PointofLaw.com
 Subscribe Subscribe   Find us on Twitter Follow POL on Twitter  
   
 
   

 

 

My election prediction

| 62 Comments


72 hours from election day! Who's going to win?

Nate Silver says that "For Romney to Win, State Polls Must Be Statistically Biased." I think this is correct. Silver's model has Obama at a 83.7% favorite, and that is likely to turn into a 90-95% favorite by election day the way his model works. Silver's argument for Obama being a favorite: he's ahead in Ohio, and if he's ahead in Ohio, he's substantially more likely to reach 270 electoral votes. This is also correct.

That said, I do believe that there is a substantial chance that the state polls are a biased prediction of where election day results turn out. Because the 538 (and Princeton) model rely so heavily on the assumption that the state polls are unbiased, if the state polls are biased, then the 538 and Princeton models will also be biased. (Note that "bias" is a neutral statistical term; it does not mean that the pollsters are conspiring to favor Obama, just that their likely-voter model is producing results that overestimate his standing.)

1) Polls have historically had a bias against Republicans relative to how election day turns out. Silver disputes this; I've explained why his analysis on that question might be wrong before. On the other hand, the 2012 polls may have fixed this problem.

2) Polls have historically had a bias overestimating the performance of sitting presidential incumbents. There's certainly a counterargument against this. On the other hand, perhaps Kerry's failure of a 2004 bounce reflects the offsetting bias against Republicans. On the other other other hand, perhaps the 1980 and 1996 collapses by Democratic incumbents from their polling positions reflect the distortion caused by a relatively strong third-party candidate, plus Clinton's 1996 decline might simply reflect a reversion to the mean. But if undecideds break 75-25 against Obama, that's a 1- or 2- point shift relative to the polls.

3) The state poll toplines, which is all that the 538 and Princeton models use in their calculations, show enormous advantages for Democratic turnout in state races, advantages greater than even Obama realized in 2008. I think that this is very likely an error that, if corrected for, will show material change. There are many many reasons to think that the Democratic advantage of 2008 will be weaker or nonexistent in 2012, when Obama won 53-46 nationally (and Ohio by 4.6 points).

(a) Obama leads Obama voters only 84-13. Now, some 2008 McCain voters have died or won't vote for Romney; some 2012 Obama voters will be new voters; some 2012 Obama voters will be 2008 McCain voters; that 84-13 may well turn into 90-10 on election day. But it's hard to see where the improvement in D turnout is going to come from when so many Obama voters are disaffected.

(b) Independents overwhelmingly favor Romney, by 20 points in some polls. It's implausible that independents have swung so wildly, yet Democrats are more enthused and more likely to turn out than Republicans. Some, such as Josh Marshall, have posited that the makeup of independents has been changed because disaffected Perot/Tea-Party/Paulite conservatives have left the Republican party. Dan McLaughlin's counter-argument strikes me as stronger. Even under Josh Marshall's calculation of voter ID, D+7 among "all adults" should not translate into a 2012 that is better than 2008 for Democrats: Republicans are more likely to vote; Republicans are more likely to vote for Romney than Democrats for Obama (especially given the shift in Catholic opinion since 2008); and Marshall necessarily concedes that the independent vote will be less pro-Obama than in 2008. And I don't believe Josh Marshall's claim that the current makeup of all adults is D+7.

(c) McCain intentionally hamstrung his campaign in 2008 with unilaterally disarming self-abnegating campaign-finance limitations, next to no on-the-ground spending for get-out-the-vote efforts ("GOTV"), and wasting money on such tactics as a national ad congratulating Obama on winning the nomination. This undoubtedly made McCain feel better about himself, but meant that the GOP had a historically unprecedented disadvantage in GOTV. Even with the supposed 2012 Obama advantage in field offices (which neglects the fact that GOP turnout efforts are largely volunteer), Romney is doing comparatively better in 2012 than McCain did in 2008, which is the relevant comparison for whether the state polls are accurately predicting that Democratic turnout advantage will be higher in 2012 than in 2008.

(d) Some polls show Republicans more excited about this election than Democrats; others show the excitement level about even. But even taking the latter polls as an accurate metric, the relevant comparison is again versus 2008, where the Democrats had a huge excitement advantage. If the Democrats are going to outperform 2008, which they have to achieve the slim lead they achieved in 2008 given the shift in independents, they need to be more relatively excited than in 2008. They're not.

(e) The Obama campaign is assuming turnout will be 72% white; polls are assuming turnout will be 74% white. But there's a strong argument that turnout will be more like 75% white. Whites favor Romney by about 61-39; non-whites Obama about 80-20. This would be a shift of 0.8% relative to the polls.

(f) Reports about early voting are mixed, with National Review and Battleground Watch reporting relative advantages for Romney versus 2008, and Democrats claiming otherwise. This election will likely have a record percentage of early voters, and it's hard to say to what extent that reflects cannibalization or improved turnout. We'll disregard this as ambiguous.

