## 6.17.2008

### Defining a "Must Win" (Technical)

I had previously been defining a "Must-Win State" (see the Swing State Anaylsis graph along the right-hand side of the page) as a state which is won the highest percentage of the time by the candidate winning the election. The problem with this definition is that it was producing some skew toward Obama-friendly states, since we now have Obama winning the election the strong majority of the time period.

The new definition is that a Must-Win State is a state won the highest percentage of the time by the winning candidate when the election is close. By "close", I mean an election in which the popular vote is within 4 percentage points. If the popular vote is outside that 4 percentage point range, it is pretty much mathematically impossible for the trailing candidate to win in the Electoral College, meaning that the electoral math becomes irrelevant. So these are the states that the candidate needs to win when winning individual states matters.

### Trendline now calculated from daily numbers

I have switched from calculating our trendline adjustment based on a weekly aggregation of polls to one instead based on daily numbers. The methodology otherwise remains the same.

The reason for the change is simply that based on some experimentation, the daily calculation appears to be a little bit more robust -- it fluctuates less from day to day as new results are added. This is also how Professor Erikson had suggested that I perform the trend calculation originally. The daily method does tend to give more deference to national tracking polls in extracting its trendlines, but since Rasmussen and Gallup are interviewing literally tens of thousands of voters between them each week, I'm not sure that this is a bad thing.

As a side benefit, the daily version of the trend adjustment seems to produce a cooler-looking graph:

### Today's Polls, 6/17

An odd day of polling, but one attention-grabbing result dominates the rest. That is from Ohio, where Public Policy Polling has Barack Obama ahead by 11 points. While Public Policy Polling developed a reputation as being somewhat Obama-friendly in the primaries, its track record is fairly strong, and its prior Ohio poll -- taken way back in March -- had shown McCain ahead by 8 points. As Ohio is probably the single most important state in this election (it's by no means the only important state, but it's pretty darned important), this result is enough to drive Obama past the 67 percent threshold in our overall electoral projection; we presently have him as about a 2:1 favorite to win the election.

In Minnesota, however, SurveyUSA has Obama with just a 1-point lead over McCain. SurveyUSA's methodology takes a more fluid view of party identification, and so it tends to produce results that can be more encouraging for the non-dominant party in a particular state. Its most recent previous Minnesota poll, taken back in May, had shown Obama ahead by 6 points.

In North Carolina, Civitas has John McCain ahead by 4 points -- down a tick from the 5-point lead he held a month ago. Obama has yet to show a lead in North Carolina, but has trailed by somewhere between 2 and 4 points in the three most recent polls of the state.

There is also a SurveyUSA poll out in Kentucky that shows Obama trailing by 12 points. This poll made it across our wires too late to be included in our metrics, but it speaks to the extent that Obama is starting to improve his numbers among lapsed, Clinton-leaning Democrats, particularly in Appalachia. Obama had trailed by 24 points in Survey USA's May poll of Kentucky, and by as many as 36 points previously.

There are also a series of national polls out, all of which have consolidated in the area of Obama +4, exactly the popular vote margin that we attribute to him based on the state-by-state polling results.

So what to make of the meme that Obama's numbers haven't been bouncing? The only way that you can come to that conclusion is if you cherrypick results. There have been a few dozen polls released since Clinton conceded the primaries, and our methodology extracts an average bounce of about 4 points between them. Four points is not so large that some individual polls won't show a bounce, particularly if the bounce is concentrated in particular states and regions. But bounce Obama has, and the longer Republicans remain in denial about it, the less time they'll have to catch up.

### Can a VP Nominee "Win" a State?

As a prelude to some work Nate is doing on VP picks, one of the common conventional wisdom canards on presidential elections is that the presidential nominee can - and maybe should - pick a vice-presidential candidate who adds a particular state into the win column in the fall. For example, Monday afternoon on the MSNBC show assigned to David Gregory, an inane conversation took place about the notion that Rudy Giuliani could bring New Jersey into play if he were McCain's VP nominee. This highly questionable theory about just what electoral gain a VP can contribute deserves closer inspection.

Consider the rationales offered for many potential choices even now: Ted Strickland for Obama or Rob Portman for McCain because they bring Ohio, one of the Virginia options for Obama because they bring Virginia, or Tim Pawlenty for McCain because he “puts Minnesota in play.” Further examples are numerous. Bill Richardson supposedly brings New Mexico and the Southwest in general; other candidates are considered to have a strike against them if they do not bring an obvious electoral benefit.

The emphasis placed on such a strategic pick is born of two consecutive nailbiter elections, where the flip of one battleground state has determined the winner. Although it’s somewhat understandable that this conventional wisdom has emerged, the evidence demonstrates it’s hypothetically sketchy at best.

In looking at the vice-presidential selections of the past five decades or so since television has expanded the regionality of presidential elections, it’s clear that, in reality, both major parties rarely have nominated VP candidates as a strategic electoral vote collector, and to the extent they have set about deliberately trying to add a state with a VP pick it has almost never worked.

Taking a look at the Republicans and working backward, Wyoming (Cheney) was always in the Republican column except for Johnson’s ’64 landslide; New York (Kemp) had been reliably blue since the 60s with the exception of the Nixon and Reagan landslide years; Indiana (Quayle) and Kansas (Dole) had been reliably red since FDR except for ’64. Even Spiro Agnew, when he was added to Nixon’s ticket in 1968, could not bring Maryland into the Republican column until 1972 as the incumbent in a national landslide. Republican VP picks in 1964 (Miller, New York) and 1960 (Cabot Lodge, Massachusetts) failed to bring those states into the fold, and it’s hard to think Republicans chose Cabot Lodge strategically in a year where the Democratic presidential nominee was from the same state.

You could argue that selecting George Bush in 1980 was a strategic pick to gather Texas, a state that had voted Democratic essentially since the Civil War except for Eisenhower’s two terms and Nixon’s ’72 landslide. But given the larger macro forces at work in Texas, a state that voted Democratic for most of the previous 100 years and then hasn’t been competitive for Dems since 1976, it’s hard to chalk that shift up to the popularity of George Bush or appreciation to Republicans for putting him on the ticket. Put another way, it would be like Obama choosing Sebelius of Kansas and then Republicans not being competitive there for the next three decades and counting.

For Republicans, one really has to go back to Richard Nixon of California where a state flipped from Democratic (5 straight elections) to Republican. Even then, the popularity of FDR and Eisenhower were far bigger macro forces than the drawing power of a young Richard Nixon.

For Democrats, John Edwards obviously did not make a competitive state out of North Carolina, whose only post-Southern Strategy flip back into the Democratic column was Carter’s 1976 win. In 2000, Democrats won Connecticut for the third consecutive presidential year as part of a larger solidification of the northeast.

Although Dems won Tennessee in 1992 and 1996 with native son Al Gore on the ticket, bringing the state back into the Democratic column for the first time since Carter’s lone post-Civil Rights Act 1976 win, the fact that as the headliner Al Gore couldn’t win his own state in 2000 indicates that Bill Clinton had more to do with winning Tennessee in the 90s than did the VP choice. Similarly, the choice of Estes Kefauver in 1956 did not win Tennessee for Dems at a time when 1952’s loss of the state was an anomaly from the previous couple decades.

Lloyd Bentsen could not bring Texas back for the Democrats, the racist Geraldine Ferraro could not hold New York in the 1984 landslide, and Sargent Shriver in 1972 could not keep Maryland’s three previous Democratic preferences going strong.

The best Democratic examples of a VP helping with a state are Walter Mondale in 1976 and 1980, Edmund Muskie in 1968 and Lyndon Johnson in 1960. Muskie is perhaps the best example, simply because with the exception of 1964’s landslide, Maine hadn’t strayed from the Republican column since Woodrow Wilson in 1912, and then promptly went back into the red column afterward. Lyndon Johnson undoubtedly helped the Catholic Kennedy in Texas, but Texas at that time was reliably Democratic anyway. And Walter Mondale certainly helped the Baptist Southerner Carter in 1976 and 1980, but Minnesota had been a reliably Democratic state since FDR, with the exception of Eisenhower’s two elections and the ’72 landslide.

