Ravings from a Man in an Old White House

I become more and more convinced every year that nominating caucuses are a horrible idea. For those who don’t know, caucus-goers gather at their local caucus site at a given time on a given date. Supporters of different candidates gather in different places in the caucus site. This works a lot like picking group projects in a classroom: “Everyone who’s going to read To Kill a Mockingbird, go to the back corner by the windows. Anyone who wants to read Dandelion Wine, stand by the pencil sharpener.” The difference is that select supporters travel from station to station to try to convince supporters of other candidates to join them. At the same time, people who don’t know who to vote for go from station to station to try to get convinced. At the appointed time, people make final decisions and everyone is counted. There are some variations on what happens next, but that is the basic pattern common to caucuses. If you cannot be at the caucus because you work, are an invalid, or will be absent from your precinct, you cannot vote.

In primaries, by contrast, voters go to their polling places, which are generally the same polling places they use in general elections. Times vary by state, and sometimes within a state, but polling places generally open before most people go to work and close after most people get home. In addition, invalids and people who will be out of the precinct on election day can vote absentee by requesting a ballot by a given day and getting it to the appropriate place by election day. In many states, including Illinois, voters can vote early without any excuse. The ballots cast in primaries are secret ballots, also known as Australian ballots.

I am not convinced that the ’70s innovation of choosing delegates for the national nominating conventions through binding votes of the party faithful is an improvement on the less democratic method of leaving the decision up to the state party organization. Having decided put the decision in the hands of the people, maximizing democracy in the process should be the priority. Caucuses leave out soldiers serving away from home, shut-ins, people providing care to out-of-town relatives, people who work during the caucus, and people who do not want to share their political views.

That brings us to the other anti-democratic problem with caucuses: public voting. The classroom analogy is helpful here. When a teacher takes a vote by raise of hands, some students look around before voting. Some even lower their hands when not enough others do. When teachers are careful, they use a heads-down vote. With caucuses, peer pressure can be worse. Because caucuses are so local, voter might feel pressured to vote for a candidate to avoid offending neighbors or customers or because he’s afraid his union will treat him unfairly if he doesn’t.

It’s unlikely the caucus will disappear anytime soon, but they are still a dumb way to pick a candidate.

No doubt a lot of people are wondering what Ted Kennedy’s endorsement of Barack Obama means. I believe it means a lot, especially with as powerful as the speech was. Ted Kennedy’s support means something to liberals, especially baby boomers and older liberals (like my grandmother) who voted for JFK. I’m not saying that all those older Democrats who have been voting for Clinton are going to switch, but many will stop and think about Obama, and some will switch. Also, Kennedy has access to organizations in several states due to his long-running leadership in the party.

Moreover, other Democrats who want to endorse Obama, but have been afraid to risk angering the Clintons in case she wins, now have cover. That is, if Clinton wins, Kennedy is powerful enough that he can force Clinton to forgive them. I think there are many such officials. One of the secrets of the Democratic party is that while rank-and-file members love Bill Clinton, many in the party establishment resent him. Bill Clinton is seen as a figure who valued his own success over the party’s. They point to the congressional losses of the nineties, among other things. Perhaps more importantly, the left wing of the party seethed beneath their smiles as Bill Clinton moved the party to the right, grafting weakened Republican planks onto the Democratic platform (e.g., welfare reform). I predict that many of those people will come out and endorse Obama, who far better represents the progressive agenda they wish to re-establish.

What’s more, this is an embarrassment for Bill and Hillary Clinton. The Clintons wanted that endorsement or, more realistically, for Kennedy to stay out of it. In addition, it is yet another major repudiation of Bill Clinton’s race-baiting in South Carolina. It is a rebuke of one of the larger-than-life leaders of the Democratic party by one of the others (if not the only other) and cannot just be brushed off.