The Princeton model takes the state polls at face value; Nate Silver has made a decision to have the 538 model disregard each of these factors and assume that the polls are as likely to be biased for Obama as for Romney. Thus, there is a substantial chance that the 538 and Princeton models are reflecting a house effect. GIGO: garbage in, garbage out. McLaughlin's post on the subject is a must-read; so is Kevin Holtsberry on Ohio; Michael Barone makes the aggressive case for Romney.

All that said, if I had a gun to my head the morning of November 3, I would rate Obama a slight favorite, with about 60% to 40% chances, a movement from the tossup I suggested on Twitter. (I'm not going to succumb to the fake precision of Nate Silver's model and suggest I can forecast with three significant digits.) Why?

(1) Each of the effects I mention above are overlapping to some extent. You can't add them up to say that the polls are biased by seven points. And Obama has some lead in the state polls, plus has more favorable electoral-college scenarios than Romney in a close election: the so-called "firewall." Thus, I could be right about three, four, or even all seven of those points about state-poll bias, and it still might not add up to enough of an effect to put Romney over the top.

(2) I have to recognize that there's some chance that I'm suffering from confirmation bias, looking only at the data that favors my preferred result, and that I'm wrong. Nate Silver is spending a lot more time thinking about the election and polls than I am, and it's more likely that he's seeing things that I'm not than I'm seeing things he's not; maybe if I were to try to duplicate his efforts, I would end up with a model that was closer to his than my current thoughts are. I thus have to discount my thinking that Romney is doing better than Silver thinks. I think partisans on both sides have been silly about Silver, with some Democrats treating him as infallible to the point that a challenge of Silver is a challenge of scientific truth, and some on the right being simply innumerate in their attacks on Silver. Still, occasionally Silver lets the mask slip and makes a statement that's pretty close to mine about the unpredictability of the election and the chances his model is wrong.

(3) There are pollsters who do take the effects I mention into consideration somewhat; in particular, Rasmussen is very aggressive in doing so. But even Rasmussen shows an election that is close to a tossup, with Obama having more "outs" in the electoral college than Romney. And I think there is a reasonable argument that Rasmussen is too aggressive in its likely-voter model.

(4) A certain number of intangible factors favor Obama. Hurricane Sandy changed news coverage dramatically in the last week of the election: a disaster brings Americans together, lets a president seem presidential, and, in this particular case, provides a conscious or subconscious reminder of Bush administration incompetence. Moreover, every minute spent on dramatic Sandy stories is a minute the news isn't spending on Obama administration failures in the economy, in energy policy, in deficit reduction, in Libya. It's hard to think that the double-standard of "fact-checking" that more harshly (and often incorrectly) scrutinizes Romney claims than Obama claims doesn't have some effect on undecided voters.

(5) What am I supposed to think about Romney throwing $12 million into Pennsylvania at the last minute? Pennsylvania has consistently voted more Democratic than the rest of the nation. It's hard to imagine a scenario where Romney wins Pennsylvania but doesn't win Ohio and/or 270 non-Pennsylvania electoral college votes. The Pennsylvania U.S. Senate race doesn't seem particularly close. There are four possibilities: (a) the Romney campaign sees a real "undertow" election, and legitimately thinks it can get meaningful coattails in Pennsylvania; (b) the Romney campaign is panicking, and throwing a Hail-Mary in Pennsylvania because it doesn't think it can get to 270 without it; (c) the Romney campaign has more money than it can spend efficiently, and is throwing money at Pennsylvania as an expensive feint to divert Obama campaign resources; (d) the Romney campaign doesn't know what it is doing or inefficiently hoarded money. If one assigns Bayesian probabilities to each of those scenarios, that doesn't look good for Romney.

(6) Finally, the way that Romney (and American Crossroads) are running campaigns suggests some problems as well. The Romney campaign feels that it has to play defense on things where Republicans are correct (such as the Lilly Ledbetter Act); it is affirmatively trumpeting protectionism instead of defending free markets. American Crossroads is blanketing Virginia with advertising attacking Democrats from the left for Medicare and education spending cuts. Romney hasn't made the Supreme Court an issue at all in the campaign. I'm clearly not the swing voter who is being targeted by these ads, and perhaps they reflect an attempt to dampen Democratic turnout. But it's hard not to think that Romney doesn't think he can win the election by selling conservative and free-market ideas. That has longer-term consequences even beyond what it says about the 2012 election.

We'll find out on Tuesday night or late Wednesday morning. That said, it's important to realize that this election is only one data point, and in the absence of an unexpected 55-45 landslide one direction or the other, the election will not be proof or disproof of any particular model of forecasting elections.

My swing-state predictions:
Likely Romney: FL, NC
Lean Romney: CO, VA
Too close to call, slight lean Obama if I have to choose: OH
Lean Obama: IA, WI
Likely Obama: MI, NH, NV, PA
Solid Romney: The other McCain 2008 states, IN, Nebraska 2nd District
Solid Obama: The other Obama 2008 states

Thus: Either 281-257 Obama or 275-263 Romney. Note, though, that the electoral-college scenarios generally favor Obama. Obama can lose WI or IA if he wins OH; if Obama wins IA and WI, he only has to win one out of CO, OH, and VA to get to 270. Note further the small chance of a Romney popular victory and Obama electoral college victory, especially if New York/New Jersey turnout is depressed.

