Blogging live on the eve of the Pennsylvania primary. The final polling data for Pennsylvania is in from Real Clear Poilitics. They show a slight increase in the percentage for Hillary Clinton and a smaller increase for Barack Obama.

The exact percentages are: Clinton–49.5; Obama–43.4. On Sunday Clinton led 47.4% to 42.3%, which registers as a jump of 2.1% for Clinton over the last 48 hours. More importantly this also puts Clinton on the threshold of the 50+% she needs to make a creditable showing in Pennsylvania. It also puts her close to the 52% which had been her high total a month ago.

It is important to note this is polling data, not exit polls of actual voters, but it suggests Hillary Clinton has regained her momentum. The downside of these data are that they raise expectations for Clinton a bit higher than they were.

Now we will see how the votes turn out.

Early Results

Turnout is reported to be high, but an even better indication of how high is that the major search engines on the net seem to be overwhelmed by the traffic.

Curiously the first exit poll data from ABC contradict the above movement saying:

Despite all the down-to-the-wire campaigning, preliminary exit poll results indicate that nearly eight in 10 Pennsylvania voters made up their minds at least a week ago, and six in 10 decided on their candidate more than a month ago — a higher number of early deciders than the norm in Democratic primaries to date.

f you read between the lines eight in ten means 20% undecided–which is a pretty high total! I wouldn’t want ABC analyzing my data!

First Data

The Associated Press has the first data, but you have to do a little manipulating to get it as the main AP web site is jammed. I found mine at the Seattle Times. Here is what our table looks like based on these data:

These VERY early data look promising for Clinton. Essentially it shows that the voters she needed to activate showed up in high numbers. Obama will now need high totals among the young and African Americans to offset this. Don’t expect these data soon.

Union voters were the key, according to my earlier analysis. If the 30% figure is accurate it suggests a turnout double their percentage of population. That will be very hard for Obama to beat.

More Data

The Las Vegas Sun reports:

One in 10 voters changed their party registration since the start of the year so they could vote in the hotly contested primary, which was open only to registered Democrats. About half of the party-switchers had been registered Republicans and the rest had been unaffiliated with either party. Another roughly 3 percent were voting for the first time in Pennsylvania.

As usual, the early exit poll data is focusing on the issues the networks think are important but not giving us any demographic data. If these data were what you were getting back from your field people at campaign headquarters, you would fire them all.

The interesting thing is the mainstream media already HAVE the demographic data, but don’t seem to think we would be interested in it.

By the way, with a bit of a lull here before the polls close in less than an hour, here is the URL for the site that does the polling for the networks and the AP. It will allow you to click back and forth between various networks to see how they are all treating the same data set.

Edison Media Research

The use of one firm by all the major news organizations does save some money, but you have to ask what if there were another firm doing the polling? Is the fact that a single firm does the polling responsible for some of the errors in the past that have been magnified because they have been reported by everyone at the same time?

A Projection

When the polls close in a little bit, I will make an early projections based on the data we have so far and explain why. Stay tuned.

While we are waiting, the Times online reports, “Early exit poll spells bad news for Hillary Clinton.” The basis for the article is a last minute poll showing Clinton winning 52-48%. They must not be reading the same data we are.

Now that the polls are closed, based on the data we have I will call Pennsylvania for Clinton. Unless the demographic data we have are totally off or those voters have suddenly switched their voting habits, I will even suggest that she will reach the 50% I said she needed to make a respectable showing. The way things are trending, the data look quite a bit like Ohio where Clinton won 54%. In Ohio the union turnout was 34%. The turnout of voters over 55 was over 40%.

In Ohio the African American turnout was 18% and that still was not enough to help Barack Obama, which he lost by 10%. A turnout of 18% in Pennsylvania would beat the previous record African American turnout of 17%. For Obama, then, the “x” factor at the moment is the turnout of young voters, for whom we do not have any data.

BTW, remember you heard it here first. The networks will not call it because they are not looking at the demographic data. They also want to keep you watching.

Next Data Set

We now have some more complete data from CNN. To their credit they withheld this until the polls closed.

The most startling numbers in this poll are the figures for older and younger voters. CNN breaks down their groups differently than I do, but their tables show that only 27% of those under 44 voted with the remainder going to those over 44. I predict one of the questions people will ask about this campaign is what happened to Obama’s young warriors?

The overall impression of these data are that Clinton could be moving towards a win like Ohio where she won by 10%.

