Array
(
    [result_info] => Array
        (
            [total_posts] => 10
            [posts_per_page] => 5
            [total_pages] => 2
            [current_page] => 1
        )

    [posts] => Array
        (
            [0] => Array
                (
                    [id] => 1344
                    [post_author] => Sam Wang
                    [post_date] => November 2, 2012
                    [raw_post_date] => DateTime Object
                        (
                            [date] => 2012-11-02 16:30:00.000000
                            [timezone_type] => 3
                            [timezone] => UTC
                        )

                    [post_title] => Where are undecideds falling?
                    [post_slug] => where-are-undecideds-falling
                    [post_url] => https://election.princeton.edu/articles/where-are-undecideds-falling/
                    [post_content] => 

(original version published on temporary site with comment thread)

In the home stretch to Tuesday, undecided voters are claimed to be an unknown that could increase President Obama’s lead – or turn the race around for Governor Romney. But many pundits (and comedians) don’t understand what it means to be “undecided.” Most undecideds probably have a preference, but are unable or unwilling to state it. Now they are starting to do so.

There is a myth that undecided voters start in the middle, from where they make a decision to support a specific candidate. As my colleague Josh Gold and I have written (The Undecided Voter’s Secret Ballot, New York Times, October 28, 2008), decisions are not like that. We are not always aware of our mental commitments, even if we have already made them. Instead, think of “candidate-preference” and “ability-to-report” as distinct qualities, analogous to how hair color and height are unrelated to one another.

Let’s look at the largest single shock to the 2012 race, Debate #1. In a single day, Romney closed nearly the entire 6-point gap (measured in Popular Vote Meta-Margin) between him and President Obama. What happened to undecided voters?
Here is an average calculated from 98 national polls from our data partners at Pollster.com, unpacked on a day-by-day basis:

The overall picture is of undecideds gradually diminishing by about 1.0% over the month of October. There might be a small hiccup of around 1% around October 4th, the day after the debate. But the key point is that there is no lasting decrease in undecideds that corresponds to the jump in the Obama-Romney margin.

Indeed, there is a very faint tendency for the two-candidate margin to go upward over the last 3 weeks – about 0.5%. Undecideds could be discovering their inner Democrat. Or decided voters could be flipping, which seems possible if one looks at the RAND survey since October 8th:

(Incidentally, the Shifts-Between-Candidates graph is the only graph at RAND I follow. It follows specific voters, so we can see them flow back and forth. I don’t regard RAND’s Election Forecast as any better than other pollsters. It might be worse; the size of their popular vote margin is certainly not looking very persuasive to me at the moment.)

In any event, there’s little evidence to say that undecideds are breaking in a big way. Note that they are also not “independents,” which are a different beast. Most independents have a distinct preference for a candidate, but not a party. Charles Franklin has analyzed undecideds and finds that they break nearly equally.

Which brings me to a major point. We fuss and mock undecided voters – but isn’t the elephant in the room the decided voters? They can vary in enthusiasm, which affects how they show up in likely-voter screens. Or they can flip, as one can see in the RAND data. Think about that. The electorate is so polarized this year. If they go right from Obama to Romney, or vice versa, they evidently don’t stop at “undecided” along the way. These flighty people, not “undecideds,” are the ones that the campaigns should be targeting in the home stretch.

[post_excerpt] => (original version published on temporary site with comment thread) In the home stretch to Tuesday, undecided voters are claimed to be an unknown that could increase President Obama’s lead – or turn the race around for Governor Romney. But many pundits (and comedians) don’t understand what it means to be “undecided.” Most undecideds probably have a […] [topics] => Array ( [2012-election] => 2012 Election [president] => President ) ) [1] => Array ( [id] => 1336 [post_author] => Sam Wang [post_date] => October 29, 2012 [raw_post_date] => DateTime Object ( [date] => 2012-10-29 10:00:00.000000 [timezone_type] => 3 [timezone] => UTC ) [post_title] => Nerds under attack! [post_slug] => nerds-under-attack [post_url] => https://election.princeton.edu/articles/nerds-under-attack/ [post_content] =>


Paul Krugman is calling out National Review Online for their attempted takedown of Nate Silver for biased methods and somehow cooking the books. Krugman writes:

This is, of course, reminiscent of the attack on the Bureau of Labor Statistics — not to mention the attacks on climate science and much more. On the right, apparently, there is no such thing as an objective calculation. Everything must have a political motive.