Because of this electoral-college firewall, if the leaning states were really independent toss-ups, Obama would have a much better than a 50-50 or 60-40 chance of winning. In fact the errors are going to be correlated, it's unlikely that Obama wins Virginia and Colorado but loses Wisconsin and Ohio, and my 60-40 estimate reflects a certain chance that Silver/Wang is completely right, a certain chance that I'm right or partially right, a small chance that Rasmussen is completely right and I'm being overconservative, and a very small chance that the state polls are completely wrong and we're seeing an undertow election like Barone and Domenech think.

(Updated to add links, fix typos, and add specific swing-state predictions.)

Update, November 4: One more reason to be a bit more skeptical of the polls.

62 Comments

""""
What am I supposed to think about Romney throwing $12 million into Pennsylvania at the last minute? Pennsylvania has consistently voted more Democratic than the rest of the nation.
""""

You should be thinking about elections after 2020... if the GOP doesn't start winning PA, WI, MI, and MN by then they will have no shot of winning the presidency. Even if the GOP gives into amnesty, the hispanic ceiling will be 40% at best for their vote. There's also the issue of a couple southern states like Georgia and North Carolina that may start slipping away too.

You say above that the 538 model assumes the state polls are unbiased, but I don't think that's correct. In his last write-up, Silver says this:

"My argument, rather, is this: we’ve about reached the point where if Mr. Romney wins, it can only be because the polls have been biased against him. Almost all of the chance that Mr. Romney has in the FiveThirtyEight forecast, about 16 percent to win the Electoral College, reflects this possibility."

As far as I can tell, the ~16% is based on his historical analysis of how often state polls have been biased enough to offset the advantage they currently show for Obama. That is of course different from your own take on the amount and direction of bias, but he's not ignoring it.

jt

Great post, Ted.

Do you have a specific technical example of what you think the state pollsters are doing wrong? It's my understanding that the D/R/I splits are an output of their models and process rather than an input/contraint. Is there specific sample rebalancing or other decisions they are making that you think are leading them to get inaccurate topline results?

Obama will win, California, New York, Illinois, Vermont and the District of Columbia.
Obama May win Washington, Oregon, Maine, and Hawaii.

All other states will vote Romney.

The republicans will gain 12 in the Senate and hold the House with 7 to 11 seats gained.

The Popular Vote Split will be - Romney 54 Obama 45

(I've been making this prediction since July)

"The Pennsylvania U.S. Senate race doesn't seem particularly close...."

When did you write this article....August?

The PA senate race is basically tied right now so it is VERY close.

My prediction: Romney with 315 electoral votes. Mass Democrat hysteria ensues. Possible on-air suicides on MSNBC...or at very least, an aneurysm or two.

Should be GREAT fun. But even if Obama should win somehow, Benghazi and Fast and Furious will move out of the Obama deep freeze, go through the courts as necessary to halt spurious 'executive privilege' claims, and then splatter like projectile vomit all over the president's rerun term. He will wish that he lost.

As a Democrat and a Progressive I stand up at and applaud at what is, to my mind, the first LOGICAL analysis of a Romney path to victory. I don't think it'll happen, but dear Lord THANK YOU for proving their is long term hope of restoring a two party, reasonable discussion.

I just wanted to thank you. I got linked here from Nate Silver's tweet and to be real honest you provide a different outlook while still looking at facts and evidence... something I find severely lacking in most conservative blogs/analysis sites.

While we may differ in some specific opinions, I think we both hail to truth and fact. Scientific analysis is mostly thrown out by the far right and I feel it discredits their positions almost universally. I 100% believe that we need sane conservatives and sane liberals involved in political discussion and in long term planning. I am mostly a moderate who is so anti-GOP right now due to their complete separation from reality.

That being said; my family is hard-core GOP (and they're all voting Obama this year, albeit painfully). We need to reign the GOP back in. We need the conservatives to be conservative, and not crazy or ideological.

Blogs/posts like these are a step in the right direction. I thank you for the time spent writing it and the analysis you've come to regarding it.

Bravo. Hope to read more in the future.

I appreciated your post - good thoughts, and some I haven't thought of, as my partisan leanings are the opposite of yours, thus our blind spots are probably a bit complementary.

One thing I think you're missing are the cases for the polls to be biased the other way. Now, I'm not saying these are more or less likely than your seven, just that they should be factored in.

For one, there's the Hispanic vote. There's some work out there Showing potential under polling of Latinos, which break strongly Democratic. This could matter in NV, CO, a bit in VA, and possibly FL, though its Hispanic population significantly differs from the rest, so that's a tougher call.

There's also the probably over-discussed cell phone problem. This has some interesting caveats which could move things either way: Democrats are actually more likely to be either wholly landline or cellphone, with Republicans more likely to have both (according to some work; I wouldn't take that as sacrosanct). Between that and polling effects lies a jungle of methodology, but the general consensus seems to be robocalling firms would slightly underpoll Democrats.

All in all, I think including these, along with your seven, probably should shift an estimate a bit toward Obama, maybe making him a 65% favorite to you.