In addition to the nuts-and-bolts importance of the endorsement, there is the sentimental importance. Kennedy could have given a standard “I believe this person is the best choice to be [insert office here]” endorsement. Newsies would have had some fun saying it associated Obama with JFK, but it would not have meant that to voters. Instead, Kennedy, his son, and JFK’s daughter Caroline did make that connection. Ted Kennedy went on to make thinly veiled swipes at the Clintons, responding to their methods and their specific criticisms against Obama. The endorsement settles upon Obama the mantle of a youthful, uniting, altogether Kennedy-esque breath of fresh air. At the same time, it lowers the Clinton mystique.

On a personal note, I am thrilled, almost giddy, about this endorsement for reasons that range from the high-minded to the petty. I am glad to see a movement coalescing around a figure who may actually be able to unite Americans around a common purpose. I am also happy that this happened and now, enabling Obama to pick-up more momentum heading into February 5. As I become more convinced that Romney will win the Republican nomination, it becomes more urgent that the Democrat be someone I can vote for. With Hillary Clinton’s disapproval rating as high as it is, it should be important to all Democrats to nominate someone else. Finally, I am glad to see the Clintons squirm a little. The Republican in me who saw his party out-maneuvered by these people over and over in some of the most cynical ways throughout the nineties is glad to see Democrats either realizing or admitting to what the Clintons do when you oppose them. It amuses me that that particular insecurity has survived, especially since the Republican party lost me ages ago, but there it is.

I don’t know how many times I’ve said it, but I hate not having access to the poll internals. This whole New Hampshire debacle was a result of people misinterpreting polls. There were a lot of undecideds, and Obama’s and Edwards’s numbers were soft. At least it is understandable. In Iowa, no one believed the one poll that showed Obama and Huckabee up by around 7. People tried to learn the lesson and so ignored the possibility that the undecideds could break for Hillary.

I saw Chris Matthews, who seems enamored of the theory that racist poll responders gave the politically correct answer (i.e., they voted for the black man), conceding that the pre-election polls may make sense, but he continued to question the exit polls. I think this is another example of overconfidence in polls. Polls have meaning, but they only mean what they mean. Often people try to make them mean more.

Let me explain how exit polls work. Pollsters stand outside select precincts and attempt to ask a random sample of voters how they voted. So how are precincts are selected? Just as there are swing voters and swing states, there are swing precincts. These precincts don’t necessarily vote reliably for the winner. They may just predict the statewide vote or part of the statewide vote for one or more of the parties. For instance, precinct X might always vote 25 percentage votes higher than the statewide Republican vote. To tally your exit polls, you adjust those precincts by adding or subtracting the historical factor. In my example, you would subtract 25 points.

I don’t totally get how this works in a primary. I guess the best you can do is to base your numbers on who past polling has shown your supporters to be. Exit pollsters do seem to have a method, whatever it is, and it generally works. No one is really talking about the exit poll internals, so it is difficult to judge why the were off. There may have been interpretation problems, but the problems may have been bigger.

Anecdotal evidence from one of MS-NBC’s contributors may help in figuring out why the exit polls were wrong this time. In a state like New Hampshire, you may just be able to put pollsters in enough representative precincts to try to get a good prediction of the vote. According to this person, who’s name I can’t remember, the pollsters outside her precinct did not seem professional or to know what they were doing and the scene was chaotic. This post is already getting long, so I am not going to go any further than to say that such a situation is seriously damaging to the poll’s randomness, and randomness is vital to a poll’s accuracy.

Having had a chance to sleep on it, I think I know the reason for the disparity between polls and results in New Hampshire. I guess this is just a lesson I have to re-learn every year. Polls, especially in Iowa and New Hampshire, are dicey things in primaries because you don’t know who will turn out. The name of the game is GOTV, get out the vote. Most polls in Iowa were showing a 1-3 point lead for Obama. Momentum was clearly a factor, but the fact that he was able to bus people in his home state was very helpful in getting the vote out. New Hampshire was just as much in Hillary Clinton’s back yard. In addition, the Clintons have had an organization in New Hampshire since 1992.