Now more commentators on the right, including Jay Cost (The Weekly Standard) and Jennifer Rubin (Washington Post), are getting in on the act. Wow, dogpile on the rabbit!

A popular approach to undermining technical knowledge is to throw mud, assert expertise, make picky points, and sow doubts among the less savvy. In this case, what’s the argument? The NRO writer, Josh Jordan, makes this core criticism:

When you weight a poll based on what you think of the pollster and the results and not based on what is actually inside the poll (party sampling, changes in favorability, job approval, etc), it can make for forecasts that mirror what you hope will happen rather than what’s most likely to happen.

Jordan sounds like many partisan polling enthusiasts – on both sides. However, his style of poll-dissection can very easily lead a person astray. The human mind has a large capacity for finding reasons to reject a piece of disagreeable evidence. I’ve written about this in the context of how people form false beliefs in politics (“Your Brain Lies To You,” NYT, June 27, 2008). Polling internals lend themselves very well to such “motivated reasoning.” It is always possible to find something not to like in a poll. This is why I discourage all of you from chewing over single polls.

Silver’s evaluations of pollster reliability are quantitative parameters. However, there isn’t full transparency about how he arrives at them and what he does with them. This leaves him open to attack.

Partly because of this risk, I have stayed with simpler rules such as

Combined with a probabilistic calculation, these rules guided our Meta-Analysis to the exact EV outcome in 2004. It missed by only 1 EV in 2008. Such simple methods are easy to make transparent. You (or Jay Cost, I guess) could download my code in an instant.

I have my own technical beefs with FiveThirtyEight (for example, see here, here, and here). I believe Silver doesn’t extract all the information and tends to add unnecessary factors, which leads to blurry probabilities and poor time resolution. However, his intuitions about the data are excellent and he is very concerned with getting things right. For purposes of popular consumption, he is a fine and honest nerd.

Jordan’s capacity for wishful thinking is apparent when he writes:

While it’s impossible to know how the late deciders will break, the historical trend has been for them to break for the challenger.

I sympathize with this, since I thought the same in 2004, and added a “turnout/undecideds” parameter. For this I received a well-deserved drubbing afterward. In fact, undecideds split about equally, as amply documented by Charles Franklin. I don’t add such parameters any more.

(However, if Jordan wants to implement his idea, he can do so easily by clicking the “With +2% for Romney” link, over in the right sidebar.)

Finally, I will state something obvious. None of this storm of criticism would be happening if “Ro-mentum” (Oct. 23) were real. In fact, Mitt Romney’s fortunes peaked around October 4-9. Since then, the race has moved back toward Obama by about 2.5 points. National polls* give the graph at left. (See update, below.)

And the Popular Vote Meta-Margin, which describes how much state polls would have to swing to generate a tie in Electoral College mechanisms, looks like this:

History of Popular Meta-Margin for Obama

The Meta-Margin may still be catching up with national polls. If so, it has a few tenths of a point to go before it stabilizes. Alternately, something different is happening in swing states. In either case, the overall picture is the same: a narrow Obama lead that is static – or perhaps widening. There is no evidence for Ro-mentum.

>>>

*Analyzed as previously described. To generate the graph above, Gallup/Rasmussen were excluded. I am not at all averse to using them, but they have large house effects and so would need to be analyzed using median-based statistics, which I did not apply above. Anyway, adding them back gives the same relative picture of rapid decline and bounceback, with the same shape, except that the entire graph is slid upward toward Romney by 1.0%.