In order for a vice-presidential candidate’s home state to be a strategic addition, it would have to be true that but for the selection, that party’s ticket would not have carried the state. And you really have to think about how this would come about. Which voters would vote for one ticket who would ordinarily vote for the other ticket or stay home? This extra margin could be a function of extra in-state voter organization and/or extra enthusiasm that makes the difference in a razor-thin race. Such hypothetical voters have to be politically plugged in enough that they know they definitely like the VP nominee, but undecided enough about the two major presidential choices that it’s the VP who closes the deal. Not only does it seem a little far-fetched that such voters would be around in any meaningful numbers to tilt an important electoral battleground one direction or another, but it seems especially far-fetched in a macro contrast election year such as this one.

My pet theory that spins off this VP-electoral vote argument, untested and probably untestable, is that such voters are more likely to exist in small, typically ignored states with 3-5 EVs. For example, the pride for North Dakotans of having one of their own in such a high profile role. Or Hawaiians. Even that might not be enough, but it’s more in alignment with intuition about history-making candidacies capturing the imagination of voters who might otherwise have stayed home or gone the other way. This theory certainly dovetails with the best example in the last 5 decades: Edmund Muskie of small-state Maine.

### Guess Who's Blogging for the National Press Club?

Last time I'll delve into this particular off-topic topic, but I was directed today to the writings of Steve O'Hearn, the chair of the National Press Club's New Media Committee who maintains a blog at the National Press Club's blogging website. Suffice it to say that I disagree with pretty much everything that Steve has to say -- from his position on Larry Sinclair to his insistence (three days after the Montana and South Dakota primaries and one day after Clinton announced that she would concede) that the Democratic nomination race was "far from over", to his calling Tom Shales, the Washington Post's Pulitzer Prize- winning media critic, a "cretin". And then there's the matter of his preferring AltaVista to Google...

But that's not what I wanted to draw your attention to. Instead, I wanted to see who else had a blog on the National Press Club's website. Not very many people do. There is O'Hearn's blog, and then blogs for several of the NPC's committees, and a blog for the Press Club's softball team. And then there is a blog for...

Jeff Gannon.

Yes, that Jeff Gannon. The fake reporter from a fake news organization who asked fake questions at White House press conferences. And who also happens to be a gay escort.

This man has a blog at the National Press Club's website.

In this context, the National Press Club's decision to host a press conference for Larry Sinclair makes a lot more sense.

### Tea Leaves

To follow up on Sean's post, Sean Reagan at The Back Forty (not our Sean) tracked down a list of 17 states in which the Obama campaign is focusing its Organizing Fellows program. We can contrast these against the 16 states that McCain identified as toss-ups in its strategy briefing, and the 17 states that appear on one of the two versions of our Swing State Analysis:

One needs to be careful when pruning through these lists, because they may have as much to do with branding as the realities of what the campaign will ultimately do (particularly McCain's list, which was couched in what amounted to a donor pitch). Nevertheless, the contrasts are interesting. Obama sees North Carolina, Virginia, New Jersey and Georgia as competitive whereas McCain does not. McCain sees Connecticut, Maine and Minnesota as competitive whereas Obama does not. Both candidates claim to want to focus on Washington and Oregon, even though the 538 list does not see them as competitive. Neither candidate claims to be focusing on Indiana, Montana or North Dakota, although our lists consider those states to competitive.

The states on McCain's list went to John Kerry by an average of 1 point. The states on Obama's list went to George W. Bush by an average of 2 points. The states on the 538 list went to Bush by an average of 6 points. McCain's list is ridiculously aggressive in a climate where the partisan advantage is shifting to the Democrats. One can understand that he does not want to lock into a defensive posture this early in the campaign, but he dawdles in organizing Virginia and North Carolina, he may find it impossible to regain the lead in the former, and difficult to shake Obama in the latter.

My critique of Obama's list would be the inclusion of Georgia rather than Indiana. This may be a reflection of the Obama campaign's belief that it can improve registration and turnout among traditionally low-turnout groups, like the African-Americans and young voters that are plentiful in Georgia. I think I like Indiana better, however, from the standpoint of portfolio theory. It's still very difficult to imagine Obama winning Georgia without winning North Carolina, and if he's won North Carolina, he almost certainly won't need Georgia. Indiana, on the other hand, offers a relatively unique set of circumstances. Obama is the first Midwestern Democrat to have received his party's nomination in years (and hails from Chicago, almost literally in Indiana's back yard). The Democrats devoted attention to Indiana for the first time in years as a result of the state's important primary. And Indiana is an extremely manufacturing-heavy state at a time of recession. I don't quite trust the couple of polls that showed Obama ahead in Indiana, but I can more easily see it being a surprise state that actually makes the difference between winning and losing the election.

### 538's Battlegrounds as of Mid-June

Those with field experience may recognize the conventional wisdom/general rule that 5 points is the most a great field program's ground game can make up against an accurate poll of voter preference.

According to 538 regression analysis, there are currently eleven states closer than 5 points, most of which show McCain with a narrow lead to defend.

In order of closeness and color coded by who currently leads, they are:

1. Virginia, 0.2%
2. Missouri, 0.3%
4. New Hampshire, 1.0%
5. Michigan, 1.5%
6. Ohio, 1.6%
7. Indiana, 1.7%
8. North Carolina, 3.3%
9. North Dakota, 3.8%
10. Florida, 4.0%
11. Montana, 4.5%

Electorally, the 39 states and DC that lie outside this 5-point range: Obama 252, McCain 157.

For what it's worth, there are twelve states between 5 and 10 points, and each candidate has 6 of them. Obama's are Colorado (+5.1%), Pennsylvania (+5.7%), New Mexico (+6.3%), Wisconsin (+6.4%), Iowa (+8.7%), and New Jersey (+9.4%), for a total of 67 EVs. McCain's are Alaska (+6.1%), South Dakota (+6.6%), West Virginia (+8.2%), Louisiana (+8.9%), South Carolina (+8.9%) and Georgia (+9.9%), for a total of 43 EVs.

What the 5-10% group tells us is that while Obama is currently pressing McCain hard in the under 5% group, if McCain could apply an across the board 2% adjustment in his favor in every state, suddenly there would be 11 states under 5% and McCain would hold a lead in seven states ranging in closeness from Ohio (+0.4%) to Indiana (+3.7%). Obama would hold a lead in four states ranging from Colorado (+3.1%) to Wisconsin (+4.4%) with Pennsylvania and New Mexico in between. And Obama would lead 207-205 electorally in the states outside the 5% margin.

If we want to arbitrarily mark the line at 8%, Obama leads 207-151 electorally with 17 states being inside that margin. Obama leads in 7 of those states that hold 86 EVs, and McCain leads in 10 with 94 EVs.

Obviously, the state races will change over time with some tightening and some being put away. Organization, energy and GOTV programs will make an impact, but we can only speculate right now as to which campaign will take better advantage of the ground game.

What's especially interesting about the close states is what a cultural range those states represent. Montana, New Hampshire, Virginia, Michigan, Nevada and Florida are all culturally such different states that it seems obvious the battle lines in this race haven't settled in yet. In a month or two after the country has had some time to size these two candidates up side by side, we should start seeing some stronger patterns emerge. But for now, let's be a little thankful that this isn't yet another dreary three-state race (OH, FL, PA).

UPDATE: With Nate's updates and new polling data tonight, some of the numbers have shifted. There are now twelve states under 5 points, as New Mexico became closer (+2.5%). Here's the new list, with the 38 + DC states outside the margin: Obama 247, McCain 157.

1. Missouri, 0.3%
3. Indiana, 0.7%
4. Michigan, 1.6%
5. Virginia, 1.8%
6. Florida, 2.0%
7. New Mexico, 2.5%
8. New Hampshire, 2.6%
9. North Carolina, 2.7%
10. Ohio, 2.9%
11. North Dakota, 3.2%
12. Montana, 4.2%

## 6.16.2008

### Today's Polls, 6/16

The headliner tonight is in Virginia, where Rasmussen has Barack Obama moving into a tiny, 1-point lead after having trailed John McCain by 3 points in May. Virginia has been a target of Democrats from the get-go in this campaign, but it's of particular importance to Obama because it enables another series of parlays for winning the election. If Plan A was holding the Kerry states and winning Ohio, and Plan B was winning Iowa and 2-3 of the Southwestern swing states, Plan C is winning the Kerry states plus Iowa and Virginia, which would get him to 272 electoral votes.

Two other polling results also show a bounce for Obama. In Kansas, John McCain leads by 10 in a Rasmussen poll; he had led by 22 a month ago. And in New York (which is polled way more than it needs to be), Obama leads by 18 in a Siena poll, up from 11 last month. If you mapped the states out in n-dimensional space according to their demographic characteristics, New York, Virginia and Kansas would form something of an equilateral triangle. Obama has received a pretty significant bounce in each of them, suggesting that his uptick is fairly widespread.