The polls tested people calling themselves likely voters; the results reflected people who actually voted, including those who didn’t expect to.

I still think Clinton has an advantage in Nevada; after all, she has been leading there by a lot for a long time. South Carolina is a different story. Obama will probably do well in South Carolina, so there is a strong possibility we will go into Super Duper Tuesday still without a frontrunner. Even if Edwards drops out after South Carolina, it may be difficult to judge how that affects the race in time for February 5. If Edwards has any meaningful organization in any of those states, that may help Obama, but I doubt it.

Tim Russert just said that Hillary Clinton just pulled a “major upset.” “It’s Fraiser-Ali.” Nonsense. Hillary Clinton was leading in New Hampshire on the eve of Iowa. The next day, the polls showed the race even or with an Obama edge. In the last 48 hours, the polls showed significant leads for Obama. He didn’t lose that support; he never had it. The polls were inaccurate. They didn’t reflect the facts on the ground. Somehow the pundits are having trouble digesting that point. I guess I shouldn’t be surprised; these are, after all, the same people who act as though the voting continues after the polls close. I have a suspicion that the polls we were seeing were weighted based on the Iowa numbers. In other words, independents were given extra weight or the like. I remember Zogby made an assumption (unwarranted, I believe) that undecideds break toward the challenger at the last minute, so he added them to Kerry’s numbers. As it turns out, the opposite is, as it was in that case, true.

Another possibility is that pollsters overestimated how many independents would break for Obama over McCain. Despite an influx of Democrats over the last few years, New Hampshire tends to go Republican. In addition, McCain has a history with those independents.

At any rate, and not for the first time in recent times, the polls were way off. Hillary Clinton is alive, but taken down a peg. We now have two races without frontrunners. This flummery about Romney being a frontrunner because he finished twice to two different people is silly to me. If he finishes second in Michigan, or even wins but fails to win convincingly, he’s probably done. The next few weeks will be interesting.

I’ve appended the first 5 posts from this evening to the end of this one.

66%

NBC calls it for Hillary Clinton.

62%

Clinton 2.74. Interestingly, turnout appears much more even between the parties so far. Democrats are on 159,375 votes at 62% in, while Republicans are at 128,899 votes at 60%. That is a far cry from the 100,000ish difference in the Democrats’ favor in Iowa.

For those who can’t think of a reason not to vote for John Edwards, his theme music between his wife’s excessive introduction and his concession was that god awful John Mellencamp song, “This Is My Country”. Grrr. Also, didn’t he talk about the same sick people who were screwed by insurance companies in his Iowa concession?

60%

Clinton is now leading by 2.47%. It appears that there are a lot of college votes out there, but I can’t see this ending any better than a close victory for Obama. I don’t know that you can claim Obama failed to meet expectations. The expectations were high because the polls were high. Regardless of who wins, I think the race continues, but with no clear frontrunner.

49%

Clinton by 2.77.

40%

I finally got sick of the lack of tenths places on the returns. What looks like a 1% lead can be a .1% lead with rounding, so I made my own spreadsheet. Clinton by 2.73.

30%

Clinton’s lead fell back to four at 25% and has held steady.

23%

Clinton’s lead went up by 2 points with an increase of 2% of precincts reporting. These must be heavily Clinton-leaning precincts.

21%

McCain is still up 9, and Clinton is now up 4. If this doesn’t start to reverse soon, then Obama had better have crushed Clinton in the remaining precincts. I don’t think I can stand another process stories about how all the polls could have gotten it wrong.

12%

With 12% reporting, McCain (declared the winner by MS-NBC) is up 9% over Romney; Clinton is up by 2% over Obama. This makes sense to me. The places in are probably more conservative. As such, they are places Clinton would get more votes, and more independents would go to McCain than to Obama. If I’m right, I would expect McCain’s lead to drop. We’ll see.