Median of all national polls, day by day

Update: OK, here is the poll-median graph, including Gallup/Rasmussen, day by day. It is not a weighted median, but it’s close enough (adverse blogging conditions). Note that the drop on October 3rd is due to a Rasmussen poll, which demonstrates their house effect, about two points relative to other pollsters. The conclusion from this plot, as well as the plot above, is that the race has been at a standstill for the last two weeks.

[post_excerpt] => Tweet Paul Krugman is calling out National Review Online for their attempted takedown of Nate Silver for biased methods and somehow cooking the books. Krugman writes: This is, of course, reminiscent of the attack on the Bureau of Labor Statistics — not to mention the attacks on climate science and much more. On the right, […] [topics] => Array ( [2012-election] => 2012 Election [president] => President ) ) [2] => Array ( [id] => 1282 [post_author] => Sam Wang [post_date] => September 26, 2012 [raw_post_date] => DateTime Object ( [date] => 2012-09-26 12:48:00.000000 [timezone_type] => 3 [timezone] => UTC ) [post_title] => Vast liberal conspiracies? [post_slug] => vast-liberal-conspiracies [post_url] => https://election.princeton.edu/articles/vast-liberal-conspiracies/ [post_content] =>

(Welcome, readers of IEEE Spectrum! A small group, but enriched for quantitative people. Do not be fooled by the picture. I usually do not glower.)

The Popular Vote Meta-Margin just hit 5.0%, its highest value of the campaign. I will briefly address the wishful thinking of Romney supporters, despite its faint whiff of hysteria. Some fantasies involve the assertion that party weighting is off. Others suggest that undecided voters will all go in one direction. However, Jay Cost and Unskewed-Man should probably drop the silliness, for the following reasons.

Party self-identification is more fluid than is commonly realized. Not only does voter enthusiasm change on either side, but some people seem to change their self-report. This hypothesis would also require a simultaneous, coordinated effort by all the pollsters (except for brave Scott Rasmussen) to stop doing what they did so well in 2004 and 2008 (see left sidebar). Bottom line: the polls are fine.

The undecided voter break is also highly unlikely, as Charles Franklin reported in 2008. As evidenced by recent polls, undecideds are currently 5.0 +/- 1.7 % (median +/- estimated SEM, n=6). Based on how last-minute undecideds have split in the past, they’ll only be good for a net shift of 0.3 +/- 1.0 % towards Romney.

Finally, there is the question of whether the debates matter. Contrary to what has been said lately, in fact they might, especially the first one. In 2004, Kerry gained about 30 EV, equivalent to about 1.5% of Meta-margin:

Median EV estimator from 2004 race

Let’s say Romney overpowered Obama in debate (1.5%) and captured an unusually high fraction of undecideds (0.3+1.0 = 1.3%). That’s 2.8% total. He would still have to make up an additional 2.2% from somewhere, almost the size of the post-Democratic-convention bounce to date. I do not see that happening.

I encourage you to focus away from the Presidential race. Of more practical interest are questions where the outcome is uncertain: a few Senate races (ND, IN, MT, CT, and MA), and House control. Karl Rove’s Crossroads GPS is probably making adjustments to reflect this polling reality. It’s hard to imagine them pouring much more money into the Presidential race. I have adjusted ActBlue as well, by de-emphasizing WI-Baldwin and VA-Kaine because they are out of the knife-edge zone.

(Update: As pointed out in comments, Senate races in Nevada and Arizona might be entering knife-edge territory. I’m watching that.)

[post_excerpt] => (Welcome, readers of IEEE Spectrum! A small group, but enriched for quantitative people. Do not be fooled by the picture. I usually do not glower.) The Popular Vote Meta-Margin just hit 5.0%, its highest value of the campaign. I will briefly address the wishful thinking of Romney supporters, despite its faint whiff of hysteria. Some fantasies […] [topics] => Array ( [2012-election] => 2012 Election [house] => House [president] => President [senate] => Senate ) ) [3] => Array ( [id] => 318 [post_author] => Sam Wang [post_date] => November 3, 2008 [raw_post_date] => DateTime Object ( [date] => 2008-11-03 14:45:00.000000 [timezone_type] => 3 [timezone] => UTC ) [post_title] => How will the last holdouts break? [post_slug] => how-will-the-last-holdouts-break [post_url] => https://election.princeton.edu/articles/how-will-the-last-holdouts-break/ [post_content] =>

The national two-candidate margin is currently Obama leading by 7.0 +/- 0.9% (median+/-SEM, n=7, 10/31-11/2). There are between 1 and 6% “undecideds.” Can they wreak havoc anywhere?