### Popular Vote v Electoral Vote

This might be my favorite graph that we've done so far: a comparison of Barack Obama's popular and electoral vote totals across the first 1,000 simulations that we ran last night:

Several interesting things to point out:

1. The relationship between the popular vote and the electoral vote is approximately linear, except at the endpoints. As a rule of thumb, a gain of one percentage point in a Obama's popular vote share results in a gain of 25 electoral votes. This is also, you will note, a pretty steep slope. If Obama wins the election by 4 percentage points, he projects to win by approximately 100 electoral votes (319-219).

2. The regression line crosses the y-intercept at 269.3 electoral votes, which is almost exactly half of 538. That means that there does not appear to be any systematic advantage in the electoral vote math to one candidate or another, at least based on our present rendering of these numbers.

3. Where you do see a little bit of skew are those scenarios where one candidate wins by about 5-15 percentage points. In those cases, the winning candidate tends to win by more electoral votes than is predicted by the regression line. This is because an especially high number of states are within reach for one or another candidate. In contrast to 2004, when 16 states and the District of Columbia were decided by 20 or more points, very few are polling that way this year.

4. The range of possible outcomes given any specific value of the popular vote is about 80-100 electoral votes wide. For example, an Obama win by 5 percentage points could easily be associated with any number from about 290 electoral votes up to as many as 390, depending on how the individual states shake out. Likewise, for any given value of the electoral vote, the range of the popular vote margin is about 6 or 7 percentage points wide. What this means, among other things, is that it's virtually impossible for a candidate to win the electoral college while losing the popular vote by more than about 3 or 3.5 percentage points.

### Weekend Poll Catchup

When my friend Geoff and I went to debate camp together (yes, there is something even dorkier than writing about polls), Geoff had a screensaver with an obnoxious lime green background and that pronounced in some very tacky, Windows 3.0 kind of font: "SLEEP IS THE ENEMY". I don't quite feel that way myself, but lately I've begun understand where Geoff was coming from. We were so busy rolling out methodological changes over the weekend that we didn't bother to document the latest polls. So let's see what we've got on the polling front.

In Arkansas, Rasmussen has Barack Obama closing to within 9 points. While this is hardly a fantastic result for a Democrat in Arkansas, it does represent an enormous improvement from Obama's prior polling in the state, which had shown him down by as many as 29 points. This result is easier to understand in light of our analysis of Obama's bounce, which seems to be concentrated in states with strong Democratic party identification but relatively few African-Americans -- Clinton country, in other words. Can Obama actually close Arkansas to the point where it becomes competitive? If Hillary Clinton is his VP choice -- Clinton still gets much more of a home-state effect in Arkansas than her adopted home of New York -- the state is probably in play. Otherwise, it probably isn't, although if the Clintons are particularly vigorous in their campaigning, it might get teasingly close.

In Oregon, Rasmussen has Obama ahead by 8. This poll has been cited by some McCain supporters a a sign that Obama's bounce is waning, since Rasmussen's last poll had him up 14 there. However, that May poll had been taken right as Obama had begun to campaign toward that state's primary; Rasmussen's March poll had shown him up by 6. No matter how you interpret the trend, Oregon looks pretty safe for Obama, but given Clinton's paucity of support in the state, there was perhaps less room for Obama to gain further ground there.

The New York Times has decided to poll the Empire State, which shows Obama leading John McCain 51-32. Although this is the Times' first poll in New York this year, Obama's 19-point margin as large as in any New York poll conducted this year. So once again, the same pattern: big Clinton state, bigger bounce.

Finally, in Nevada, a Mason-Dixon poll shows John McCain ahead by 2 points, 44-42. Mason-Dixon had last polled Nevada way back in December, at which point McCain led by 6. Although you have to go back to March to find a survey where Obama led in Nevada (the state is notoriously hard to poll and hasn't been polled much), the state's demographics should be relatively friendly to him, and he should theoretically be on the right side of the Yucca Mountain issue. With that said, in most of the scenarios* where Obama wins by way of the Southwestern swing states, he only needs two out of three to win, so Colorado and New Mexico may remain his first targets.

* For example, Kerry states + Iowa + Colorado + New Mexico is a winning combination; if he wins those states but loses New Hampshire, the election ends in a 269-269 tie that would probably be resolved in Obama's favor in the House.

### A Refinement to the Adjustment, Part II

The principal criticism of the Trend Adjustment that I introduced on Saturday is that it assumed that the trend was uniform across all states. Even if we can demonstrate that Barack Obama has gained, say, 3 points in his polling on average, and even if that average was taken across a fairly robust group of state and national polls, it might not hold that the bounce would be felt the same in Utah as it might be in Massachusetts.

I agree entirely with this criticism in theory. I would also argue that it is probably better to assume a uniform trend than no trend at all. The polling has become dense enough (particuarly if we include national polls) that we're getting a pretty fair mix of state and national polls in any given week. It is unlikely that Obama could improve his position in say 10 out of 12 state polls, and 5 out of 6 national polls, without his also being likely to have improved his position in other states that weren't polled during this period.

Nevertheless, it would clearly be best if we could have our cake and eat it too: adjust for the most recent trends (in a somewhat cautious way) without having to take some of the state-by-state specificity out of our model. I think I've developed a reasonable way to accomplsih that.

The basic way that we developed the trend estimator was to express each polling result as a combination of two dummy variables, one representing the state/pollster combination (e.g. "Quinnipac-Florida" or "Zogby-Delaware") and the other the week in which the poll was conducted. Each poll in our database can thus can be expressed in the form of a regression equation:

...'Margin' represents the polling result (Obama's total less McCain's), whereas the squiggly little 'e' you see is a term denoting the residual error/uncertainty. Technically speaking, there are coefficient terms on the two dummy variables, though over the long run, these coefficients will by definition equal one. Likewise, the error term will definitionally equal zero over the long run. However, just because the coefficients equal one on average does not mean that they do so in every single case. Another way to express our regression would be to embed the uncertainty term in the time-trend dummy, as follows:

In this equation, m represents a multiplier on the weekly trend variable. It is trivial to solve for m.

In a state which is more impacted by a time-defendant trend, m will be greater than one. In a state that is less impacted by the trend, it will be less than one.

Once we have a derived an m for each poll in our database, we can then regress it against a series of demographic variables in the state where the poll was conducted to see whether there is any pattern to the residuals. Since our particular concern is with recent trends, we weight recent polls much more heavily when conducting this analysis. (A couple of technical notes: we discard any cases in which the pollster has polled the state just once, as m will always be one in these cases. Also, we discard cases where the weekly dummy is a very small number -- anything less than one, in fact -- as this can produce very large, highly erratic values of m).

The demographic regression that I perform on m includes relatively few variables. This is because there aren't all that many useful data points to work with -- we need very recent polls, and for those polls to have been conducted in a state that the pollster surveyed previously -- so there is more risk of overfitting the model. The particular variables we include are a state's partisan ID index, its Kerry vote share in 2004, its black population, its Hispanic population, its average per capita income, its percentage of senior citizens, and its percentage of evangelicals. With the exception of the Kerry and 'partisan' variables, which are too fundamental to the model to be excluded, these variables have the virtue of not being strongly intercorrelated with one another.

As it turns out, there are some patterns in where Obama's bounce is showing up. It is coming in states where Democrats have a strong party identification advantage (no surprise), and seems to be especially strong in states where many voters are registered as Democrats, but where John Kerry did not perform well in 2004. This particularly describes states like West Virginia and Arkansas, where Obama's numbers have improved significantly, and where (assuredly not coincidentally) Hillary Clinton also performed well. The other observable trend is that Obama's bounce has been larger in states where there are not a lot of African-American voters, simply because there are few marginal gains for him to make among that demographic. It will probably always be the case in this election that states with lots of African-American voters will be less responsive to trends in the polling numbers.