It’s About Time

The networks have finally decided to call the contest. Remember they knew what I knew when I called the race an hour ago. This late call was about keeping viewers on the hook as long as possible. The other interesting piece, as one analyst pointed out, is this will get Clinton on the evening news.

Some new data from CNN, but only a few changes.

The only large shift in the last hour has been the increase in the total of women voting. EMILY’s list played a big part in Clinton’s performance in Ohio and Texas. They deserve credit also for Pennsylvania.

I will post some more detailed analysis in a bit as I review some of the data from the previous primaries. The question is: does this hurt Obama and how much?

Analysis: Five Questions

1) The Delegates: The idiots at the networks, etc, have forgotten the rules of the Pennsylvania primary. Remember that the delegates are allocated by PRECINCT, not statewide. In other words, in each precinct each candidate gets a percentage of delegates equal to their percent of the vote in that precinct. Why is that relevant? From the data I have seen Barack Obama may have won as much as 80%–perhaps even more–in some Philadelphia precincts. It is unlikely Clinton won any precincts by that big a margin.

I predict that the final delegate count will be much closer than the actual vote. Pennsylvania could even end up like Texas, with Obama ending up with the most delegates–although I think that would be unlikely. I could see the delegates splitting so Clinton’s margin of victory IN THE DELEGATES is only a few percentage points. This will not be enough to substantially impact Obama’s overall lead.

2) Clinton’s Campaign: It has become pretty clear that given Obama’s delegate lead that the only way Hillary Clinton can win the nomination is to cut up Barack Obama so badly that Democrats start wondering if he can win. Make no mistake, the Clintons know this campaign is about who shall control the Democratic Party. That is why this is the most ferocious nomination contest since the Humphrey-McCarthy-Kennedy campaign of 1968 and the Goldwater-Rockefeller Republican campaign of 1964. This scorched-earth strategy does the Democratic Party no good.

3) Where are the Democratic Leaders? While Hillary Clinton has been cutting up Obama, those Democratic leaders who supposedly support him have largely been silent. This has produced the bizarre spectacle of an ex-President acting like an attack dog while the Party leadership has behaved like poodles. When someone is subject to character assassination then it is up to the friends and supporters of that person to stand up. Joe McCarthy was successful for awhile because so many people were deserted by their friends and supporters when they were blacklisted. Barack Obama cannot govern by himself. If his supporters do not come to his defense now, what will happen when he needs them for something really serious?

4) Obama’s Base: One major question remaining about Barack Obama is one I have raised before–can he expand his base? In 1996, 34% of all voters had an income under $30,000. In the Pennsylvania primary, only 20% of those voting fell into that group. I will say this for the umpteenth time: the Democrats cannot win without these voters. They just barely won in 1996 with a sitting President! I am still waiting for Obama to make that speech about justice, equality and the need for a level playing field.

Even more troubling is that in Pennsylvania, Obama’s base did not turn out. The young people who have been such a refreshing addition to the Democratic Party seem to be cooling off. This group inspired thoughts of the Obama campaign becoming another “Children’s Crusade” like 1968. Pennsylvania tarnishes that hope.

5) Troubling Data: Besides what has become increasingly clear–Clinton and Obama’s supporters do not like each other–there are some troubling findings in the exit poll data from Pennsylvania.

Some of these data suggest the GOP may have been meddling in this primary. Eleven percent said neither candidate was trustworthy. Six percent believed neither could improve the economy and less than half believed both could. But the key statistic is that eleven percent said they would vote for McCain for President! According to the exit poll data, 69% of these McCain voters went to Obama.

The other troubling thread is race. Of the nineteen percent who said the race of the candidate was important to them, 58% voted for Clinton. Which means they voted for her because she is white. If you do the math, 58% of 19% is 11%–more than the margin of victory for Clinton. Thirteen percent of whites said race was important versus only 4% of African Americans. A troubling 74% of those whites voted for Clinton.

Final Thoughts

The numbers are clear. Hillary Clinton cannot win the nomination unless she cruises to huge victories in the remaining primaries, convinces superdelegates to change their declared preference, or bloodies Obama so badly that the Party decides he cannot win. As you will see in Friday’s essay, Clinton does not have the money to mount the kind of campaign that would yield that 60%, nor has she ever won 60% in any contest. That leaves the superdelegates or the scorched strategy. Now the question is at what point does this become destructive to the Democratic Party?

Tagged with: Barack Obama • democratic primary results • equality • exact percentages • exit polls • Hillary Clinton • justice • level playing field • major search engines • Pennsylvania • Pennsylvania exit polls • polling • polling data • real clear poilitics • turnout