Today, Charles Franklin at Pollster.com writes about how undecideds have broken since 1948. These data can give us outer bounds on what may happen this year. Bottom line: they don’t matter.

In these data from the National Election Study, the largest break ever occurred in 2000, when undecided voters broke 3:1 against the incumbent party. Overall, the average swing has been a net loss of 7 +/- 21 (mean+/-SD) % of undecided voters for the incumbent party. There’s no statistically significant bias in either direction.

We can put an upper limit on how much effect these unaware souls will have. The most-extreme-case scenario is 6% undeclared (NBC/WSJ) with the same size break as in 2000, for a maximum net swing of 2.6% for either Obama or McCain. Assuming an average of 3.1% undecideds across surveys, the net swing has a 95% confidence interval of -1.4% to +1.4%. These values are fairly close to the “Obama +2%” and “McCain +2%” scenarios that you can see by clicking on the right sidebar.

Using current median margins (second entry on each line), the two-candidate margin is smaller than 2.6% in four states: Indiana (tie), Missouri (tie), North Carolina (Obama +1%), and North Dakota (tie). In all other cases, the win probabilities would not flip (i.e. change between >50% and <50%). In the case of a break toward McCain, what’s left is 278 safe EV for Obama, and a median outcome of approximately Obama 338 EV, McCain 200 EV.

This uncertainty does not affect the overall Median EV Estimate, as I wrote back in August. My topic at the time was the effect of correlated fluctuations in polls, i.e. “what if all the polls are off in the same direction?” My conclusion at the time was that unless we know in advance what the error is (for instance, with the cell-phone effect), there’s not much change to the analysis.

I will probably include this in my final estimate tonight. You’ll still see what the hypothetical effect would be in my bias analysis, where I consider the possible effect of biased polls due to missed cell-phone users, the [inverse] Bradley effect, and so on.

[post_excerpt] => The national two-candidate margin is currently Obama leading by 7.0 +/- 0.9% (median+/-SEM, n=7, 10/31-11/2). There are between 1 and 6% “undecideds.” Can they wreak havoc anywhere? Today, Charles Franklin at Pollster.com writes about how undecideds have broken since 1948. These data can give us outer bounds on what may happen this year. Bottom line: […] [topics] => Array ( [2008-election] => 2008 Election ) ) [4] => Array ( [id] => 307 [post_author] => Sam Wang [post_date] => October 28, 2008 [raw_post_date] => DateTime Object ( [date] => 2008-10-28 11:20:00.000000 [timezone_type] => 3 [timezone] => UTC ) [post_title] => Covert decisionmaking and the Bradley effect [post_slug] => covert-decisionmaking-and-the-bradley-effect [post_url] => https://election.princeton.edu/articles/covert-decisionmaking-and-the-bradley-effect/ [post_content] =>

Today’s New York Times is a target-rich environment for a brain geek. In addition to my own op-ed on the brains of undecided voters, there’s an op-ed by David Brooks on irrational decision-making in financial markets. All decisions, all the time. And for my second geek, there’s the piece on polling analysis (in which I am quoted). There are a few points that didn’t fit in our article.

The undecided v. persuadables. Seemingly paradoxically, the undecided voters are not necessarily the ones that need to be won over at this point. Charles Franklin has data showing that considerable swings in the race can occur rapidly, while the number of undecided voters declines only slowly. Focus groups of undecided voters do not, on average, jump on or off bandwagons quickly. These observations are consistent with the idea that a large share of movement in a race comes from voters switching allegiances, i.e. from McCain to Obama, or Obama to McCain.