This demographic regression allows us to estimate a unique value of m for each state. I cap the values of m at 0.0 and 2.0, respectively. The average value of m will not necessarily be 1.0, as it could be the case that particular kinds of states are especially predisposed to a bounce, and those states have also been polled more frequently (in fact, this does appear to have been the case to a small degree over the past couple weeks). The present m values for some representative states are as follows:
`Kentucky       1.98Arkansas       1.93Massachusetts  1.76Oklahoma       1.66New York       1.37Michigan       1.05North Carolina 1.01California     0.97Pennsylvania   0.93Florida        0.71Nevada         0.70Ohio           0.54Arizona        0.29Utah           0.00`
In adjusting our polling numbers, we take the trend from our LOESS estimator and multiply it by m. For example, say that our LOESS curve estimates that Barack Obama is polling 3 points stronger now on average than he was three weeks ago. If we take a 3-week old poll from Kentucky, we will adjust it upward (toward Obama) by (3 x 1.98) = 5.94 points. In California, we will adjust it by (3 x 0.97) = 2.91 points. And in Arizona, we would adjust it by only 0.87 points.

Taking into account the sensitivity of individual states to time trends produces a slightly less impressive result for Obama than we had been figuring on over the weekend, as his bounce seems to be most profound in states where he was already well ahead (like Massachusetts), or where he is probably too far behind to catch up (like Oklahoma). Still, we have seen at least some bounce for Obama across a large and relatively diverse array of states, and can expect to see that trend manifested in other states where new polls will come out unless his bounce begins to recede nationally.

### A Refinement to the Adjustment, Part I

In consideration of everyone's feedback, I am making two refinements to the timeline adjustment that I introduced yesterday.

The first refinement is to slightly dampen the effect of the timeline adjustment at the endpoints of the curve. The second is to use a state-specific timeline adjustment, rather than a one-size-fits all model. I will describe the first adjustment in this post.

Before I continue, I want to make clear what the goal of this project is. I want to provide you, at any given moment in time, with the best possible projection of what's going to happen in the November election. This is inherently a forward-looking exercise. If what you're interested in instead is simply a summation of what the polls are telling you now, there are plenty of other websites that can provide that for you. I do require that the projections be based on objective and quantifiable evidence. For example, I'm not going to say: "McCain is awful on the campaign trail, and people don't realize it yet. Let's take 5 points off his averages". Nor am I going to say "I heard from a well-connected source that the Republicans have put together a devastating attack ad on Barack Obama. We'd better cut his win percentage by 10 points". But that doesn't mean I'm going to limit myself to simply averaging the current polls.

* * *

In the long methodological discussion that we have had over the past couple days, there is one important point that hasn't been raised. Suppose you grant me that my timeline adjustment does an essentially optimal job of telling you what would happen if the election were held today? Does it necessarily follow that that the best projection of what would happen if the election were held today is also the best projection available to us of what would happen if the election were held tomorrow?

In other words, suppose that we are holding an election for the President of Hell. The candidates are Gary Condit and Mark Foley. In June, Foley leads by 2 points. In July, Foley leads by 5 points. What is our best possible projection in July of what the outcome will be in November? There are three possible answers to that question.

1. The random walk hypothesis. There is no way to guess whether the polls will move upward or downward in any given future period. Therefore, if a candidate's current lead in the polling is 5 points, our best guess at the eventual election outcome is 5 points.

2. The bounce hypothesis. Polls have some tendency to regress back to the mean established in previous periods. Therefore, if a candidate leads by 2 points in June, and by 5 points in July, our best guess is that he will probably finish somewhere between 2 points and 5 points ahead.

3. The trend hypothesis. This is sort of the opposite of the bounce hypothesis. Polling from previous periods does tell us something, but those polls are inversely related with the eventual outcome. So if Foley leads by 2 points in June and 5 points in July, that is evidence that he is trending upward, and is likely to eventually win by some number greater than 5 points.

I've tried to produce an answer to this question in several different ways, revisiting it this weekend by using Andrew Gelman's dataset. In some cases, like in 1988 or the summer of 1992, when the movement in the polls was fairly unidirectional for long periods of time, the more recent your poll was, the better off you'd be. In other cases, like in 2000 and 2004, the polls tended to oscillate, as though regressing back toward the mean; a bounce was usually just a bounce.

We can model this more formally by using different LOESS curves. The smoothness of a LOESS curve is determined by something called the smoothing parameter. A smoothing parameter of .7 or .8 will give you a very conservative curve that reacts slowly to new information (put differently, it still places some value in old information). A smoothing parameter of .3, on other hand, will give you an extremely volatile curve that gives a strong presumption to the most current information.

I went back and tried to evaluate whether there was an optimal smoothing parameter based on the weekly national polling averages from 1988, 1992, 2000 and 2004 (skipping 1996 because my dataset is scattershot for that year). I was looking for an answer in the following form: with X weeks to go until the general election, you will minimize your error by using smoothing parameter Y. If Y is a smaller number, like .3, that would be evidence for the random walk hypothesis or perhaps even the trend hypothesis. If Y is closer to .8, that would be evidence for the bounce hypothesis.

Unfortunately, there is no clear answer to this question. Different parameters performed better or worse in different elections, and at different points in those elections. All smoothing parameters from about .3 to .8 produced roughly the same average error when applied to the weekly polling data, with a possible exception of the two weeks immediately prior to the election, when a smaller parameter (e.g. a more sensitive curve) may be more desirable.

What this tells us is that it's frankly a judgment call as to how much emphasis we want to give to the most recent polling results. Neither the random walk hypothesis nor the bounce hypothesis can really be ruled out (we can probably rule out the trend hypothesis, however, as that would require low smoothing parameters to be demonstrably better than higher ones).

What I wound up doing was using a hybrid smoothing parameter, which is conservative toward the endpoints of the curve, but more aggressive in the middle of the curve.

There is a good, logical reason to do this, namely that we have less information available to us at the endpoints of the curve than we do in the middle. We can fairly clearly isolate the impact of something like Jeremiah Wright's first appearance on the scene, because we can look at polling both before and afterward: we see Obama's polls tumbling and then recovering. However, in trying to evaluate the polls right now, we only know what the polls were in the past; we do not know in which direction they'll move in the future. The hybrid curve allows us both to be fairly aggressive in isolating events that might have impacted the polls in the past, but also erring on the side of caution about the present direction of the polls.

The net effect of all of this is a somewhat more conservative estimate of Barack Obama's current strength in the polling; we know he's bouncing, but we don't know how long that bounce is going to last. If his polling remains strong into next week, that will be three weeks in a row where his numbers have shown a marked improvement, and even the most conservative estimator will start to give him credit for more or less the entirety of his bounce. If he and McCain regress back to a tie, on the other hand, we may even start to take a point or two away from polls that were conducted over the past couple of weeks. This is one thing, by the way, that I think some of the McCain supporters around here are missing. If Obama's post-nomination bounce does prove to be a temporary thing, we will be able to adjust for this more quickly, and recognize that states that were polled frequently during this period may not be as strong for him as they appear.

## 6.15.2008

### Convention Choices and Electoral Offense

One of the striking elements early on in the 2008 presidential race was how both major parties selected convention sites designed for offense. The Republicans chose Minneapolis-St. Paul, reflecting an ambition to make gains in the Upper Midwest. The Democrats chose Denver, reflecting their ambition to build upon recent gains in the Intermountain West.

In recent years, the Democrats chose safe locations – Boston (2004), Los Angeles (2000), Chicago (1996), and New York (1992). Not since 1988 (Atlanta) had the Democrats strayed outside their comfort zone, and before that came San Francisco (1984), New York (1980) and New York (1976).

The Republicans have generally chosen equally non-competitive territory – New York City (2004), San Diego (1996), Houston (1992), New Orleans (1988), and Dallas (1984). Only Philadelphia in 2000 marked a convention site in a swing state.

Both parties had reason to be hopeful. The once-liberal bastion Minnesota, which had been the sole Mondale state in 1984, had been surprisingly close in 2004. John Kerry only carried the state by 3.5%. Neighboring Iowa flipped blue-to-red from 2000 to 2004 by 10,000 votes, and neighboring Wisconsin was the single closest percentage state in 2004, a state Kerry carried by a mere 0.4%. Looking at the map objectively, this region seemed ripest to add new electoral votes into the Republican column.

The Democrats looked over the same red-blue stalemate map of the previous two presidential elections and chose Colorado to host its convention despite the fact that labor problems and funding concerns in Denver made New York seem the easier logistical choice. Bush beat Kerry in the state by just under 100,000 votes out of a little over 2.1 million cast, less than 5%. That margin was down from 9% in 2000. Similarly, Nevada was tightening as its population boomed, from 3.6% in 2000 down to 2.6% in 2004. New Mexico was under 1% margin in both elections.