Regardless, good post, and I appreciate the lack of rancor and acknowledgment of partisan blinders - refreshing in this ugly cycle.

"I'm not going to succumb to the fake precision of Nate Silver's model and suggest I can forecast with three significant digits."

I dont think that is what Nate Silver does or implies. If you have lots of factors and probabilities, you're bound to get an uneven result.

That said, i like that you dont come across as so many others favouring Romney : arrogant and ignorant at the same time.

You make some good points, but if someone put a gun to my head, i would choose Nate Silver's assessment over yours.

I'm a Democrat, but this is exceptionally good analysis. If more Republicans (and Democrats) thought and communicated like this, we'd all be better off.

I think the general point is especially valid, and lost in all the 538 hubbub. Nate isn't actually going out and asking anyone anything. His analysis can only be as good as the aggregate value of the information he's given. I believe in his method for aggregating polls, but if the polls themselves are all suffering from the same incorrect assumptions, Nate will be, too -- reinforced over and over again.

Very well thought out and very interesting! Makes you feel better if you are an Obama supporter and a lil nervous if you are a Romney supporter. Good write up!

The polls have been so much spin. This analysis is refreshing.

I live in Ohio, and so like any person who lives in a swing state, I've a better vantage point for what is happening here than those outside of the state. We are the targets of the ads, we have our local news outlets constantly reporting on the topic from a local perspective, we have internal state polls published everywhere, we see the candidates and their rallies, etc...etc... That being said, I can say quite confidently that Obama's lead in Ohio is strong, and has in fact remained strong even after Romney's first-debate bounce. Local polls have Obama ahead between 5 and 8 points as recently as this morning (Saturday Nov 3rd, 2012) but these numbers have been consistent since summer. When all Ohio polling data is aggregated Obama's lead stands at a firm 4 points here (the advantage of aggregating is that there is negligible or no margin of error). As you and everyone else knows, Romney cannot win without Ohio; his electoral vote tally simply cannot reach the required 270 electoral votes. But when we aggregate and graph the polling data in the other swing states such as VA, CO and even FL, the movement has clearly been toward Obama over the past 10 days. So I will propose an alternative to your closing state predictions as follows: Romney will almost certainly win NC (current aggregate lead of 1.5 points) but Obama will carry CO (current aggregate lead of 2 points), VA (2 points), OH (4 points), IA (3 points) and WI (6 points). Florida remains very close, but in the newest polls released between the Weds to Fri time period Obama has gained an aggregated polling lead of 1 point which is consistent with predictions earlier this year for the number Obama would win FL by, namely 1 point. So the final electoral vote tally that the world will see on the morning of Weds Nov 7th will be: without a FL win Obama 303 - Romney 235. With a FL win Obama 332 - Romney 206. Personally, I expect it will be the latter. Many might consider an Obama win in FL to be the surprise of the night, but the real surprise would be if Obama managed to win NC. It's unlikely but because Romney's lead is so small he is vulnerable so it is barely possible. Obama only won NC by 1/3rd of 1 point in 2008. If Obama carries NC he will definitely carry FL making the final EV tally Obama 347 - Romney 191. But again I think Obama 332-206 will be the final total.

Unfortunately, your election prediction betrays your own bias. Silver has repeatedly emphasized that his model accounts for the possibility of across the board bias in state-level polls. That is why Romney still has about a 16% chance of passing 270.

You can keep trying to build an argument through six or seven or eight layers of statistical "analysis" about why all of the polls are wrong, but, as you pointed out above, there is no historical reason to think that this is the case. Further, there is no contemporary reason to think this is the case. And finally, it is highly implausible that *every* pollster this cycle has gotten it wrong *every* time.

Nice try, although a huge waste of time.

Hi Ted; long time, no see. (Granted, rec.sports.baseball lost relevance way back in the day.) Not much time to respond here, but one point that I think may be important that you are underestimating is: recent polling trends have been pretty much uniformly negative for Romney. And in this case, we know the numbers could move the other way, as they did following Obama's comatose first debate performance. I do not for a minute buy into the "it's over" message that Wang has been trumpeting, but if the latest polling trend does *not* concern you...you aren't looking at the numbers hard enough.

One other point: Romney blowing $12 million on Pennsylvania really should either strike you as an incredibly stupid move (think pre-money ball baseball GMs) or else a move of complete desperation, which is how I read it. In any case, we will know the answer shortly. I don't expect we would see Romney tip any concession on the main race to help, for example, the closest senatorial races, but any appearance by Romney in an expected red state would be a total giveaway. So if Mitt does an event in Arizona or Indiana or Montana or North Dakota on Monday, I think we can agree what the meaning would be.

Wow, an intelligent and coherent rebuttal to Nate Silver that doesn't descend into either innumeracy or right-wing bluster. I'm impressed. I'm also a strong D supporter politically, so I hope Obama wins, but that's besides the point. Congratulations on a clear and honest analysis that lays out where you're coming from and why you reached your conclusions, and we'll see what happens Tuesday night.

nice analysis but misses at least one big factor that favors Obama: Latino and Asian immigration. Particularly important is Romney under performing Latino vote which expects a record turnout rate. Romney also doing worse than McCain with Blacks, though a lower turnout is expected. These issues may offset the passion gap among other groups that favor Romney.