I suspect that “undecided” voters are not easily distinguishable from “decided” voters in what moves them. In past elections they have broken unevenly between the candidates. But they may have biases as strong as those of decideds. In other words, what you see in decided respondents is what you get. In general, campaigns should really think of all voters as committed, and work on persuading them.

The Bradley effect. Undecided voters may differ from the decided mainly in their awareness of their own voting tendency. Although I am unaware of data to speak to this point, this could be relevant to the Bradley effect. Poll consumers have speculated that respondents lie about their preference. But what if the respondents are in fact being as truthful as they can, and realize that their preference is different once they are in the voting booth?

I consider this hypothesis to be more likely than the idea that they feel constrained to lie to a stranger on the telephone. Lying makes little sense since respondents can give any reason, or none at all, when expressing a candidate preference.

In other words, respondents may be simply unaware that they have a race-based bias. Such a bias is well-documented. For a popular description, see Malcolm Gladwell’s book Blink.

[post_excerpt] => Today’s New York Times is a target-rich environment for a brain geek. In addition to my own op-ed on the brains of undecided voters, there’s an op-ed by David Brooks on irrational decision-making in financial markets. All decisions, all the time. And for my second geek, there’s the piece on polling analysis (in which I […] [topics] => Array ( [2008-election] => 2008 Election ) ) ) [filter-list] => Array ( [0] => Array ( [slug] => article [label] => Articles ) [1] => Array ( [slug] => forecast [label] => Forecasts ) ) [topic-list] => Array ( [0] => Array ( [term_id] => 21 [name] => 2004 Election [slug] => 2004-election [term_group] => 0 [term_taxonomy_id] => 21 [taxonomy] => topics [description] => [parent] => 0 [count] => 79 [filter] => raw ) [1] => Array ( [term_id] => 22 [name] => 2006 Elections [slug] => 2006-elections [term_group] => 0 [term_taxonomy_id] => 22 [taxonomy] => topics [description] => [parent] => 0 [count] => 10 [filter] => raw ) [2] => Array ( [term_id] => 23 [name] => 2008 Election [slug] => 2008-election [term_group] => 0 [term_taxonomy_id] => 23 [taxonomy] => topics [description] => [parent] => 0 [count] => 130 [filter] => raw ) [3] => Array ( [term_id] => 27 [name] => 2012 Election [slug] => 2012-election [term_group] => 0 [term_taxonomy_id] => 27 [taxonomy] => topics [description] => [parent] => 0 [count] => 189 [filter] => raw ) [4] => Array ( [term_id] => 28 [name] => 2014 Election [slug] => 2014-election [term_group] => 0 [term_taxonomy_id] => 28 [taxonomy] => topics [description] => [parent] => 0 [count] => 98 [filter] => raw ) [5] => Array ( [term_id] => 29 [name] => 2016 Election [slug] => 2016-election [term_group] => 0 [term_taxonomy_id] => 29 [taxonomy] => topics [description] => [parent] => 0 [count] => 201 [filter] => raw ) [6] => Array ( [term_id] => 24 [name] => 2018 Election [slug] => 2018-election [term_group] => 0 [term_taxonomy_id] => 24 [taxonomy] => topics [description] => [parent] => 0 [count] => 52 [filter] => raw ) [7] => Array ( [term_id] => 26 [name] => 2020 Election [slug] => 2020-election [term_group] => 0 [term_taxonomy_id] => 26 [taxonomy] => topics [description] => [parent] => 0 [count] => 105 [filter] => raw ) [8] => Array ( [term_id] => 4 [name] => Congress [slug] => congress [term_group] => 0 [term_taxonomy_id] => 4 [taxonomy] => topics [description] => [parent] => 