Ironically, despite the narrow previous elections, the realistic opportunity for Dems to play offense with a state like Colorado and the Mountain West as a whole was not as clear-cut at the time Denver was chosen. When Denver was announced on
January 11, 2007, the conventional wisdom was that Hillary Clinton would be the nominee, and it’s hard to think of a worse match than the gun control poster child Hillary Clinton and the libertarian-striped Mountain West Dem brand. (Barack Obama is to Appalachia as Hillary Clinton is to the Mountain West, and the polling data backs this up.)

But now that Obama is the nominee, we see that his opportunity to win Colorado is significant. Based on 538’s projection model, Obama currently leads by 5.9% over McCain. In New Mexico, Obama projects to win by 5% even, and in Nevada by 2.4%.

On the flip side, McCain is not currently competitive in Minnesota, projecting to lose by 12%. Iowa, where McCain finished a distant fourth in his party’s caucus vote behind Fred Thompson (Fred Thompson!) looks equally dismal for Republican hopes at a 9.1% deficit. Wisconsin isn’t much brighter a prospect, where Obama projects to win by 7%.

Perhaps McCain has an opportunity in Michigan, a Democratic state since 1992. Obama has gotten off the mark slowly there, in no small part because it was one of the two states where Democrats did not campaign during the primaries thanks to geniuses like Carl Levin.

All in all, however, the Democratic opportunity to play offense with its regional convention pick seems much better than the Republican opportunity.

## 6.14.2008

### We know more than we think (Big Change #2)

The other major change to our methodology (which I am surprised nobody guessed in the teaser thread) is that we are now making adjustments to the results of all states based on a time trend.

One of the problems with our previous way of doing things is that polling data tends to roll in at different times in different states. Both state and national polls conducted since the conclusion of the Democratic nomination process have reflected a bounce of a few points for Barack Obama. For example, we know that Barack Obama has experienced a bounce in his polling results in states like Wisconsin, Michigan and New Jersey, as well as in both the Rasmussen and Gallup national tracking polls. It would be naive to assume that Obama won't also experience a bounce in other states like Pennsylvania and Ohio where new polling data has yet to come out. However, we've had no way to account for these changes in states where the polling data is not fresh.

Our objective, then, is to infer what is likely to happen in states where we don't have fresh polling data based on those states where we do. In order to make such an inference, I apply a four-step process. A version of this process was suggested to be by Professor Robert Erikson of Columbia University, who has spent his lifetime studying polling and public opinion, and who is also a family friend.

Step 1: All polls are placed into groups based on (i) the week of the election; and (ii) the state-pollster unit. A state-pollster unit is a combination of a particular state and a particular pollster; for example "Alabama-SurveyUSA" or "New York-Quinnipiac". The current week is defined as having begun seven days before the current date, with weeks progressing backward from there to the start of the calendar year 2008. One very important note: we treat national polls as a "state". For example, there are units for "USA-Rasmussen Tracker" and "USA-Gallup Tracker". One of the most useful elements of national polls, and particularly national tracking polls, is that they provide a robust baseline for measuring changes in candidate support. We do not include national polls directly in our averages. We do use them, however, to help infer trends, which in turn can inform our state-by-state projections.

Step 2: We run a linear regression with a large number of dummy variables. Specifically, we include one dummy variable for each week, and one dummy variable for each state-pollster unit. The coefficients of the weekly dummy variables give us an inkling of a time trend. Specifically, the time trend looks like this:

Let me explain exactly what is going on here. Suppose that in that in Week 15, Rasmussen shows Barack Obama 6 points ahead in Minnesota. Then, in Week 22, it shows him 9 points ahead in Minnesota. This is a piece of information implying that Obama's standing was 3 points better in Week 22 than in Week 15. If we apply this process to all state-pollster units, we get quite a lot of information about in which way the polls are changing. That's all that this process is doing. It's taking the changes that we see in each poll where we have a baseline for comparison, and inferring an overall time trend based on those changes.

Step 3: The time trend is smoothed by means of a LOESS regression. You probably don't think you know what a LOESS regression is, but if you've ever been over to Pollster.com, you have seen one. A LOESS regression is way to create smooth curves through time series data. In our case, that curve looks like this:

When running a LOESS regression, one may choose a "smoothing parameter" that determines how sensitive the regression line is to changes in the data. I use a fairly conservative smoothing parameter, tending toward a smoother rather than a jerkier curve. Nevertheless, we can make out a few fairly clear trends. Obama's numbers surged in February, when he was winning one primary after another. They slumped in March and early April, as stories like bittergate and Jeremiah Wright dominated the landscape. They have since been gradually improving, but particularly so in the last two weeks since he wrapped up the nomination.

Step 4: Polls from previous weeks are adjusted to match the LOESS estimate from the current week. For example, our LOESS regression line tells us that an average poll in the current week has been about 2.5 points stronger for Barack Obama than a poll in the week ending 5/17. Thus, the Quinnipiac poll of Florida taken on 5/17, which showed John McCain ahead by 4 points, is treated as though it had shown McCain ahead by 1.5 points (i.e. 2.5 points better for Obama). The idea, simply put, is to make all old data match the current polling landscape.

* * *

From there, everything proceeds as it always has. We still run a demographic regression, although it is based on the trend-adjusted polls rather than the original ones. (Also, I am now referring to our result in each state as a "projection" rather than an "average", as that nomenclature is more consistent with our process.

This adjustment presently results in an increase of about 2 points in Barack Obama's projected popular vote margin. Because a large number of states in this election are very close, this results in a somewhat dramatic-seeming change in Obama's win percentage and electoral vote projection. Interestingly, Obama's current win percentage of 64.7 percent almost exactly matches the price of Democratic contracts on Intrade, which also has the Democrats with a 64 percent chance of winning the election.

### We don't know as much as we think (Big Change #1)

There are two major changes to my methodology, which you already see reflected in the new charts and graphics that are presently on the site. This is the easier of the two to explain, so let's handle it first.

Andrew Gelman of Columbia University was kind enough to share some of his old national polling data with me. His dataset runs from 1952 through 1992. I took his data from 1988 and 1992 (before 1988, there are only a limited number of polls available), then combined it with the data I already had for 2000 and 2004, and tracked down some 1996 data in a magical place on the internet.

If you had looked back at the polls in June in the five previous election cycles, what would you have found?

In 1988, Michael Dukakis was ahead by an average of 8.2 points in 5 June polls. In November, George Bush won by 7.8 points.

In 1992, George Bush was ahead by an average of 4.9 points in 14 June polls. In November, Bill Clinton won by 5.6 points.

I don't actually have any June polls for 1996 (if anybody's sitting on a big stash of Clinton-Dole data, you know where to find me). But in Gallup's July poll, Bill Clinton led by 17 points. In November, Clinton won by 8.5 points.

In 2000, George W. Bush was ahead by an average of 4.7 points in 14 June polls. In November, Al Gore won the popular vote by 0.5 points.

In 2004, John Kerry was ahead by an average of 0.9 points in 16 June polls (this was pretty much his high-water mark all year). In November, George W. Bush won by 2.4 points.

So in four out of the last five elections, an average of June polls would have incorrectly picked the winner of the popular vote. That's kind of a problem for anybody who is overly confident about how this election is going to turn out.

Previously, I had modeled the error in our polling averages based on 2004 data (simply because that's the data I had access to). The issue with that is that the polls were unusually stable in 2004. From April onward, John Kerry never held a lead of more than about 2 points in the Real Clear Politics national average, and George W. Bush never held a lead of more than 6 or 7 points. Those numbers pretty well framed the actual result of Bush +2.4. But as the Gelman data reveals, there was much more fluidity in previous years, and so modeling the error based on 2004 data alone would lead one to underestimate the degree of uncertainty inherent in a general election.

My error estimates are now modeled instead on the 1988-2004 dataset. I do give somewhat more weight to more recent cycles, as in general, the polls have tended to get closer to the actual margin more quickly in recent years. There are a few reasons to think this might not be an accident. For example, (i) the country has tended to get more partisan over time, meaning that there may be fewer true undecided voters than there used to be; (ii) with the proliferation of the Internet and cable news, voters now have more information about the candidates sooner than they used to, and (iii) the science of polling has probably improved over time. Nevertheless, we are accounting for quite a bit more error than we had been before.

