snowing today, climate change must be a hoax. (at night: “The sun will never rise again!”)

People who get into a tizzy over post-convention bounces are a little like that. We are now in the middle of a post-DNC convention bounce for Hillary Clinton. What if her numbers come back down? In this sense, DeLong is correct – extrapolating from current conditions is not really a forecast. Though in that case, the “Brownian motion” (note: polling changes do not actually act like Brownian motion, so it would be better to just say random drift) plays no meaningful role, and should be dropped.

To generate a long-term prediction (their polls-plus forecast), FiveThirtyEight establishes a prior based on other factors such as the economy. (No, dear reader, I do not want you to tell me what those factors are!) This method can work – Drew Linzer at Votamatic and Benjamin Lauderdale have used it to make a long-range forecast, and have analyzed their model rigorously.

In my view, FiveThirtyEight’s polls-only approach does not take full advantage of this year’s polling data. It is possible to generate a long-term prediction using polls only. Here at PEC, that’s what we do to create a prior expectation for where polls will drift to. Think of it as “polls-plus-more-polls.” Here is how we do it. [Read more →]

→ 47 CommentsTags: 2016 Election · President


August 2nd, 2016, 1:00pm by Sam Wang

Three months before the general election, why would a candidate claim that the election will be rigged? Either to foment unrest afterward, or to claim that he was robbed. In either case, the remark suggests that he expects to lose.

The high accuracy of poll aggregation acts as a safeguard against blatant fraud. In the last few weeks of the Presidential campaigns of 2004, 2008, and 2012, poll aggregates (including those served up here) gave the correct winner in all states where the poll median was more than 1% for either candidate. Polls for individual Senate races and national Congressional popular vote also do very well in Presidential years – though less so in off-years like 2010 and 2014.

At the Princeton Election Consortium, we use state polls only, which means that it takes some time to catch up with changes in opinion. The aggregate above includes only three post-convention polls: one each in Pennsylvania, Nevada, and Georgia. At the noon update, Georgia should go to a median of Trump +1%. In 2012, Mitt Romney beat President Obama by 8 percentage points. Previously I pointed out that Trump runs about 9 percentage points worse than Romney in Republican-leaning states. The persistence of that gap suggests that the Republican convention, and Trump’s continuing focus on the late Captain Khan’s parents, have failed to bring his party’s voters home.

→ 47 CommentsTags: 2016 Election · President

Post-Democratic convention bounce: 7 points for Clinton

August 1st, 2016, 8:14am by Sam Wang

So far, four five six pollsters have released national surveys using samples taken after the end of the Democratic National Convention, and have data from the post-RNC period. The median swing is a 7.0 ± 1.1 2.0 2.1 % (± estimated SEM) move toward Hillary Clinton.

Here are this year’s bounces, added to the previous graph.

Individual polls: [

→ 69 CommentsTags: 2016 Election · President

Politics & Polls #5: Courting the Working Class

July 30th, 2016, 12:00pm by Sam Wang

This week, Democrats invoked patriotism, diversity, and strength at their convention. They also made an explicit play for people who typically vote Republican. Among these are white working-class voters. In this week’s Politics and Polls, Julian Zelizer and I talk about what these voters might want. [WWS] [iTunes] [SoundCloud]

→ 22 CommentsTags: 2016 Election · President

A Local Experiment In Crowd Wisdom

July 29th, 2016, 4:00pm by Sam Wang

(updated 10:10pm)
I have finally converted my statistical politics hobby into material gain.

Here at Princeton, the Center for the Study of Democratic Politics has announced the winners of its quadrennial election-prediction contest, a game that is open to CSDP and the Princeton community more broadly. Late in 2015, they put out an open call. Entrants were asked to predict the nominees for the two major parties. The deadline was January 2, 2016. Vice-presidential nominations were used as tie-breakers. I won! The prize: a hooded sweatshirt with the CSDP logo.

Near the end of December, I submitted my entry to the organizer, Michele Epstein along with a note: “If I’m wrong, and there is a good chance I will be, consider this merely a snapshot of what the data pointed toward as of January 1, 2016.” As it turns out, 2016 has been an excellent test of what a polling data-centered approach can accomplish.

Taken as a whole, the contest entries describe the collective opinion of knowledgeable experts – and reveals what a strange year this has been. [Read more →]

→ 39 CommentsTags: 2016 Election · President