Come on. You don't even know that there's an EV at play in Maine. You guys are funny, coming into political talk wagging your alpha dog statistic smak and you don't even know the lay of the land.

fyi, PPP just polled ME02 and found it at 51-46, it's probably in the Lean Obama column.

First of all, thank you for writing a coherent argument for why the polls might be wrong. Too many pieces from the right reed like screeds, full of ad-hominems rather than logical arguments.

Your main argument seems to be that the pollsters are getting the composition of the electorate wrong. Essentially that it is unlikely that the electorate will be D+7. But don't pollsters ask a random sample of the electorate, and get them to self-report the results. Most don't delibrately poll a certain number or democrats and a certain number of republicans. The results come as they poll the electorate. So you must believe then, that Republicans are less likely to answer the phone and give an answer, but given the democratic coalition (young voters, minorities etc), wouldn't they be more likely to be missed than Republican voters? And don't most polls show a large difference between registered voter results and likely voter results, so aren't pollsters already taking the lack of Democratic enthusiasm already into account? What would be your theories on why pollsters are showing a vastly different electorate than you would expect, is my question. Are republicans not answering the phone or are they being missed all together in pollsters survey's of various states.

"I have to recognize that there's some chance that I'm suffering from confirmation bias, looking only at the data that favors my preferred result"

This. You've devoured wholesale every crumb of wishful thinking available to you.

Nicely done. I'm a fan of Silver's work but have been troubled by his reliance on some of results from 2008. Your perspective is refreshing, especially in a time of "unskewedpolls.com" and similar criticisms of Silver that evince a fundamental misunderstanding of basic statistical theory. Thanks for laying out a rational, call-it-as-I-see-it case for a tighter election; one that doesn't rely on hyperbole or ad hominem attacks. It's both rare and welcome.

One factor that may work in Obama's favor--and which is also hard to quantify--is the degree to which the Republican brand has been further tarnished since 2008 because of the Tea Party. I think that's a headwind for Romney that could counteract some of the assumptions you've made.

One thing that is often missed in analysis of polls based on party affiliation is that pollsters do not actually ask about registration, but rather what someone considers themselves at the moment. So Romney leading independents broadly doesn't really mean that he is leading people registered as independent.

It is just as plausible that a large lead in "independents" as reported by polls could simply mean that a large portion of registered independents are siding with Obama and saying they "consider themselves" Democrats, and all that are left calling themselves independent are people who lean Republican but aren't particularly enthused by this campaign.

The same "errors" in "over-sampling" democrats were noted in the 2008 election, where polls were incredibly accurate.

I also find it somewhat questionable to act as though top-line polling is somehow less accurate than the breakdowns of poll subsets, which obviously have much more room for systemic error because of smaller samples. The bottom line is those subsets *have to* add up to the top-line results, so finding subset results that call a state differently than the top-line results *must mean* cherry picking data to support one's own view.

I can't really bring myself to call Ohio a toss up when Obama has led virtually every poll held there in 3 months. Obama is clearly a heavy favorite to win this election unless, of course, polls turn out to be poor predictors of outcomes. If that's the case, then, well, we'll really have to start re-thinking politics.

At last - an intelligent discussion about possible outcomes. Thank you.

You're dreaming..............................

jd: seriously?
so you believe Romney will win CT, RI, MA, MD, DE ??? when was the last time any of those voted Republican?

I could see making an argument for Romney winning the "tossup" states as most people see them, but you're predicting something that NOBODY ELSE thinks is even remotely possible.

I'm assuming your comment was a joke, right? If not, I look forward to your "I admit I was wrong" comment on Wednesday.

Pennsylvania has always been Republican's fools gold in Presidential elections.

Centrist Democrat here via silver tweet. I found this interesting and appreciated the tone of the article and the comments.

I found your comment dismissive and snarky.

I second the many commenters who have thanked you, Mr.Frank, for your very well-articulated arguments. I too was directed here by a link from Nate Silver's site, and it's refreshing to have a careful and thoughtful approach from the right. It's been disappointing and outright disturbing how much misinformation about statistics and polling has been put forth by professionals and non-professionals alike.

I am a survey research professional. I'm in market research, so not the political arena, but it's very similar - i.e., trying to estimate what a population thinks about things. As a survey researcher, I am most interested in how these issues of bias might occur from a poll methodology perspective. On this topic, I wanted to share my thoughts on some of your points, and hear your thoughts as well.

As I read your arguments, I think what you are primarily talking about is the "likely voter" screen that is crucial to polls. As other commenters have correctly pointed out, the vast majority of polls DO NOT weight their samples by reported party affiliation. That is an outcome of the research, not a weighting variable. What almost all pollsters DO weight on are demographic variables (i.e., gender, age, race/ethnicity, usually education, usually geographic region for national polls). The goal is to collect a sample that accurately represents the adult registered voter electorate, which is obtained through census estimates. This is exactly what I do in my market surveys - I want a respondent pool that mirrors the US population. (By the way, only one major polling firm that I'm aware of - Rasmussen - deliberately DOES weight by reported party affiliation. As a researcher, I think this is a bad idea, but that post can be for another time.)