0 [count] => 2 [filter] => raw ) [9] => Array ( [term_id] => 10 [name] => Democracy [slug] => democracy [term_group] => 0 [term_taxonomy_id] => 10 [taxonomy] => topics [description] => [parent] => 0 [count] => 2 [filter] => raw ) [10] => Array ( [term_id] => 12 [name] => Election [slug] => election [term_group] => 0 [term_taxonomy_id] => 12 [taxonomy] => topics [description] => [parent] => 0 [count] => 1 [filter] => raw ) [11] => Array ( [term_id] => 13 [name] => Governors [slug] => governors [term_group] => 0 [term_taxonomy_id] => 13 [taxonomy] => topics [description] => [parent] => 0 [count] => 17 [filter] => raw ) [12] => Array ( [term_id] => 3 [name] => Health [slug] => health [term_group] => 0 [term_taxonomy_id] => 3 [taxonomy] => topics [description] => [parent] => 0 [count] => 13 [filter] => raw ) [13] => Array ( [term_id] => 16 [name] => House [slug] => house [term_group] => 0 [term_taxonomy_id] => 16 [taxonomy] => topics [description] => [parent] => 0 [count] => 132 [filter] => raw ) [14] => Array ( [term_id] => 14 [name] => Meta-analysis [slug] => meta-analysis [term_group] => 0 [term_taxonomy_id] => 14 [taxonomy] => topics [description] => [parent] => 0 [count] => 37 [filter] => raw ) [15] => Array ( [term_id] => 5 [name] => Moneyball [slug] => moneyball [term_group] => 0 [term_taxonomy_id] => 5 [taxonomy] => topics [description] => [parent] => 0 [count] => 21 [filter] => raw ) [16] => Array ( [term_id] => 11 [name] => Politics [slug] => politics [term_group] => 0 [term_taxonomy_id] => 11 [taxonomy] => topics [description] => [parent] => 0 [count] => 113 [filter] => raw ) [17] => Array ( [term_id] => 6 [name] => President [slug] => president [term_group] => 0 [term_taxonomy_id] => 6 [taxonomy] => topics [description] => [parent] => 0 [count] => 291 [filter] => raw ) [18] => Array ( [term_id] => 15 [name] => Princeton [slug] => princeton [term_group] => 0 [term_taxonomy_id] => 15 [taxonomy] => topics [description] => [parent] => 0 [count] => 50 [filter] => raw ) [19] => Array ( [term_id] => 7 [name] => Redisctricting [slug] => redisctricting [term_group] => 0 [term_taxonomy_id] => 7 [taxonomy] => topics [description] => [parent] => 0 [count] => 1 [filter] => raw ) [20] => Array ( [term_id] => 25 [name] => Redistricting [slug] => redistricting [term_group] => 0 [term_taxonomy_id] => 25 [taxonomy] => topics [description] => [parent] => 0 [count] => 208 [filter] => raw ) [21] => Array ( [term_id] => 17 [name] => Senate [slug] => senate [term_group] => 0 [term_taxonomy_id] => 17 [taxonomy] => topics [description] => [parent] => 0 [count] => 166 [filter] => raw ) [22] => Array ( [term_id] => 8 [name] => Site News [slug] => site-news [term_group] => 0 [term_taxonomy_id] => 8 [taxonomy] => topics [description] => [parent] => 0 [count] => 46 [filter] => raw ) [23] => Array ( [term_id] => 9 [name] => Supreme Court [slug] => supreme-court [term_group] => 0 [term_taxonomy_id] => 9 [taxonomy] => topics [description] => [parent] => 0 [count] => 18 [filter] => raw ) [24] => Array ( [term_id] => 18 [name] => U.S. Institutions [slug] => us-institutions [term_group] => 0 [term_taxonomy_id] => 18 [taxonomy] => topics [description] => [parent] => 0 [count] => 46 [filter] => raw ) [25] => Array ( [term_id] => 19 [name] => Uncategorized [slug] => uncategorized [term_group] => 0 [term_taxonomy_id] => 19 [taxonomy] => topics [description] => [parent] => 0 [count] => 49 [filter] => raw ) [26] => Array ( [term_id] => 20 [name] => United Kingdom [slug] => united-kingdom [term_group] => 0 [term_taxonomy_id] => 20 [taxonomy] => topics [description] => [parent] => 0 [count] => 3 [filter] => raw ) ) )