If I look at the total miss for each poll based on the number of days until the election, I get the following, very pretty graph:

There is quite a lot of noise there, but the error can be modeled reasonably well as a function of the square root of the number of days until the election. Specifically, the curve I use looks like this:

Presently, with about 145 days to go until Election Day, we would anticipate that a typical national poll will be off my around 6-7 points. We do not know, unfortunately, which direction that miss is likely to be in. But there is reason to believe that the range of possible outcomes -- including scenarios where the election doesn't turn out to be especially close -- is wider than we had been assuming before.

### Major Update Coming

I'm implementing a couple of significant changes to our methodology. Don't flip out. Everything will be explained in due time.

### McCain's Atheist Problem?

According to Gallup, John McCain trails Barack Obama by 25 points among voters for whom religion is not "an important part of [their] daily life". McCain leads by 5 points among those who answer that question in the affirmative.

These sorts of numbers are generally described as a problem for the Democratic candidate. However, as Ruy Teixeira pointed out four years ago, if you had to pick a sign of this divide to be on, it might be on the side of the secular. That is because by almost all indicators, religious participation in the United States is decreasing. According to a Pew poll, 45 percent of Americans now completely agree with the statement that "prayer is an important part of my daily life", down from a peak of 55 percent in 1999. (There does appear to have a bit of a "God Bounce"/mini-revival in the mid-late 1990s -- not so much in the number of religious Americans, but in the activity and enthusiasm of those that do practice).

Moreover, the younger generation is less religious than the older generation. 19 percent of those born after 1977 say they are atheist or agnostic, as compared with 11 percent of Boomers (born 1946-1964), and 5 perecnt of pre-Boomers (born before 1946).

Barack Obama, of course, does need to at least hold his own among actively religious voters, who constitute 65 percent of the electorate according to Gallup. He is able to do so thanks to substantial support from African-American and Latino voters, while trailing McCain by 25 points among actively religious, non-Hispanic whites. Nevertheless, if these generational trends hold, then each year a coalition based on actively religious voters will become marginally less successful.

## 6.13.2008

### Tim Russert

Rest in peace, big guy.

It's better to ask one question too many than one question too few.

(Added 4:19 PM) I'm still trying to collect my thoughts about this. Everyone knew, I think, that Russert was a big deal. But this might also be one of those cases where people didn't realize quite how big a deal until after his passing. In particular, the institution of Meet The Press -- precisely because it asked such tough questions and presented such real risks for candidates -- also gave them the opportunity to shift the national media narrative in a way no other news program could.

### CNN Video

I've done sports radio fairly regularly over the past several years, and a small handful of television appearances on regional cable. But this CNN gig was a different thing entirely. One thing I didn't fully appreciate is how labor-intensive all of this is. Morning shows are intended to give the appearance of spontaneity -- and to some extent, because of the fast, almost caffeinated pace, they are. But there is also an enormous amount of due diligence involved. Between the anchors and the guest and especially the producers, there are probably 100 minutes of preparation for every minute of live television that you see.

### Today's Polls, 6/13

The Republicans have given every signal of wanting to make a play for Minnesota. Their convention will be held in St. Paul, and Tim Pawlenty is perhaps the odds-on favorite to become John McCain's Vice Presidential nominee. There just isn't much indication, however, that the state is liable to be competitive.

Rasmussen's newest poll in Minnesota has Barack Obama leading John McCain by 13 points. This is technically not a bounce: Obama led by 12 and 13 points in Rasmussen's April and May polls, respectively. But Minnesota also does not appear to be close enough where little things like the selection of Pawlenty as McCain's running mate would matter (I'm sitting on some research about this, but the home state advantage of a VP selection is not all that it's cracked up to be). Indeed, the entire Northwest quadrant of the country -- draw a line from the southern tip of Illinois everywhere northward and westward -- has polled extremely well for Barack Obama, both absolutely and relative to John Kerry.

Rasmussen also has polling out in North Carolina, where John McCain holds on to a slim 2-point advantage. This result is not entirely surprising, as several polling firms have shown North Carolina within the margin of error at some point in the cycle. Barack Obama has every reason to give North Carolina a try -- the Research Triangle portion of the state might go for him 3:2 or even 2:1. But at some point, he's going to want to show an actual lead in the polling there, lest it become a tease state like the Republicans have had with New Jersey.

In Oklahoma, a Research 2000 / DailyKos poll has John McCain leading by 14 points. This might actually be Obama's best result of the day, as other Oklahoma polling had shown McCain ahead by as many as 40 points. Obama won't win Oklahoma, but the internals of the survey -- which show a bare plurality of Oklahomans identifying their party ID as Democrat -- are a reminder of just how difficult the partisan landscape is for John McCain.

Finally, I wanted to announce that FiveThirtyEight will be partnering with Rasmussen Reports and providing them with our state-by-state averages for inclusion in their Balance of Power Calculations. Between that and my appearance on CNN a bit earlier (video if and when it becomes available), I'm starting out my day pretty wired.

## 6.12.2008

### Etc. Etc. Etc.

I not a big fan of catch-all threads, but a few items of note:

1. There is a Kansas poll out showing John McCain with just a 4-point lead over Barack Obama in the Sunflower State. But we're not going to be listing it as it was an internal poll conducted for Jim Slattery, the Democratic candidate for US Senate. The problem with internal polls is not necessarily that they're bad polls, but that they only become public knowledge under certain circumstances. Slattery is trying to raise money and to build enthusiasm for his campaign -- to some extent, he is competing against other Democrats in order to do so. I'm sure that this poll was conducted honestly and fairly, but would a (strongly-worded) press release have been put out if Slattery had been trailing Pat Roberts by 30 points instead of 12? Probably not. For this reason, we do not list internal polls conducted on behalf of a candidate.

2. I had a longer item posted previously about one of the controversies described at Barack Obama's Fight the Smears. After about five minutes, I decided the issue was old news and took it down. So for the couple dozen of you who might have remembered seeing the post, rest assured that you aren't going crazy.

I did, however, want to encourage people to read Jane Hamsher at firedoglake about the matter of Larry Sinclair and to sign her petition. I am not much of a viral action kind of guy by default, but some things are so ridiculous as to merit an exception.

3. It does look like the CNN interview is going to go off tomorrow, barring further breaking news or natural disasters. My time slot is approximately 8:20 AM Eastern.

### World's Simplest Election Projection

My latest ZohoSheet project. This will give you a simple popular vote projection for November based on (i) a candidate's support from each party and (ii) turnout rates.

A couple of notes: the party ID values you see here are based on a combination of three recent surveys: Gallup, Rasmussen , and Pew. All three show the Democrats with between a 9 and 10 point partisan edge on the Republicans, although the surveys differ somewhat in how many independents they identify.

I recommend that you not play with the party ID numbers, since those numbers move glacially and are at least somewhat exogenous to the political contest in any given year. Instead, you can manipulate the turnout rates. I have Democrats and Republicans each turning out at 62 percent, with independent turnout being slightly lower. I believe this is roughly what the turnout rates looked like in 2004, although I can't find any hard evidence on that.

The other numbers you see in the worksheet, while not coming from anywhere in particular, are not entirely arbitrary either. Fundamentally, it is very challenging for McCain to be working from this deficit in partisan ID.

That is a picture of the signage at an extremely popular (and very good) taqueria in my neighborhood in Chicago. They have a conspicuously large sign which is visible from a major intersection and is probably worth a couple thousand bucks a month in advertising impressions, and have decided to devote roughly the bottom third of it to a homemade testament to Barack Obama.

Not that this anecdote should mean very much in the larger scheme of things. Obama barely carried my home district, the ridiculously gerrymanderd, Hispanic-majority IL-4, in the Illinois primary in February.

The Latino experience in America is primarily still an immigrant experience, and immigration boomed in the 1990s when the Clintons were in office. A lot of Hispanics who are now becoming citizens, or who are now registering to vote for the first time, came to America then, or had friends and family who did.

But I've been saying for a long time that one should not confuse the outpouring of support among Latinos for Hillary Clinton in the primaries with a lack of support for Barack Obama in the general election. Voting is intrinsically relativistic. I've made this comparison before, but would Hillary Clinton be regarded as running strong among Catholics if her opponent were John F. Kennedy?

I mention this because a couple of national polls -- the new NBC-WSJ survey as well as Gallup's extensive May tracking data -- now show Obama with roughly a 2:1 lead among Hispanic voters. Exit polls had John Kerry winning Hispanics by only about 55-45 in 2004 (although that figure is disputed), so this is a gain the Democrats are making right at the same time that Latinos begin to vote in much greater numbers.