Then - and this is the crucial step - each pollster tries to figure out whether the respondent is a likely voter. (Yes, there are "registered voter" polls too, but we de-emphasize those this late in the cycle.) This is done differently by different pollsters, but the GOAL is to separate likely voters from all others, and then the poll reports those numbers.

I think it's this screen that is at the heart of your proposed "issues," but I don't think I agree with you that these issues are left unaddressed. For example, you very reasonably propose that Republicans might be more enthusiastic this election, or perhaps they have been better reached by GOTV efforts. Both ideas are possible. BUT, a pollster should be capturing already this through the "likely voter" screen. Likewise, the "independents breaking for Romney" should already be captured in the poll results. *** In other words, I suspect that the items in your Point #3, especially items b-c-d, are already "baked in" to the poll results, so you shouldn't really count on them showing up on Election Day for the first time.

I also wouldn't put too much weight on your subpoints (a) and (f). You rightly dismiss (f) - We just have no way to know how early voters compare with Tuesday voters, so it's not worth it to chase too hard. As for (a), there is an enormous tendency for people to (1) remember themselves voting for the winner in previous elections, even when they didn't, and (2) construct a narrative to align with their current position (e.g., "I'm really angry with Obama, and that anger will come across more clearly if I'm perceived as someone who voted FOR him before"). So, I would be dubious of claims of large groups of people who claim to have voted for Obama but have now switched. I absolutely believe those people are out there and exist - but, measuring them objectively will be very difficult.

Where I *do* think you have a valid thing to hope for is in your Point 3-e - racial/ethnic turnout. Yes, if the actual turnout is very white, Romney will likely do better. But, this largely comes down to the age-old statement, "Gotta get out the vote." It's not an outcome that is estimated by polls. So, if Tuesday sees a particular, unforeseen mix of voters, then yes, that will impact the results. But, I don't know of any evidence that that will happen in EITHER direction. (The Battleground link you provided is plausible, but not convincing. That can be for another post too.)

As a survey researcher, therefore, my post-election goal to improve surveys will likely focus most on the Likely Voter screens. If Tuesday goes as is predicted by the vast majority of pollsters, then I might conclude that the screens are doing what they're meant to do. (Always room for improvement, but the evidence would be that they're on the right track.) If Tuesday breaks very differently, then we (survey researchers and pollsters) need to figure out how we were screening people incorrectly, and somehow mis-labeling or undercounting some people who did indeed decide to vote.

As you impressively admit with integrity, the weight of the available evidence is not on your side at this point. But, I thank you for pointing us toward the areas that we should perhaps look at if the predictions turn out very wrong. I hope I've helped add to and/or provide focus to the possibilities you've raised.

Mass insanity herein, albeit I have seen this type of delusional thinking many times before. Mathematics is by design a finite and explicit; quantitative science. The Silver and Wang math is clean and correct. The election is all but over. Democratic Whitehouse and Senate...possible change in the House leadership.

It really is amazing to me that someone would waste so much time writing this piece based on a false premise that proves the writer was either too lazy or too dishonest to even read Nate Silver to understand his analysis.

It's simple: 538 includes the chance of poll bias in the model. That's why the probability of Obama winning is not 100%.

Seriously, writing elegently or politely, but still lying, makes you even worse than the idiots that are agreeing with you here.

Really enjoyed this post. The bottom line is that Silver's analysis is a basically a slightly better mousetrap as far as poll averaging, no more, no less. The only place where I think he may have loaded the deck a little is in adding in an economic weight that slides out of the polling pretty fast in the last month. He had to know that given current conditions, this would produce the impression of slight movement towards Obama. But in any event, Nate's predictions boil down to a slightly more sophisticated RCP no-tossup EC map. If the polls on average missed the composition of the electorate, then it will be a case of garbage in, garbage out, for RCP and 538 alike.

Where you disagree with Silver is on the likelihood that the polls are systematically inaccurate. He looks at historical inaccuracy in assigning Romney a 17% chance to win. Virtually all of Romney's chance is due to that error, by the way, so his metric is unlikely to show Obama 95% on election day. However, there are several reasons to think that the polls are inaccurate this year, and you've highlighted all of them quite intelligently. I would just add that with the rise of caller ID and cellphone-only households driving response rates way below 20%, I think there is an above average change of there being an election sometime in the next decade that the pollsters blow badly. The only question is whether it will be this one.

Ironic that you accuse me of being "too lazy or too dishonest" to read who I'm critiquing, given that you were plainly too lazy or too dishonest to see that I said "Nate Silver has made a decision to have the 538 model disregard each of these factors and assume that the polls are as likely to be biased for Obama as for Romney."

Again, than you for such a coherent and thoughtful analysis. Given the current ideologically warped alternate reality that passes for knowledge and wisdom in the current GOP it is refreshing to hear a sane voice from the right.