As of today, the night before the New Hampshire Democratic primary, it is widely predicted that Barack Obama is going to defeat Hillary Clinton for the second time in five days. In fact, most of the polls tracked at RealClearPolitics suggest that Ms. Clinton will be trounced. Her candidacy is in serious trouble, and things may be about to get worse. According to a Wall Street Journal article, several Democratic senators who have remained neutral are talking to Mr. Obama’s staff about endorsing him. Also according to that article, the culinary union in Nevada is supposedly going to endorse Sen. Obama. The article implies that the endorsement is strong enough to sway the caucus results. Another problem for Sen. Clinton is the fact that its rules are similar to Iowa’s, including the second choice voting. Losing Nevada would be really bad news for her because she had a double-digit lead there well into December. Hillary Clinton will however win at least one primary before about half the delegates (other than super-delegates) are allocated on Super Duper Tuesday, February 5, 2008: Michigan. That is not much to feel good about since she is the only serious candidate on the ballot. The fact that no delegates are at stake there makes the victory a bit more hollow.

All the same, I am not convinced that the junior senator from New York is toast. I fully admit that my reluctance to call the fight probably stems from spending eight-and-a-half years watching the Clintons pull-off improbable victories. The Clinton campaign does face difficult questions about how its candidate can win. The following are just a few:

(1) Why did Ms. Clinton’s huge national lead dissipate, and how did her significant leads in Iowa and New Hampshire become significant losses?
(2) How can she quickly redirect her campaign and change the narrative in time for February 5?
(3) Why has she not made any changes in her campaign? The leads have been slipping for quite sometime, so I’m surprised she has not fired anyone yet.
(4) If the Clinton campaign can remain viable through February 5, but the primary vote is not conclusive, will it be able to continue during the months of primaries that follow?
(5) It’s unbelievable that this is still a question, but how will she deal with her above 50% negatives when the general election comes around?
(6) This is more of a follow up, but why has she not already began to work on her high negatives?

There are a few lessons we can take from Sen. Clinton’s situation. Co-opting your opponent’s message does not always work. Her husband was able to co-opt the Republican message throughout his presidency: he consistently took the Republican position, watered it down, and scored political points. She had been co-opting Republican positions on the war and terrorism since 9/11. All of that may be genuine, but I doubt it. At any rate, Sen. Clinton had been touting her experience, but then Mr. Obama’s “change” message caught on. How did she respond? “I can bring real change.” “I have 35 years’ experience of change.” That has not worked (although she continues to try).

Also, you cannot count on your lead to hold up. You can’t just get conservative and start running the ball; you must play your game all the way to the end.

Finally, the primary calendar is flat-out weird this year. Without looking to be sure, I think Iowa and New Hampshire have been held in January and were followed by a steady stream of primaries until Super Tuesday in early March. One of the two early votes would generally begin an avalanche because the money would dry-up and the less successful candidates could not afford to continue for a 1.5-2 months. This year, half the delegates are chosen only 33 days after Iowa. The two top Democrats have each raised enough money to make it entirely plausible to stick around until the mega-election day even if the funding dries up. Winning on the fifth would be enough to start the funding spigots again.

Due to Clinton’s problems with the liberal base and everyone’s knowledge of her high disapproval ratings, coupled with her weakening position, I would not bet on her winning the Democratic nomination. Given her and her husband’s pasts, I would not bet against her, either.

Note: The writing of this post has gone late enough that I can report that Dixville Notch and Hart’s Location, New Hampshire, have given Barack Obama an early lead with 16 delegates votes to 3 each for Clinton and Edwards and 1 for Richardson. On the Republican side, it is McCain 10, Huckabee 5, Paul 4, Romney 3, and Giuliani 1. Under New Hampshire law, towns with a population of 100 or fewer to open the polls at midnight and close when everyone has voted.