One saving grace for McCain: Hispanics do not vote monolithically. Mexican Americans vote differently from Puerto Rican Americans from Cuban Americans. And so we should not necessarily assume that Obama's numbers are going to improve among Cuban voters in Florida.

### Today's Polls, 6/12

Beginning today, I will be cross-posting the daily polling thread at The Plank. For those of you who don't know me, I am the proprietor of FiveThirtyEight.com, which is sort of a self-help group for polling junkies. Most all of the rest of my blogging will remain exclusive to FiveThirtyEight, except when I feel like making fun of Jonathan Chait. We are, however, also contemplating a weekly, graphics-intensive feature in TNR's print edition.

It's a good day to get started, because the pollsters are up bright and early. Yesterday, we noted that Obama had experienced about a 5-point bounce in his state-by-state polls since Hillary Clinton's withdraw from her campaign, and today we are continuing to see some favorable results for him in other states.

In Wisconsin, Obama leads John McCain by 13 points in a University of Wisconsin / WisPolitics.com poll. Strictly speaking, this is the debut edition of this poll, and so we have no trendlines against which to compare. But the poll is conducted by Charles Franklin of pollster.com and his colleague Ken Goldstein, and so should be pretty solid. The continuum of Midwestern states goes something like Michigan- Ohio- Pennsylvania- Wisconsin- Iowa- Minnesota in order of most competitive to least competitive (one can argue that the order of Michigan and Ohio should be inverted). In each of these states, the Democrats have a pretty strong advantage in terms of party identification, and Wisconsin and Pennsylvania are the two that might come off John McCain's board if Obama's bounce has some legs.

Meanwhile, Rasmussen shows Obama with an 18-point lead in Washington. We have gotten used to seeing double-digit leads for Obama on the West Coast, but this is nevertheless an improvement from his 11-point lead in Rasmussen's May poll. We now show Obama as having a 98 percent chance of winning Washington. For the sake of comparison, Obama is roughly as likely to win Mississippi or Wyoming as he is to lose Washington.

In Massachusetts, a Suffolk University poll shows Obama with a 23-point lead. While it's not intrinsically surprising to see a Democrat with a large lead in Massachusetts, the state had not been polled that much, and one of the two pollsters who had polled it (SurveyUSA) was showing a relatively close race. Massachusetts has a lot of Hillary Clinton supporters, so it should not be surprising to see Obama's numbers improve there as he consolidates their support.

The modest exception to all of this is in New Jersey, where Quinnipiac shows Obama with a relatively tepid 6-point lead; Obama had led by 7 points in Quinnipiac's February poll of the Garden State. Other New Jersey polling has shown Obama with a somewhat larger lead. Whether the state becomes a fall battleground may depend as much on the Senate race, where some polling has shown Frank Lautenberg surprisingly vulnerable, as anything that takes place at the Presidential level.

Overall, our simulations give Obama a 54.9 percent chance of winning the election; this is his highest figure since March 18. As new polling begins to roll in from states like Ohio and Pennsylvania, that lead is likely to get larger before it gets smaller.

UPDATE: More late-bouncing developments in Iowa, where Rasmussen has Obama ahead 45-38. That 7-point lead is an improvement from a 2-point lead that Obama held in Rasmussen's prior Iowa poll.

## 6.11.2008

### 'Zona Defense

Well, it worked. Ever since the McCain campaign put Arizona in a festive pink on its electoral map, Democratic blogs are starting to pick up and run with the notion that Arizona is a swing state. The warrant for the claim is pretty thin -- an internal poll leaked by a Democratic congressional candidate showed McCain ahead by just 5 points in an R+6 district. Against that, we have a relative abundance of statewide polling showing Obama no closer than 9 points (and often down by more than that) and the fact that it's McCain's home state.

The home state effect seems generally to be on the order of 6-7 points, but can vary a lot from candidate to candidate and state to state. It might be weaker for McCain than for some other candidates, as he tends more to be a United States Senator than someone who porkbarrels for his home state.

Still, even without that home state effect, Arizona would lag a few points behind pickup opportunities like Colorado, New Mexico and Nevada, largely because of its large retiree population. Is it possible that Obama could win Arizona? Sure (although our model assigns him only a 4 percent chance). Is Arizona likely to make the difference between winning and losing the election? Probably not.

Just for our collective edification, however, you can construct some pure Western strategies for Obama that come up with winning electoral margins. Let's say that Obama loses Ohio and Michigan, but wins Iowa, and everything East of the Mississippi goes as it did in 2004. If Obama sweeps the small "sorta, kinda, maybe" Western states that are polling in the single-digits -- these are Montana, Alaska, and North Dakota -- plus the more talked-about Colorado, Nevada and New Mexico, that would get him to 271 EV. So would Colorado, Nevada, New Mexico and Arizona. But we're a long way from being able to declare the state competitive.

### 51,199,463,116,367 (Shout-Outs)

Congratulations to Isabel Lugo, a third-year PhD student at the University of Pennsylvania, who was the first person to come up with a solution to the number of unique winning scenarios in the electoral college, as well all the people who helped Isabel to get there and confirm the result. To my surprise, this became the most commented-upon thread ever at FiveThirtyEight.com, and is quite possibly my favorite.

...

In unrelated news that probably isn't quite worth a thread unto itself, I will be a guest tomorrow morning on CNN's American Morning with John Roberts. The segment will air live at roughly 8:15 AM Eastern. Postponed because of CNN's tornado coverage. This is why one should always be very cautious about touting a live TV gig. But we're trying to get rescheduled for tomorrow.

### Today's Polls, 6/11

In Michigan, Rasmussen shows Barack Obama with a 45-42 lead over John McCain. In Rasmussen's Michigan poll from last month, Obama had trailed by one point. In New York, Quinnipiac has Barack Obama ahead by 14 points. While this result is not inherently all that surprising, it does represent a 6-point improvement from Quinnipiac's prior poll of the state.

There have now been six polls that were in the field since the Democratic primaries were concluded, and for which we have a previous trendline against which to compare. Barack Obama has gained ground in all six of those polls; his average bounce has been about 5 points.
`GA    Rasmussen 5/6    -14      Rasmussen 6/4    -10NJ    Rasmussen 3/27   -1       Rasmussen 6/4    +9WI    Rasmussen 5/5    -4       Rasmussen 6/5    +2NY    Quinnipiac 4/15  +8       Quinnipiac 6/6   +14WA    SurveyUSA 5/17   +16      SurveyUSA 6/8    +17MI    Rasmussen 5/7    -1       Rasmussen 6/9    +3-----------------------------------------------------AVG                    +0.7                      +5.8`

### Tonight's Polls, 6/10

I wish pollsters were a little more consistent in when they released their data. Otherwise, there is no particularly good time of day for the polling thread. Among the more prolific pollsters, SurveyUSA and ARG tend to release their polls in the afternoon; Mason-Dixon and Quinnipiac in the early AM; Zogby in the middle of the night, and Rasmussen is all over the board. But since I don't keep a particularly consistent schedule either, I suppose that's neither here nor there.

Two polls out tonight. In Georgia, Rasmuseen shows John McCain with a 51-41 lead over Barack Obama. All of the Georgia polling has been pretty consistent, showing a lead of somewhere between 10 and 14 points for McCain. This may be because opinion of Barack Obama is pretty well polarized in Georgia (and elsewhere in the South). 34 percent have a very favorable impression of Obama -- presumably this includes the usual mix of African-Americans, college kids, and urban professionals in Georgia's burgeoning tech sector. But 36 percent have a very unfavorable view of Obama. With those goalposts set up, the polls are likely to be fairly stubborn in Georgia unless Bob Barr's candidacy gains traction (he was not included in this poll).

In Washington, SurveyUSA shows Obama with an impressive 17-point lead over John McCain. Obama is now a 97 percent favorite to win Washington.
The Washington poll is noteworthy as being the first statewide survey completed in its entirety following Hillary Clinton's endorsement of Obama. Obama had an 89-7 lead among Democrats in this poll.

## 6.10.2008

### Candidate Health Report: John McCain

Will Carroll is my colleague at Baseball Prospectus and analyzes baseball player's medical histories for a living. So naturally, when he heard about FiveThirtyEight.com, he wanted to do the same thing for John McCain. Here, then, is Will's take on McCain, which presents an essentially optimistic picture.