Regarding your point 3(d), that Republicans are more excited than Democrats in this election, I think you are right. There is always more enthusiasm for throwing out the current status quo than keeping it. On the other hand, what I sense in the Democrats I know (and yes, I'm biased to D) is more an anger at the GOP than an excitement for Obama. This anger comes from 8 years of Bush and 4 years of Republican obstruction in congress. Many, many D voters are eager to vote Obama back in because they can't stand the idea that thug tactics will be validated in this election.

For these voters it doesn't really matter who the Republican candidate is, they (a) can't stand the idea that destructive obstructionism will become the new normal on the hill, and (b) don't believe the country will thrive under another Republican administration. Yes, Reid in the Senate can block legislation, but all appropriation bills originate in the House, and reconciliation is a powerful tool if the White House is ready to sign the bills.

So if the pollster asks, "Are you enthusiastic about your choice?" after 4 years of a "Great Recession" economy many D voters might say no, or not much even if they really think Obama has been a great president. That doesn't mean that they are not smoldering hot voters, though. Most that I know would crawl over broken glass to keep Romney from becoming Commander in Chief.

Long story short - Many D voters are voting D as in Defense, not O, this year.

Again, than you for such a coherent and thoughtful analysis. Given the current ideologically warped alternate reality that passes for knowledge and wisdom in the current GOP it is refreshing to hear a sane voice from the right.

Regarding your point 3(d), that Republicans are more excited than Democrats in this election, I think you are right. There is always more enthusiasm for throwing out the current status quo than keeping it. On the other hand, what I sense in the Democrats I know (and yes, I'm biased to D) is more an anger at the GOP than an excitement for Obama. This anger comes from 8 years of Bush and 4 years of Republican obstruction in congress. Many, many D voters are eager to vote Obama back in because they can't stand the idea that thug tactics will be validated in this election.

For these voters it doesn't really matter who the Republican candidate is, they (a) can't stand the idea that destructive obstructionism will become the new normal on the hill, and (b) don't believe the country will thrive under another Republican administration. Yes, Reid in the Senate can block legislation, but all appropriation bills originate in the House, and reconciliation is a powerful tool if the White House is ready to sign the bills.

So if the pollster asks, "Are you enthusiastic about your choice?" after 4 years of a "Great Recession" economy many D voters might say no, or not much even if they really think Obama has been a great president. That doesn't mean that they are not smoldering hot voters, though. Most that I know would crawl over broken glass to keep Romney from becoming Commander in Chief.

Long story short - Many D voters are voting D as in Defense, not O, this year.

Hi Ted. Hope you are well. Do you still remember standing in Grand Central Station together during AFU NYC II, holding a paper plate that said "Diane Kelly of AFU" so we could bring her to the party?

Just wanted to say thanks for this cogent analysis. I'm generally a bleeding-heart liberal with some moderate fiscal sympathies, but most of all I'm a scientist who appreciates clear thinking and logical analysis on both sides of an issue.

This addressed my point. Thanks even if it was not for me.

Romney probably put money in Pennsylvania to reach Eastern Ohio, just like he put money in Minnesota for no other reason than to reach western Wisconsin.

I literally laughed out loud to this gem:

"Obama will win, California, New York, Illinois, Vermont and the District of Columbia.
Obama May win Washington, Oregon, Maine, and Hawaii.

All other states will vote Romney."

I followed Silver's tweet here. Great post. My one comment is that I do appreciate Silver posting his odds to three digits, especially since he does such a good job explaining why the probabilities change day-to-day. The one thing I'd love to see on his site in addition is a means by which the reader can toggle the various assumptions described above and watch how the odds change. I would be curious to see the odds that Romney wins if, for instance, the average poll for the swing states were x% biased and the Romney turnout were y% better than '08.

Based upon early polling Romney easily wins Florida and North Carolina.

Romney wins by a little in Colorado, Ohio and Iowa.

Nevada is too close to call.

It is easy to demonstrate the bias in the state polls by just comparing early vote in comparison to what is represented within the polls. Romney is winning by 2 points in Colorado and yet within the polls that show Obama is winning the state they are showing that Obama is winning the early vote by 6. This is impossible unless Obama is crushing with Independents or is polling more Republican votes than Romney is pulling Democrat votes. Both scenarios which are very unlikely. What is likely is that they are overcorrecting for ethnicity and pushing a Democrat bias.

Votes in heavily Democrat leaning Cuyahoga are down substantially over 2008 as are Democrat votes in all of the swing states.

White vote in 2008 was 76% per the Census Bureau, not the 74% that everyone parrots from the CNN exit polls which were shown to be inaccurate.

The national polls are over correcting for ethnic vote according to a projection of 74% white vote which yields the D+7 polls which are not accurate. It is difficult to believe that anyone believes that Obama will have a 2008 turnout when every poll of excitement and partisan make up suggest that we will see an even or R+1 election versus the D+7 of 2008.

If the turnout is D+7, of course Obama wins and Silver continues to be the Silver Boy. Otherwise Romney wins and Silver looks like a partisan hack. Silver does not treat all polls the same. He downgrades polls like Rasmussen which are Republican leaning, and gives extra credit to NYTimes and PPP which is a purely partisan poll paid for by SEIU. That is bias, plain and simple.

My prediction is 295 Romney, 243 Obama with a 70% chance that Romney wins.

Just wanted to echo some of my fellow Dems and say thank you for a rational, fact-based argument free of personal attacks. Also want to thank those of my fellow Dems who have come here and not acted like jerks.

In an era where there is huge political benefit in ratcheting up the emotional factors in an election, it's good to see that reasoned, respectful discourse can exist.

I generally agree with your conclusion to all this, that 538 is overstating the President's chances of being re-elected.

But, with all due respect, I think your dismissal of the 2000 elections on the basis of the Bush DUI october surprise is a reach. It's not an event that was forgotten because of the recount, it's event that is mentioned along with other insignificant october surprises.
In 2000 George Bush's past drug and alcohol use was common knowledge, it's about on the level of the revelation of Obama's illegal immigrant aunt, or the Sadaam Hussain verdict. I can't imagine it would move the polls 3+ points. Did Mitt's 47% remarks, or Obama's terrible debate performance even move things that much?


I wrote a similar comment after you October 24th posting, but to somewhat repeat myself - I've looked at 5 reputable aggregate polling sites (Silver/Wang plus Blogging Caesar, RealClearPolitics, and TPM) and they all are making very similar predictions. I've followed the Blogging Caesar (very conservative in his politics, but quite honest in his meta-polling)since 2006 and his track record has been very good (missed 2004 by 3 EV's & missed Indiana & NC in 2008). I've noticed that RealClearPolitics has had Obama in the lead every single day in their "No Toss Ups" map - Romney's high water mark was Obama 277 EV's. Given the track record of the sites that I've mentioned, and throwing in IEM having Obama at 70% now, will go with proven track record & put Obama as a 70+% favorite at this point.

Your final predicts do seem to lean too much in the hope column. RCP has Obama leading Ohio at 2.8%, but Iowa at 2.5%, so why would you have Iowa lean Obama but Ohio too close to call? I would have to assume it's because Romney can afford to lose Iowa and still win, but not Ohio.

Your question about $$$$ in PA. Have read that now with only 11 (or fewer) swing states, campaigns no longer have to make hard choices of where to spend their money. If you still have millions sitting in your campaign chest, why not spend every last cent you can - it's not like it costs you personally anything. (Tman's argument sounds true as well.)

Well, it's 3 days to the election, so will look forward to reading your post the morning after.

There is little chance of that happening.

A reasonable and well thought analysis, however I disagree with the result. The reason is that you state that you believe that the only way Romney can win is if the state polls are biased. So, what you're saying is that you believe there is a 40% chance that the state polls are so biased that Obama won't win the election. I find that hard to believe. Obama is being shown as a 3 point favorite in Ohio right now, and 538's average error was 2.5 percent in 2008. If Nate Silver's model is biased, it is likely biased by the same amount it was in 2004. Is it possible for Romney to win? Yes, I just don't think it's at a 40% shot.

As a reader of Nate Silver's, I don't agree with you about the "Democratic over-weighting" concern: I believe it's a non-issue. Other comments above state the case.

Nonetheless, looking at your final summary:
"My swing-state predictions:
Likely Romney: FL, NC
Lean Romney: CO, VA
Too close to call, slight lean Obama if I have to choose: OH
Lean Obama: IA, WI
Likely Obama: MI, NH, NV, PA
Solid Romney: The other McCain 2008 states, IN, Nebraska 2nd District
Solid Obama: The other Obama 2008 states

Thus: Either 281-257 Obama or 275-263 Romney"

My two top guesses are exactly the same as yours! The make or break state is Ohio.

Thank you Ted, from Italy.

Bill Clinton is doing 4 campaign stops in PA this monday. What's to think about that?

Your analysis, coming from a conservative, is exceptional. Good step towards proving : "Most insane people are replublicans. But all republicans are insane" wrong. ;)
Hope to read more from you now on.

Leave a comment

Once submitted, the comment will first be reviewed by our editors and is not guaranteed to be published. Point of Law editors reserve the right to edit, delete, move, or mark as spam any and all comments. They also have the right to block access to any one or group from commenting or from the entire blog. A comment which does not add to the conversation, runs of on an inappropriate tangent, or kills the conversation may be edited, moved, or deleted.

The views and opinions of those providing comments are those of the author of the comment alone, and even if allowed onto the site do not reflect the opinions of Point of Law bloggers or the Manhattan Institute for Policy Research or any employee thereof. Comments submitted to Point of Law are the sole responsibility of their authors, and the author will take full responsibility for the comment, including any asserted liability for defamation or any other cause of action, and neither the Manhattan Institute nor its insurance carriers will assume responsibility for the comment merely because the Institute has provided the forum for its posting.

Related Entries:

 

 


Isaac Gorodetski
Project Manager,
Center for Legal Policy at the
Manhattan Institute
igorodetski@manhattan-institute.org

Katherine Lazarski
Press Officer,
Manhattan Institute
klazarski@manhattan-institute.org

 

Published by the Manhattan Institute

The Manhattan Insitute's Center for Legal Policy.