Let's get right past the age issue. Simple chronological age isn't a good gauge for what we're trying to look at here. McCain has kept up a normal, active Senate calendar and held his own on the campaign trail, a grueling march that never seemed to get to McCain.

Unlike Bob Dole, a comparable that many have brought up due to war injuries and an advancing age at the time of their campaign, McCain never seems to wear down. Given his workload recently, made up of mostly fundraisers and media opportunities, he's had a chance to rest that his opponent has not. Sure, Obama's relative youth and athleticism should give him some recovery advantage, but the fall campaign is not going to be the same kind of long-term grind that could wear on McCain. Focused on a few big performances and keyed to his electoral needs, McCain will be able to pick his spots.

McCain's two most significant injuries are to his shoulders and his history of melanoma. The shoulders are a visible sign of his captivity in Viet Nam, leaving the Senator unable to raise either arm significantly above his head. While it prevents a vigorous Nixonian wave of victory, it isn't noticeable and without prompting, most voters don't notice a deficit. The signs of melanoma, a puffy cheek and a long, five-inch scar on his left cheek, are far more noticeable. His appearance on "Saturday Night Live" highlighted his need for careful control of the media environment. McCain needs to be seated head-on to his audience, lit from his right and does not like to turn his head to stage right, exposing his scar. While the signs of melanoma have not recurred, the data does raise some concerns.

There was, buried in his extensive medical records, one glaring warning sign. In an article in the New York Times, Elisabeth Bumiller found that McCain's physicians had noted not one, but two melanomas and suggested some spread. Doctors aggressively removed lymph nodes - hence the large scar - and halted the spread. McCain himself knows well that melanoma can recur. He's had two other instances of melanomas being removed, once from his shoulder in the early 1990s and another on his nose in 2002, two years after the more noted melanoma removal. McCain's health and vigor eight years after the most significant melanoma is a good sign. Patients with this type of cancer have a 65% survival rate, but this is more a curve than a line, trending back up after a period of time.

Finally, McCain claims a genetic advantage and does appear to have it. His mother is 96 years old but notably vigorous and mentally sharp. The rest of his family history is less notable. His father, a Navy Admiral, lived to age 70 while his grandfather died at 61, worn down by the stress of combat during World War II.

Overall, McCain is in good shape for a 71-year-old who has been through harrowing torture and multiple bouts with cancer. McCain's most obvious comparables, Bob Dole and Dick Cheney, offer interesting lessons. Both would have had far more negative Health Reports during their campaigns, but both are alive and well at the end of their terms (in Dole's case, the hypothetical). History is not destiny, but neither is destiny predictable. Age will surely be an issue, though health it appears, should not be.

### Homework Assignment

I was asked this question by a highly-respected political writer and couldn't come up with any convenient way to provide him with an answer. Nor does there appear to be any guidance on Google. So let me pose it to the collective:

How many unique ways are there to acquire at least 270 electoral votes without any excess?

For example, one combination would be to win California, Connecticut, the District of Columbia, Hawaii, Illinois, Maine, Massachusetts, Michigan, Minnesota, New Hampshire, New Jersey, New York, Ohio, Oregon, Pennsylvania, Rhode Island, Vermont, Washington and Wisconsin. That would be equal to 272 electoral votes (not coincidentally, these are the John Kerry states plus Ohio).

Note that there are no excess electoral votes in this combination: if you remove one of the states with three electoral votes, the number falls to 269, which is below the 270-EV cut-off. So winning all of these states plus North Dakota would not qualify, since the candidate has superfluous electoral votes. On the other hand, replacing Vermont with North Dakota would make for a unique combination.

Nothing of monetary value to be provided to the winner, but I will give you a big thank you and shout-out on the front page, and your name will be immortalized in Google and possibly in a national column. I'm hoping that there's some genius out there who can solve this problem in 15 minutes.

p.s. To keep things at least somewhat simple, we probably should not worry about the split electoral votes in Maine and Nebraska.

### Liberal-Conservative Rankings Done Right

I think I'm a pretty smart guy, but every now and then I come across something and say to myself "man, that shit is deep".

This morning, in doing a little bit of remote-term planning for features that we might add to the site in the distant future, I was doing some background research on liberal-conservative scores and other methods of vote classification and stumbled across a website called voteview.com, created by a University of California at San Diego professor named Keith Poole. Voteview uses an extremely rigorous methodology for ordering Senators from most liberal to most conservative which to my mind produces some fairly intuitive results. (Five most liberal senators thus far this year? Russ Feingold, Chris Dodd, Bernie Sanders, Sheldon Whitehose, and Ted Kennedy).

This is how Voteview classifies Senators McCain and Obama over the last four Congresses; for good measure we'll also throw in Senator Clinton:
`Congress   Obama      McCain      Clinton107th      --         57/102      22/102108th      --         96.5/100    21.5/100109th      21/101     100/101     25/101110th      10.5/101   94/101      20/101`
By this method, Obama is liberal, but not that liberal. He was the 21st most liberal senator in the 109th Congress and has been the 10th or 11th most liberal thus far in the 110th. The surprising result is John McCain, who rates as the 8th most conservative senator in the 110th Congress, the 2nd most conservative in the 109th, and the 5th or 6th most conservative in the 108th.

In the 107th Congress, however, McCain was quite moderate. Voteview doesn't have rankings before the 107th, so I'm not sure whether there was some permanent change in McCain's political philosophy on or around 2003 (perhaps coinciding with the start of the Iraq War) or whether it was his behavior in the 107th that was unusual (perhaps he took some pleasure in being a thorn in President Bush's side after having lost the primary to him). But this is more evidence for the notion that the 2008 version of John McCain is a very different politician than the 2000-2002 version of John McCain.

## 6.09.2008

### The best move of the general election season (so far)

Obama to set up anti-disinformation SWAT team:
Barack Obama is recruiting senior staff to a new unit which will combat virulent rumor campaigns on the internet that threaten to cost him votes in the presidential election against John McCain.

The unit is part of a huge expansion of Obama's campaign team as he shifts from the Democratic nomination race to the campaign for November's election.
It's become hackneyed to say this, but the underlying dynamics of the election do favor Obama. We'll see if he begins to get a bounce in his state-by-state polling over the coming weeks. I think he very well might, and we may wind up talking about an endgame in which in order to win McCain will either have to (i) wait for a mistake; (ii) play perfect Election Night poker and sweep the swing states, or (iii) go negative.

But as McCain himself does not seem inclined to go negative, he'll have to rely (knowingly or not) on off-label elements on the Internet. Hence, Obama's SWAT Team (you'll have to excuse me for being a little Giordanoesque in my prose). We do not know exactly what the SWAT Team will consist of (the Obama campaign probably does not want us to know), but a safe guess is some combination of: public relations staff, law-enforcement officers, hackers, Internet security experts (i.e. more hackers), bloggers, and lawyers.

One wonders if this is a response to the recent plagiarisms and fabrications from the well-trafficked anti-Obama site NoQuarter, which gained enough traction to provoke a question from (and a repudiation of) a McClatchy reporter. If I were a proprietor of such a site, I would be thinking about retaining an attorney.

### Today's Polls, 6/9

Three new polls out this afternoon from Rasmussen. In Wisconsin, Barack Obama holds a 45-43 lead over John McCain. Rasmussen's March and May polls had shown McCain with a lead in Wisconsin, so this brings them into line with other polling in the state, all of which shows a small lead for Obama. The McCain team may need to decide relatively early on whether they want to make a serious play in Wisconsin, or concentrate more exclusively on Michigan and Ohio, each of which look somewhat more favorable to them. Both candidates get pretty good favorability scores in Wisconsin, but Obama's support is firmer, so McCain may be playing for a smaller-than-usual number of swing voters.

In New Jersey, Obama leads by 9 points. New Jersey has not been polled as much as it probably should be, but this is another case of the polls coalescing a little bit, as Rasmussen's prior polling in New Jersey had shown the state to be a toss-up. If there's bad news here, it's for Frank Lautenberg, since the poll seems to indicate that many Dems are thinking about splitting their ticket.

Lastly, in Texas, McCain holds a 52-39 lead. This is a "pre-bounce" poll, as the field work was conducted last Monday. Nevertheless, this is yet another case of Rasmussen's numbers gravitating back toward what other pollsters like Research 2000 and Baselice have found in the state. Obama may make some pretense of competing in Texas, but it will be pretense only.